diff --git a/.gitattributes b/.gitattributes
index 89ccfbd806a7cca94beeb4db6364a841b202c709..69f858f217914ed5808e48e94104bfe9c856684e 100644
--- a/.gitattributes
+++ b/.gitattributes
@@ -34,3 +34,31 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
*tfevents* filter=lfs diff=lfs merge=lfs -text
Ocean_Pexels_[[:space:]]8953474_512x512.mp4 filter=lfs diff=lfs merge=lfs -text
pexels-jill-burrow-7665249_512x512.mp4 filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/github_docs/imgs/anime_3.png filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/github_docs/imgs/anime_4.png filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/github_docs/imgs/canny_1.png filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/github_docs/imgs/inpaint_before_fix.png filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/github_docs/imgs/ip2p_1.png filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/github_docs/imgs/ip2p_2.png filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/github_docs/imgs/ip2p_3.png filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/github_docs/imgs/lineart_1.png filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/github_docs/imgs/lineart_2.png filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/github_docs/imgs/lineart_3.png filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/github_docs/imgs/mlsd_1.png filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/github_docs/imgs/normal_1.png filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/github_docs/imgs/normal_2.png filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/github_docs/imgs/scribble_2.png filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/github_docs/imgs/seg_2.png filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/github_docs/imgs/shuffle_1.png filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/github_docs/imgs/shuffle_2.png filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/github_docs/imgs/softedge_1.png filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/github_docs/imgs/tile_new_1.png filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/github_docs/imgs/tile_new_2.png filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/github_docs/imgs/tile_new_3.png filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/github_docs/imgs/tile_new_4.png filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/test_imgs/bird.png filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/test_imgs/building.png filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/test_imgs/building2.png filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/test_imgs/girls.jpg filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/test_imgs/person-leaves.png filter=lfs diff=lfs merge=lfs -text
+ControlNet-v1-1-nightly-main/test_imgs/sn.jpg filter=lfs diff=lfs merge=lfs -text
diff --git a/ControlNet-v1-1-nightly-main/.gitignore b/ControlNet-v1-1-nightly-main/.gitignore
new file mode 100644
index 0000000000000000000000000000000000000000..a0b77e4c92c6792a0174ea32a2e096b436c7701c
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/.gitignore
@@ -0,0 +1,140 @@
+.idea/
+
+training/
+lightning_logs/
+image_log/
+
+*.pth
+*.pt
+*.ckpt
+*.safetensors
+
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+pip-wheel-metadata/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+# Usually these files are written by a python script from a template
+# before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py,cover
+.hypothesis/
+.pytest_cache/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+.python-version
+
+# pipenv
+# According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+# However, in case of collaboration, if having platform-specific dependencies or dependencies
+# having no cross-platform support, pipenv may install dependencies that don't work, or not
+# install all needed dependencies.
+#Pipfile.lock
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule
+celerybeat.pid
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/
diff --git a/ControlNet-v1-1-nightly-main/README.md b/ControlNet-v1-1-nightly-main/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..a9a34ed7151b3bae92ff4bf40e4b14c6fef5460d
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/README.md
@@ -0,0 +1,620 @@
+# ControlNet 1.1
+
+This is the official release of ControlNet 1.1.
+
+ControlNet 1.1 has the exactly same architecture with ControlNet 1.0.
+
+We promise that we will not change the neural network architecture before ControlNet 1.5 (at least, and hopefully we will never change the network architecture). Perhaps this is the best news in ControlNet 1.1.
+
+ControlNet 1.1 includes all previous models with improved robustness and result quality. Several new models are added.
+
+Note that we are still working on [updating this to A1111](https://github.com/Mikubill/sd-webui-controlnet/issues/736).
+
+This repo will be merged to [ControlNet](https://github.com/lllyasviel/ControlNet) after we make sure that everything is OK.
+
+**Note that we are actively editing this page now. The information in this page will be more detailed and finalized when ControlNet 1.1 is ready.**
+
+# This Github Repo is NOT an A1111 Extension
+
+Please do not copy the URL of this repo into your A1111.
+
+If you want to use ControlNet 1.1 in A1111, you only need to install https://github.com/Mikubill/sd-webui-controlnet , and only follow the instructions in that page.
+
+This project is for research use and academic experiments. Again, do NOT install "ControlNet-v1-1-nightly" into your A1111.
+
+# How to use ControlNet 1.1 in A1111?
+
+The Beta Test for A1111 Is Started.
+
+The A1111 plugin is: https://github.com/Mikubill/sd-webui-controlnet
+
+Note that if you use A1111, you only need to follow the instructions in the above link. (You can ignore all installation steps in this page if you use A1111.)
+
+**For researchers who are not familiar with A1111:** The A1111 plugin supports arbitrary combination of arbitrary number of ControlNets, arbitrary community models, arbitrary LoRAs, and arbitrary sampling methods! We should definitely try it!
+
+Note that our official support for “Multi-ControlNet” is A1111-only. Please use [Automatic1111 with Multi-ControlNet](https://github.com/Mikubill/sd-webui-controlnet#Multi-ControlNet) if you want to use multiple ControlNets at the same time. The ControlNet project perfectly supports combining multiple ControlNets, and all production-ready ControlNets are extensively tested with multiple ControlNets combined.
+
+# Model Specification
+
+Starting from ControlNet 1.1, we begin to use the Standard ControlNet Naming Rules (SCNNRs) to name all models. We hope that this naming rule can improve the user experience.
+
+
+
+ControlNet 1.1 include 14 models (11 production-ready models and 3 experimental models):
+
+ control_v11p_sd15_canny
+ control_v11p_sd15_mlsd
+ control_v11f1p_sd15_depth
+ control_v11p_sd15_normalbae
+ control_v11p_sd15_seg
+ control_v11p_sd15_inpaint
+ control_v11p_sd15_lineart
+ control_v11p_sd15s2_lineart_anime
+ control_v11p_sd15_openpose
+ control_v11p_sd15_scribble
+ control_v11p_sd15_softedge
+ control_v11e_sd15_shuffle
+ control_v11e_sd15_ip2p
+ control_v11f1e_sd15_tile
+
+You can download all those models from our [HuggingFace Model Page](https://huggingface.co/lllyasviel/ControlNet-v1-1/tree/main). All these models should be put in the folder "models".
+
+You need to download Stable Diffusion 1.5 model ["v1-5-pruned.ckpt"](https://huggingface.co/runwayml/stable-diffusion-v1-5/tree/main) and put it in the folder "models".
+
+Our python codes will automatically download other annotator models like HED and OpenPose. Nevertheless, if you want to manually download these, you can download all other annotator models from [here](https://huggingface.co/lllyasviel/Annotators/tree/main). All these models should be put in folder "annotator/ckpts".
+
+To install:
+
+ conda env create -f environment.yaml
+ conda activate control-v11
+
+Note that if you use 8GB GPU, you need to set "save_memory = True" in "config.py".
+
+## ControlNet 1.1 Depth
+
+Control Stable Diffusion with Depth Maps.
+
+Model file: control_v11f1p_sd15_depth.pth
+
+Config file: control_v11f1p_sd15_depth.yaml
+
+Training data: Midas depth (resolution 256/384/512) + Leres Depth (resolution 256/384/512) + Zoe Depth (resolution 256/384/512). Multiple depth map generator at multiple resolution as data augmentation.
+
+Acceptable Preprocessors: Depth_Midas, Depth_Leres, Depth_Zoe. This model is highly robust and can work on real depth map from rendering engines.
+
+ python gradio_depth.py
+
+Non-cherry-picked batch test with random seed 12345 ("a handsome man"):
+
+
+
+**Update**
+
+2023/04/14: 72 hours ago we uploaded a wrong model "control_v11p_sd15_depth" by mistake. That model is an intermediate checkpoint during the training. That model is not converged and may cause distortion in results. We uploaded the correct depth model as "control_v11f1p_sd15_depth". The "f1" means bug fix 1. The incorrect model is removed. Sorry for the inconvenience.
+
+**Improvements in Depth 1.1:**
+
+1. The training dataset of previous cnet 1.0 has several problems including (1) a small group of greyscale human images are duplicated thousands of times (!!), causing the previous model somewhat likely to generate grayscale human images; (2) some images has low quality, very blurry, or significant JPEG artifacts; (3) a small group of images has wrong paired prompts caused by a mistake in our data processing scripts. The new model fixed all problems of the training dataset and should be more reasonable in many cases.
+2. The new depth model is a relatively unbiased model. It is not trained with some specific type of depth by some specific depth estimation method. It is not over-fitted to one preprocessor. This means this model will work better with different depth estimation, different preprocessor resolutions, or even with real depth created by 3D engines.
+3. Some reasonable data augmentations are applied to training, like random left-right flipping.
+4. The model is resumed from depth 1.0, and it should work well in all cases where depth 1.0 works well. If not, please open an issue with image, and we will take a look at your case. Depth 1.1 works well in many failure cases of depth 1.0.
+5. If you use Midas depth (the "depth" in webui plugin) with 384 preprocessor resolution, the difference between depth 1.0 and 1.1 should be minimal. However, if you try other preprocessor resolutions or other preprocessors (like leres and zoe), the depth 1.1 is expected to be a bit better than 1.0.
+
+## ControlNet 1.1 Normal
+
+Control Stable Diffusion with Normal Maps.
+
+Model file: control_v11p_sd15_normalbae.pth
+
+Config file: control_v11p_sd15_normalbae.yaml
+
+Training data: [Bae's](https://github.com/baegwangbin/surface_normal_uncertainty) normalmap estimation method.
+
+Acceptable Preprocessors: Normal BAE. This model can accept normal maps from rendering engines as long as the normal map follows [ScanNet's](http://www.scan-net.org/) protocol. That is to say, the color of your normal map should look like [the second column of this image](https://raw.githubusercontent.com/baegwangbin/surface_normal_uncertainty/main/figs/readme_scannet.png).
+
+Note that this method is much more reasonable than the normal-from-midas method in ControlNet 1.0. The previous method will be abandoned.
+
+ python gradio_normalbae.py
+
+Non-cherry-picked batch test with random seed 12345 ("a man made of flowers"):
+
+
+
+Non-cherry-picked batch test with random seed 12345 ("room"):
+
+
+
+**Improvements in Normal 1.1:**
+
+1. The normal-from-midas method in Normal 1.0 is neither reasonable nor physically correct. That method does not work very well in many images. The normal 1.0 model cannot interpret real normal maps created by rendering engines.
+2. This Normal 1.1 is much more reasonable because the preprocessor is trained to estimate normal maps with a relatively correct protocol (NYU-V2's visualization method). This means the Normal 1.1 can interpret real normal maps from rendering engines as long as the colors are correct (blue is front, red is left, green is top).
+3. In our test, this model is robust and can achieve similar performance to the depth model. In previous CNET 1.0, the Normal 1.0 is not very frequently used. But this Normal 2.0 is much improved and has potential to be used much more frequently.
+
+## ControlNet 1.1 Canny
+
+Control Stable Diffusion with Canny Maps.
+
+Model file: control_v11p_sd15_canny.pth
+
+Config file: control_v11p_sd15_canny.yaml
+
+Training data: Canny with random thresholds.
+
+Acceptable Preprocessors: Canny.
+
+We fixed several problems in previous training datasets.
+
+ python gradio_canny.py
+
+Non-cherry-picked batch test with random seed 12345 ("dog in a room"):
+
+
+
+**Improvements in Canny 1.1:**
+
+1. The training dataset of previous cnet 1.0 has several problems including (1) a small group of greyscale human images are duplicated thousands of times (!!), causing the previous model somewhat likely to generate grayscale human images; (2) some images has low quality, very blurry, or significant JPEG artifacts; (3) a small group of images has wrong paired prompts caused by a mistake in our data processing scripts. The new model fixed all problems of the training dataset and should be more reasonable in many cases.
+2. Because the Canny model is one of the most important (perhaps the most frequently used) ControlNet, we used a fund to train it on a machine with 8 Nvidia A100 80G with batchsize 8×32=256 for 3 days, spending 72×30=2160 USD (8 A100 80G with 30 USD/hour). The model is resumed from Canny 1.0.
+3. Some reasonable data augmentations are applied to training, like random left-right flipping.
+4. Although it is difficult to evaluate a ControlNet, we find Canny 1.1 is a bit more robust and a bit higher visual quality than Canny 1.0.
+
+## ControlNet 1.1 MLSD
+
+Control Stable Diffusion with M-LSD straight lines.
+
+Model file: control_v11p_sd15_mlsd.pth
+
+Config file: control_v11p_sd15_mlsd.yaml
+
+Training data: M-LSD Lines.
+
+Acceptable Preprocessors: MLSD.
+
+We fixed several problems in previous training datasets. The model is resumed from ControlNet 1.0 and trained with 200 GPU hours of A100 80G.
+
+ python gradio_mlsd.py
+
+Non-cherry-picked batch test with random seed 12345 ("room"):
+
+
+
+**Improvements in MLSD 1.1:**
+
+1. The training dataset of previous cnet 1.0 has several problems including (1) a small group of greyscale human images are duplicated thousands of times (!!), causing the previous model somewhat likely to generate grayscale human images; (2) some images has low quality, very blurry, or significant JPEG artifacts; (3) a small group of images has wrong paired prompts caused by a mistake in our data processing scripts. The new model fixed all problems of the training dataset and should be more reasonable in many cases.
+2. We enlarged the training dataset by adding 300K more images by using MLSD to find images with more than 16 straight lines in it.
+3. Some reasonable data augmentations are applied to training, like random left-right flipping.
+4. Resumed from MLSD 1.0 with continued training with 200 GPU hours of A100 80G.
+
+## ControlNet 1.1 Scribble
+
+Control Stable Diffusion with Scribbles.
+
+Model file: control_v11p_sd15_scribble.pth
+
+Config file: control_v11p_sd15_scribble.yaml
+
+Training data: Synthesized scribbles.
+
+Acceptable Preprocessors: Synthesized scribbles (Scribble_HED, Scribble_PIDI, etc.) or hand-drawn scribbles.
+
+We fixed several problems in previous training datasets. The model is resumed from ControlNet 1.0 and trained with 200 GPU hours of A100 80G.
+
+ # To test synthesized scribbles
+ python gradio_scribble.py
+ # To test hand-drawn scribbles in an interactive demo
+ python gradio_interactive.py
+
+Non-cherry-picked batch test with random seed 12345 ("man in library"):
+
+
+
+Non-cherry-picked batch test with random seed 12345 (interactive, "the beautiful landscape"):
+
+
+
+**Improvements in Scribble 1.1:**
+
+1. The training dataset of previous cnet 1.0 has several problems including (1) a small group of greyscale human images are duplicated thousands of times (!!), causing the previous model somewhat likely to generate grayscale human images; (2) some images has low quality, very blurry, or significant JPEG artifacts; (3) a small group of images has wrong paired prompts caused by a mistake in our data processing scripts. The new model fixed all problems of the training dataset and should be more reasonable in many cases.
+2. We find out that users sometimes like to draw very thick scribbles. Because of that, we used more aggressive random morphological transforms to synthesize scribbles. This model should work well even when the scribbles are relatively thick (the maximum width of training data is 24-pixel-width scribble in a 512 canvas, but it seems to work well even for a bit wider scribbles; the minimum width is 1 pixel).
+3. Resumed from Scribble 1.0, continued with 200 GPU hours of A100 80G.
+
+## ControlNet 1.1 Soft Edge
+
+Control Stable Diffusion with Soft Edges.
+
+Model file: control_v11p_sd15_softedge.pth
+
+Config file: control_v11p_sd15_softedge.yaml
+
+Training data: SoftEdge_PIDI, SoftEdge_PIDI_safe, SoftEdge_HED, SoftEdge_HED_safe.
+
+Acceptable Preprocessors: SoftEdge_PIDI, SoftEdge_PIDI_safe, SoftEdge_HED, SoftEdge_HED_safe.
+
+This model is significantly improved compared to previous model. All users should update as soon as possible.
+
+New in ControlNet 1.1: now we added a new type of soft edge called "SoftEdge_safe". This is motivated by the fact that HED or PIDI tends to hide a corrupted greyscale version of the original image inside the soft estimation, and such hidden patterns can distract ControlNet, leading to bad results. The solution is to use a pre-processing to quantize the edge maps into several levels so that the hidden patterns can be completely removed. The implementation is [in the 78-th line of annotator/util.py](https://github.com/lllyasviel/ControlNet-v1-1-nightly/blob/4c9560ebe7679daac53a0599a11b9b7cd984ac55/annotator/util.py#L78).
+
+The perforamce can be roughly noted as:
+
+Robustness: SoftEdge_PIDI_safe > SoftEdge_HED_safe >> SoftEdge_PIDI > SoftEdge_HED
+
+Maximum result quality: SoftEdge_HED > SoftEdge_PIDI > SoftEdge_HED_safe > SoftEdge_PIDI_safe
+
+Considering the trade-off, we recommend to use SoftEdge_PIDI by default. In most cases it works very well.
+
+ python gradio_softedge.py
+
+Non-cherry-picked batch test with random seed 12345 ("a handsome man"):
+
+
+
+**Improvements in Soft Edge 1.1:**
+
+1. Soft Edge 1.1 was called HED 1.0 in previous ControlNet.
+2. The training dataset of previous cnet 1.0 has several problems including (1) a small group of greyscale human images are duplicated thousands of times (!!), causing the previous model somewhat likely to generate grayscale human images; (2) some images has low quality, very blurry, or significant JPEG artifacts; (3) a small group of images has wrong paired prompts caused by a mistake in our data processing scripts. The new model fixed all problems of the training dataset and should be more reasonable in many cases.
+3. The Soft Edge 1.1 is significantly (in nealy 100\% cases) better than HED 1.0. This is mainly because HED or PIDI estimator tend to hide a corrupted greyscale version of original image inside the soft edge map and the previous model HED 1.0 is over-fitted to restore that hidden corrupted image rather than perform boundary-aware diffusion. The training of Soft Edge 1.1 used 75\% "safe" filtering to remove such hidden corrupted greyscale images insider control maps. This makes the Soft Edge 1.1 very robust. In out test, Soft Edge 1.1 is as usable as the depth model and has potential to be more frequently used.
+
+## ControlNet 1.1 Segmentation
+
+Control Stable Diffusion with Semantic Segmentation.
+
+Model file: control_v11p_sd15_seg.pth
+
+Config file: control_v11p_sd15_seg.yaml
+
+Training data: COCO + ADE20K.
+
+Acceptable Preprocessors: Seg_OFADE20K (Oneformer ADE20K), Seg_OFCOCO (Oneformer COCO), Seg_UFADE20K (Uniformer ADE20K), or manually created masks.
+
+Now the model can receive both type of ADE20K or COCO annotations. We find that recognizing the segmentation protocol is trivial for the ControlNet encoder and training the model of multiple segmentation protocols lead to better performance.
+
+ python gradio_seg.py
+
+Non-cherry-picked batch test with random seed 12345 (ADE20k protocol, "house"):
+
+
+
+Non-cherry-picked batch test with random seed 12345 (COCO protocol, "house"):
+
+
+
+**Improvements in Segmentation 1.1:**
+
+1. COCO protocol is supported. The previous Segmentation 1.0 supports about 150 colors, but Segmentation 1.1 supports another 182 colors from coco.
+2. Resumed from Segmentation 1.0. All previous inputs should still work.
+
+## ControlNet 1.1 Openpose
+
+Control Stable Diffusion with Openpose.
+
+Model file: control_v11p_sd15_openpose.pth
+
+Config file: control_v11p_sd15_openpose.yaml
+
+The model is trained and can accept the following combinations:
+
+* Openpose body
+* Openpose hand
+* Openpose face
+* Openpose body + Openpose hand
+* Openpose body + Openpose face
+* Openpose hand + Openpose face
+* Openpose body + Openpose hand + Openpose face
+
+However, providing all those combinations is too complicated. We recommend to provide the users with only two choices:
+
+* "Openpose" = Openpose body
+* "Openpose Full" = Openpose body + Openpose hand + Openpose face
+
+You can try with the demo:
+
+ python gradio_openpose.py
+
+Non-cherry-picked batch test with random seed 12345 ("man in suit"):
+
+
+
+Non-cherry-picked batch test with random seed 12345 (multiple people in the wild, "handsome boys in the party"):
+
+
+
+**Improvements in Openpose 1.1:**
+
+1. The improvement of this model is mainly based on our improved implementation of OpenPose. We carefully reviewed the difference between the pytorch OpenPose and CMU's c++ openpose. Now the processor should be more accurate, especially for hands. The improvement of processor leads to the improvement of Openpose 1.1.
+2. More inputs are supported (hand and face).
+3. The training dataset of previous cnet 1.0 has several problems including (1) a small group of greyscale human images are duplicated thousands of times (!!), causing the previous model somewhat likely to generate grayscale human images; (2) some images has low quality, very blurry, or significant JPEG artifacts; (3) a small group of images has wrong paired prompts caused by a mistake in our data processing scripts. The new model fixed all problems of the training dataset and should be more reasonable in many cases.
+
+## ControlNet 1.1 Lineart
+
+Control Stable Diffusion with Linearts.
+
+Model file: control_v11p_sd15_lineart.pth
+
+Config file: control_v11p_sd15_lineart.yaml
+
+This model is trained on awacke1/Image-to-Line-Drawings. The preprocessor can generate detailed or coarse linearts from images (Lineart and Lineart_Coarse). The model is trained with sufficient data augmentation and can receive manually drawn linearts.
+
+ python gradio_lineart.py
+
+Non-cherry-picked batch test with random seed 12345 (detailed lineart extractor, "bag"):
+
+
+
+Non-cherry-picked batch test with random seed 12345 (coarse lineart extractor, "Michael Jackson's concert"):
+
+
+
+Non-cherry-picked batch test with random seed 12345 (use manually drawn linearts, "wolf"):
+
+
+
+
+## ControlNet 1.1 Anime Lineart
+
+Control Stable Diffusion with Anime Linearts.
+
+Model file: control_v11p_sd15s2_lineart_anime.pth
+
+Config file: control_v11p_sd15s2_lineart_anime.yaml
+
+Training data and implementation details: (description removed).
+
+This model can take real anime line drawings or extracted line drawings as inputs.
+
+Some important notice:
+
+1. You need a file "anything-v3-full.safetensors" to run the demo. We will not provide the file. Please find that file on the Internet on your own.
+2. This model is trained with 3x token length and clip skip 2.
+3. This is a long prompt model. Unless you use LoRAs, results are better with long prompts.
+4. This model does not support Guess Mode.
+
+Demo:
+
+ python gradio_lineart_anime.py
+
+
+Non-cherry-picked batch test with random seed 12345 ("1girl, in classroom, skirt, uniform, red hair, bag, green eyes"):
+
+
+
+Non-cherry-picked batch test with random seed 12345 ("1girl, saber, at night, sword, green eyes, golden hair, stocking"):
+
+
+
+Non-cherry-picked batch test with random seed 12345 (extracted line drawing, "1girl, Castle, silver hair, dress, Gemstone, cinematic lighting, mechanical hand, 4k, 8k, extremely detailed, Gothic, green eye"):
+
+
+
+## ControlNet 1.1 Shuffle
+
+Control Stable Diffusion with Content Shuffle.
+
+Model file: control_v11e_sd15_shuffle.pth
+
+Config file: control_v11e_sd15_shuffle.yaml
+
+Demo:
+
+ python gradio_shuffle.py
+
+The model is trained to reorganize images. [We use a random flow to shuffle the image and control Stable Diffusion to recompose the image.](github_docs/annotator.md#content-reshuffle)
+
+Non-cherry-picked batch test with random seed 12345 ("hong kong"):
+
+
+
+In the 6 images on the right, the left-top one is the "shuffled" image. All others are outputs.
+
+In fact, since the ControlNet is trained to recompose images, we do not even need to shuffle the input - sometimes we can just use the original image as input.
+
+In this way, this ControlNet can be guided by prompts or other ControlNets to change the image style.
+
+Note that this method has nothing to do with CLIP vision or some other models.
+
+This is a pure ControlNet.
+
+Non-cherry-picked batch test with random seed 12345 ("iron man"):
+
+
+
+Non-cherry-picked batch test with random seed 12345 ("spider man"):
+
+
+
+**Multi-ControlNets** (A1111-only)
+
+Source Image (not used):
+
+
+
+Canny Image (Input):
+
+
+
+Shuffle Image (Input):
+
+
+
+Outputs:
+
+
+
+(From: https://github.com/Mikubill/sd-webui-controlnet/issues/736#issuecomment-1509986321)
+
+**Important If You Implement Your Own Inference:**
+
+Note that this ControlNet requires to add a global average pooling " x = torch.mean(x, dim=(2, 3), keepdim=True) " between the ControlNet Encoder outputs and SD Unet layers. And the ControlNet must be put only on the conditional side of cfg scale. We recommend to use the "global_average_pooling" item in the yaml file to control such behaviors.
+
+~Note that this ControlNet Shuffle will be the one and only one image stylization method that we will maintain for the robustness in a long term support. We have tested other CLIP image encoder, Unclip, image tokenization, and image-based prompts but it seems that those methods do not work very well with user prompts or additional/multiple U-Net injections. See also the evidence [here](https://github.com/lllyasviel/ControlNet/issues/255), [here](https://github.com/Mikubill/sd-webui-controlnet/issues/547), and some other related issues.~ After some more recent researches/experiments, we plan to support more types of stylization methods in the future.
+
+## ControlNet 1.1 Instruct Pix2Pix
+
+Control Stable Diffusion with Instruct Pix2Pix.
+
+Model file: control_v11e_sd15_ip2p.pth
+
+Config file: control_v11e_sd15_ip2p.yaml
+
+Demo:
+
+ python gradio_ip2p.py
+
+This is a controlnet trained on the [Instruct Pix2Pix dataset](https://github.com/timothybrooks/instruct-pix2pix).
+
+Different from official Instruct Pix2Pix, this model is trained with 50\% instruction prompts and 50\% description prompts. For example, "a cute boy" is a description prompt, while "make the boy cute" is a instruction prompt.
+
+Because this is a ControlNet, you do not need to trouble with original IP2P's double cfg tuning. And, this model can be applied to any base model.
+
+Also, it seems that instructions like "make it into X" works better than "make Y into X".
+
+Non-cherry-picked batch test with random seed 12345 ("make it on fire"):
+
+
+
+Non-cherry-picked batch test with random seed 12345 ("make it winter"):
+
+
+
+We mark this model as "experimental" because it sometimes needs cherry-picking. For example, here is non-cherry-picked batch test with random seed 12345 ("make he iron man"):
+
+
+
+
+## ControlNet 1.1 Inpaint
+
+Control Stable Diffusion with Inpaint.
+
+Model file: control_v11p_sd15_inpaint.pth
+
+Config file: control_v11p_sd15_inpaint.yaml
+
+Demo:
+
+ python gradio_inpaint.py
+
+Some notices:
+
+1. This inpainting ControlNet is trained with 50\% random masks and 50\% random optical flow occlusion masks. This means the model can not only support the inpainting application but also work on video optical flow warping. Perhaps we will provide some example in the future (depending on our workloads).
+2. We updated the gradio (2023/5/11) so that the standalone gradio codes in main ControlNet repo also do not change unmasked areas. Automatic 1111 users are not influenced.
+
+Non-cherry-picked batch test with random seed 12345 ("a handsome man"):
+
+
+
+See also the Guidelines for [Using ControlNet Inpaint in Automatic 1111](https://github.com/Mikubill/sd-webui-controlnet/discussions/1143).
+
+## ControlNet 1.1 Tile
+
+Update 2023 April 25: The previously unfinished tile model is finished now. The new name is "control_v11f1e_sd15_tile". The "f1e" means 1st bug fix ("f1"), experimental ("e"). The previous "control_v11u_sd15_tile" is removed. Please update if your model name is "v11u".
+
+Control Stable Diffusion with Tiles.
+
+Model file: control_v11f1e_sd15_tile.pth
+
+Config file: control_v11f1e_sd15_tile.yaml
+
+Demo:
+
+ python gradio_tile.py
+
+The model can be used in many ways. Overall, the model has two behaviors:
+
+* Ignore the details in an image and generate new details.
+* Ignore global prompts if local tile semantics and prompts mismatch, and guide diffusion with local context.
+
+Because the model can generate new details and ignore existing image details, we can use this model to remove bad details and add refined details. For example, remove blurring caused by image resizing.
+
+Below is an example of 8x super resolution. This is a 64x64 dog image.
+
+
+
+Non-cherry-picked batch test with random seed 12345 ("dog on grassland"):
+
+
+
+Note that this model is not a super resolution model. It ignores the details in an image and generate new details. This means you can use it to fix bad details in an image.
+
+For example, below is a dog image corrupted by Real-ESRGAN. This is a typical example that sometimes super resolution methds fail to upscale images when source context is too small.
+
+
+
+Non-cherry-picked batch test with random seed 12345 ("dog on grassland"):
+
+
+
+If your image already have good details, you can still use this model to replace image details. Note that Stable Diffusion's I2I can achieve similar effects but this model make it much easier for you to maintain the overall structure and only change details even with denoising strength 1.0 .
+
+Non-cherry-picked batch test with random seed 12345 ("Silver Armor"):
+
+
+
+More and more people begin to think about different methods to diffuse at tiles so that images can be very big (at 4k or 8k).
+
+The problem is that, in Stable Diffusion, your prompts will always influent each tile.
+
+For example, if your prompts are "a beautiful girl" and you split an image into 4×4=16 blocks and do diffusion in each block, then you are will get 16 "beautiful girls" rather than "a beautiful girl". This is a well-known problem.
+
+Right now people's solution is to use some meaningless prompts like "clear, clear, super clear" to diffuse blocks. But you can expect that the results will be bad if the denonising strength is high. And because the prompts are bad, the contents are pretty random.
+
+ControlNet Tile can solve this problem. For a given tile, it recognizes what is inside the tile and increase the influence of that recognized semantics, and it also decreases the influence of global prompts if contents do not match.
+
+Non-cherry-picked batch test with random seed 12345 ("a handsome man"):
+
+
+
+You can see that the prompt is "a handsome man" but the model does not paint "a handsome man" on that tree leaves. Instead, it recognizes the tree leaves paint accordingly.
+
+In this way, ControlNet is able to change the behavior of any Stable Diffusion model to perform diffusion in tiles.
+
+**Gallery of ControlNet Tile**
+
+*Note:* Our official support for tiled image upscaling is A1111-only. The gradio example in this repo does not include tiled upscaling scripts. Please use the A1111 extension to perform tiled upscaling (with other tiling scripts like Ultimate SD Upscale or Tiled Diffusion/VAE).
+
+From https://github.com/Mikubill/sd-webui-controlnet/discussions/1142#discussioncomment-5788601
+
+(Output, **Click image to see full resolution**)
+
+
+
+(Zooming-in of outputs)
+
+
+
+
+
+
+
+From https://github.com/Mikubill/sd-webui-controlnet/discussions/1142#discussioncomment-5788617
+
+(Input)
+
+
+
+(Output, **Click image to see full resolution**)
+
+
+From: https://github.com/lllyasviel/ControlNet-v1-1-nightly/issues/50#issuecomment-1541914890
+
+(Input)
+
+
+
+(Output, **Click image to see full resolution**, note that this example is extremely challenging)
+
+
+
+From https://github.com/Mikubill/sd-webui-controlnet/discussions/1142#discussioncomment-5796326:
+
+(before)
+
+
+
+(after, **Click image to see full resolution**)
+
+
+**Comparison to Midjourney V5/V5.1 coming soon.**
+
+# Annotate Your Own Data
+
+We provide simple python scripts to process images.
+
+[See a gradio example here](github_docs/annotator.md).
diff --git a/ControlNet-v1-1-nightly-main/annotator/canny/__init__.py b/ControlNet-v1-1-nightly-main/annotator/canny/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..cb0da951dc838ec9dec2131007e036113281800b
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/canny/__init__.py
@@ -0,0 +1,6 @@
+import cv2
+
+
+class CannyDetector:
+ def __call__(self, img, low_threshold, high_threshold):
+ return cv2.Canny(img, low_threshold, high_threshold)
diff --git a/ControlNet-v1-1-nightly-main/annotator/ckpts/ckpts.txt b/ControlNet-v1-1-nightly-main/annotator/ckpts/ckpts.txt
new file mode 100644
index 0000000000000000000000000000000000000000..1978551fb2a9226814eaf58459f414fcfac4e69b
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/ckpts/ckpts.txt
@@ -0,0 +1 @@
+Weights here.
\ No newline at end of file
diff --git a/ControlNet-v1-1-nightly-main/annotator/hed/__init__.py b/ControlNet-v1-1-nightly-main/annotator/hed/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..003c66768666296ef59bcbd144dc132a2b362dbe
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/hed/__init__.py
@@ -0,0 +1,80 @@
+# This is an improved version and model of HED edge detection with Apache License, Version 2.0.
+# Please use this implementation in your products
+# This implementation may produce slightly different results from Saining Xie's official implementations,
+# but it generates smoother edges and is more suitable for ControlNet as well as other image-to-image translations.
+# Different from official models and other implementations, this is an RGB-input model (rather than BGR)
+# and in this way it works better for gradio's RGB protocol
+
+import os
+import cv2
+import torch
+import numpy as np
+
+from einops import rearrange
+from annotator.util import annotator_ckpts_path, safe_step
+
+
+class DoubleConvBlock(torch.nn.Module):
+ def __init__(self, input_channel, output_channel, layer_number):
+ super().__init__()
+ self.convs = torch.nn.Sequential()
+ self.convs.append(torch.nn.Conv2d(in_channels=input_channel, out_channels=output_channel, kernel_size=(3, 3), stride=(1, 1), padding=1))
+ for i in range(1, layer_number):
+ self.convs.append(torch.nn.Conv2d(in_channels=output_channel, out_channels=output_channel, kernel_size=(3, 3), stride=(1, 1), padding=1))
+ self.projection = torch.nn.Conv2d(in_channels=output_channel, out_channels=1, kernel_size=(1, 1), stride=(1, 1), padding=0)
+
+ def __call__(self, x, down_sampling=False):
+ h = x
+ if down_sampling:
+ h = torch.nn.functional.max_pool2d(h, kernel_size=(2, 2), stride=(2, 2))
+ for conv in self.convs:
+ h = conv(h)
+ h = torch.nn.functional.relu(h)
+ return h, self.projection(h)
+
+
+class ControlNetHED_Apache2(torch.nn.Module):
+ def __init__(self):
+ super().__init__()
+ self.norm = torch.nn.Parameter(torch.zeros(size=(1, 3, 1, 1)))
+ self.block1 = DoubleConvBlock(input_channel=3, output_channel=64, layer_number=2)
+ self.block2 = DoubleConvBlock(input_channel=64, output_channel=128, layer_number=2)
+ self.block3 = DoubleConvBlock(input_channel=128, output_channel=256, layer_number=3)
+ self.block4 = DoubleConvBlock(input_channel=256, output_channel=512, layer_number=3)
+ self.block5 = DoubleConvBlock(input_channel=512, output_channel=512, layer_number=3)
+
+ def __call__(self, x):
+ h = x - self.norm
+ h, projection1 = self.block1(h)
+ h, projection2 = self.block2(h, down_sampling=True)
+ h, projection3 = self.block3(h, down_sampling=True)
+ h, projection4 = self.block4(h, down_sampling=True)
+ h, projection5 = self.block5(h, down_sampling=True)
+ return projection1, projection2, projection3, projection4, projection5
+
+
+class HEDdetector:
+ def __init__(self):
+ remote_model_path = "https://huggingface.co/lllyasviel/Annotators/resolve/main/ControlNetHED.pth"
+ modelpath = os.path.join(annotator_ckpts_path, "ControlNetHED.pth")
+ if not os.path.exists(modelpath):
+ from basicsr.utils.download_util import load_file_from_url
+ load_file_from_url(remote_model_path, model_dir=annotator_ckpts_path)
+ self.netNetwork = ControlNetHED_Apache2().float().cuda().eval()
+ self.netNetwork.load_state_dict(torch.load(modelpath))
+
+ def __call__(self, input_image, safe=False):
+ assert input_image.ndim == 3
+ H, W, C = input_image.shape
+ with torch.no_grad():
+ image_hed = torch.from_numpy(input_image.copy()).float().cuda()
+ image_hed = rearrange(image_hed, 'h w c -> 1 c h w')
+ edges = self.netNetwork(image_hed)
+ edges = [e.detach().cpu().numpy().astype(np.float32)[0, 0] for e in edges]
+ edges = [cv2.resize(e, (W, H), interpolation=cv2.INTER_LINEAR) for e in edges]
+ edges = np.stack(edges, axis=2)
+ edge = 1 / (1 + np.exp(-np.mean(edges, axis=2).astype(np.float64)))
+ if safe:
+ edge = safe_step(edge)
+ edge = (edge * 255.0).clip(0, 255).astype(np.uint8)
+ return edge
diff --git a/ControlNet-v1-1-nightly-main/annotator/lineart/LICENSE b/ControlNet-v1-1-nightly-main/annotator/lineart/LICENSE
new file mode 100644
index 0000000000000000000000000000000000000000..16a9d56a3d4c15e4f34ac5426459c58487b01520
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/lineart/LICENSE
@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2022 Caroline Chan
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
\ No newline at end of file
diff --git a/ControlNet-v1-1-nightly-main/annotator/lineart/__init__.py b/ControlNet-v1-1-nightly-main/annotator/lineart/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..fd92f9981158a61827109c9ca4009c2a441ad2b2
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/lineart/__init__.py
@@ -0,0 +1,124 @@
+# From https://github.com/carolineec/informative-drawings
+# MIT License
+
+import os
+import cv2
+import torch
+import numpy as np
+
+import torch.nn as nn
+from einops import rearrange
+from annotator.util import annotator_ckpts_path
+
+
+norm_layer = nn.InstanceNorm2d
+
+
+class ResidualBlock(nn.Module):
+ def __init__(self, in_features):
+ super(ResidualBlock, self).__init__()
+
+ conv_block = [ nn.ReflectionPad2d(1),
+ nn.Conv2d(in_features, in_features, 3),
+ norm_layer(in_features),
+ nn.ReLU(inplace=True),
+ nn.ReflectionPad2d(1),
+ nn.Conv2d(in_features, in_features, 3),
+ norm_layer(in_features)
+ ]
+
+ self.conv_block = nn.Sequential(*conv_block)
+
+ def forward(self, x):
+ return x + self.conv_block(x)
+
+
+class Generator(nn.Module):
+ def __init__(self, input_nc, output_nc, n_residual_blocks=9, sigmoid=True):
+ super(Generator, self).__init__()
+
+ # Initial convolution block
+ model0 = [ nn.ReflectionPad2d(3),
+ nn.Conv2d(input_nc, 64, 7),
+ norm_layer(64),
+ nn.ReLU(inplace=True) ]
+ self.model0 = nn.Sequential(*model0)
+
+ # Downsampling
+ model1 = []
+ in_features = 64
+ out_features = in_features*2
+ for _ in range(2):
+ model1 += [ nn.Conv2d(in_features, out_features, 3, stride=2, padding=1),
+ norm_layer(out_features),
+ nn.ReLU(inplace=True) ]
+ in_features = out_features
+ out_features = in_features*2
+ self.model1 = nn.Sequential(*model1)
+
+ model2 = []
+ # Residual blocks
+ for _ in range(n_residual_blocks):
+ model2 += [ResidualBlock(in_features)]
+ self.model2 = nn.Sequential(*model2)
+
+ # Upsampling
+ model3 = []
+ out_features = in_features//2
+ for _ in range(2):
+ model3 += [ nn.ConvTranspose2d(in_features, out_features, 3, stride=2, padding=1, output_padding=1),
+ norm_layer(out_features),
+ nn.ReLU(inplace=True) ]
+ in_features = out_features
+ out_features = in_features//2
+ self.model3 = nn.Sequential(*model3)
+
+ # Output layer
+ model4 = [ nn.ReflectionPad2d(3),
+ nn.Conv2d(64, output_nc, 7)]
+ if sigmoid:
+ model4 += [nn.Sigmoid()]
+
+ self.model4 = nn.Sequential(*model4)
+
+ def forward(self, x, cond=None):
+ out = self.model0(x)
+ out = self.model1(out)
+ out = self.model2(out)
+ out = self.model3(out)
+ out = self.model4(out)
+
+ return out
+
+
+class LineartDetector:
+ def __init__(self):
+ self.model = self.load_model('sk_model.pth')
+ self.model_coarse = self.load_model('sk_model2.pth')
+
+ def load_model(self, name):
+ remote_model_path = "https://huggingface.co/lllyasviel/Annotators/resolve/main/" + name
+ modelpath = os.path.join(annotator_ckpts_path, name)
+ if not os.path.exists(modelpath):
+ from basicsr.utils.download_util import load_file_from_url
+ load_file_from_url(remote_model_path, model_dir=annotator_ckpts_path)
+ model = Generator(3, 1, 3)
+ model.load_state_dict(torch.load(modelpath, map_location=torch.device('cpu')))
+ model.eval()
+ model = model.cuda()
+ return model
+
+ def __call__(self, input_image, coarse):
+ model = self.model_coarse if coarse else self.model
+ assert input_image.ndim == 3
+ image = input_image
+ with torch.no_grad():
+ image = torch.from_numpy(image).float().cuda()
+ image = image / 255.0
+ image = rearrange(image, 'h w c -> 1 c h w')
+ line = model(image)[0][0]
+
+ line = line.cpu().numpy()
+ line = (line * 255.0).clip(0, 255).astype(np.uint8)
+
+ return line
diff --git a/ControlNet-v1-1-nightly-main/annotator/lineart_anime/LICENSE b/ControlNet-v1-1-nightly-main/annotator/lineart_anime/LICENSE
new file mode 100644
index 0000000000000000000000000000000000000000..16a9d56a3d4c15e4f34ac5426459c58487b01520
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/lineart_anime/LICENSE
@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2022 Caroline Chan
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
\ No newline at end of file
diff --git a/ControlNet-v1-1-nightly-main/annotator/lineart_anime/__init__.py b/ControlNet-v1-1-nightly-main/annotator/lineart_anime/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..acdc35b8c813857b08b3e7bab9a1b6ef530efa63
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/lineart_anime/__init__.py
@@ -0,0 +1,150 @@
+# Anime2sketch
+# https://github.com/Mukosame/Anime2Sketch
+
+import numpy as np
+import torch
+import torch.nn as nn
+import functools
+
+import os
+import cv2
+from einops import rearrange
+from annotator.util import annotator_ckpts_path
+
+
+class UnetGenerator(nn.Module):
+ """Create a Unet-based generator"""
+
+ def __init__(self, input_nc, output_nc, num_downs, ngf=64, norm_layer=nn.BatchNorm2d, use_dropout=False):
+ """Construct a Unet generator
+ Parameters:
+ input_nc (int) -- the number of channels in input images
+ output_nc (int) -- the number of channels in output images
+ num_downs (int) -- the number of downsamplings in UNet. For example, # if |num_downs| == 7,
+ image of size 128x128 will become of size 1x1 # at the bottleneck
+ ngf (int) -- the number of filters in the last conv layer
+ norm_layer -- normalization layer
+ We construct the U-Net from the innermost layer to the outermost layer.
+ It is a recursive process.
+ """
+ super(UnetGenerator, self).__init__()
+ # construct unet structure
+ unet_block = UnetSkipConnectionBlock(ngf * 8, ngf * 8, input_nc=None, submodule=None, norm_layer=norm_layer, innermost=True) # add the innermost layer
+ for _ in range(num_downs - 5): # add intermediate layers with ngf * 8 filters
+ unet_block = UnetSkipConnectionBlock(ngf * 8, ngf * 8, input_nc=None, submodule=unet_block, norm_layer=norm_layer, use_dropout=use_dropout)
+ # gradually reduce the number of filters from ngf * 8 to ngf
+ unet_block = UnetSkipConnectionBlock(ngf * 4, ngf * 8, input_nc=None, submodule=unet_block, norm_layer=norm_layer)
+ unet_block = UnetSkipConnectionBlock(ngf * 2, ngf * 4, input_nc=None, submodule=unet_block, norm_layer=norm_layer)
+ unet_block = UnetSkipConnectionBlock(ngf, ngf * 2, input_nc=None, submodule=unet_block, norm_layer=norm_layer)
+ self.model = UnetSkipConnectionBlock(output_nc, ngf, input_nc=input_nc, submodule=unet_block, outermost=True, norm_layer=norm_layer) # add the outermost layer
+
+ def forward(self, input):
+ """Standard forward"""
+ return self.model(input)
+
+
+class UnetSkipConnectionBlock(nn.Module):
+ """Defines the Unet submodule with skip connection.
+ X -------------------identity----------------------
+ |-- downsampling -- |submodule| -- upsampling --|
+ """
+
+ def __init__(self, outer_nc, inner_nc, input_nc=None,
+ submodule=None, outermost=False, innermost=False, norm_layer=nn.BatchNorm2d, use_dropout=False):
+ """Construct a Unet submodule with skip connections.
+ Parameters:
+ outer_nc (int) -- the number of filters in the outer conv layer
+ inner_nc (int) -- the number of filters in the inner conv layer
+ input_nc (int) -- the number of channels in input images/features
+ submodule (UnetSkipConnectionBlock) -- previously defined submodules
+ outermost (bool) -- if this module is the outermost module
+ innermost (bool) -- if this module is the innermost module
+ norm_layer -- normalization layer
+ use_dropout (bool) -- if use dropout layers.
+ """
+ super(UnetSkipConnectionBlock, self).__init__()
+ self.outermost = outermost
+ if type(norm_layer) == functools.partial:
+ use_bias = norm_layer.func == nn.InstanceNorm2d
+ else:
+ use_bias = norm_layer == nn.InstanceNorm2d
+ if input_nc is None:
+ input_nc = outer_nc
+ downconv = nn.Conv2d(input_nc, inner_nc, kernel_size=4,
+ stride=2, padding=1, bias=use_bias)
+ downrelu = nn.LeakyReLU(0.2, True)
+ downnorm = norm_layer(inner_nc)
+ uprelu = nn.ReLU(True)
+ upnorm = norm_layer(outer_nc)
+
+ if outermost:
+ upconv = nn.ConvTranspose2d(inner_nc * 2, outer_nc,
+ kernel_size=4, stride=2,
+ padding=1)
+ down = [downconv]
+ up = [uprelu, upconv, nn.Tanh()]
+ model = down + [submodule] + up
+ elif innermost:
+ upconv = nn.ConvTranspose2d(inner_nc, outer_nc,
+ kernel_size=4, stride=2,
+ padding=1, bias=use_bias)
+ down = [downrelu, downconv]
+ up = [uprelu, upconv, upnorm]
+ model = down + up
+ else:
+ upconv = nn.ConvTranspose2d(inner_nc * 2, outer_nc,
+ kernel_size=4, stride=2,
+ padding=1, bias=use_bias)
+ down = [downrelu, downconv, downnorm]
+ up = [uprelu, upconv, upnorm]
+
+ if use_dropout:
+ model = down + [submodule] + up + [nn.Dropout(0.5)]
+ else:
+ model = down + [submodule] + up
+
+ self.model = nn.Sequential(*model)
+
+ def forward(self, x):
+ if self.outermost:
+ return self.model(x)
+ else: # add skip connections
+ return torch.cat([x, self.model(x)], 1)
+
+
+class LineartAnimeDetector:
+ def __init__(self):
+ remote_model_path = "https://huggingface.co/lllyasviel/Annotators/resolve/main/netG.pth"
+ modelpath = os.path.join(annotator_ckpts_path, "netG.pth")
+ if not os.path.exists(modelpath):
+ from basicsr.utils.download_util import load_file_from_url
+ load_file_from_url(remote_model_path, model_dir=annotator_ckpts_path)
+ norm_layer = functools.partial(nn.InstanceNorm2d, affine=False, track_running_stats=False)
+ net = UnetGenerator(3, 1, 8, 64, norm_layer=norm_layer, use_dropout=False)
+ ckpt = torch.load(modelpath)
+ for key in list(ckpt.keys()):
+ if 'module.' in key:
+ ckpt[key.replace('module.', '')] = ckpt[key]
+ del ckpt[key]
+ net.load_state_dict(ckpt)
+ net = net.cuda()
+ net.eval()
+ self.model = net
+
+ def __call__(self, input_image):
+ H, W, C = input_image.shape
+ Hn = 256 * int(np.ceil(float(H) / 256.0))
+ Wn = 256 * int(np.ceil(float(W) / 256.0))
+ img = cv2.resize(input_image, (Wn, Hn), interpolation=cv2.INTER_CUBIC)
+ with torch.no_grad():
+ image_feed = torch.from_numpy(img).float().cuda()
+ image_feed = image_feed / 127.5 - 1.0
+ image_feed = rearrange(image_feed, 'h w c -> 1 c h w')
+
+ line = self.model(image_feed)[0, 0] * 127.5 + 127.5
+ line = line.cpu().numpy()
+
+ line = cv2.resize(line, (W, H), interpolation=cv2.INTER_CUBIC)
+ line = line.clip(0, 255).astype(np.uint8)
+ return line
+
diff --git a/ControlNet-v1-1-nightly-main/annotator/midas/LICENSE b/ControlNet-v1-1-nightly-main/annotator/midas/LICENSE
new file mode 100644
index 0000000000000000000000000000000000000000..277b5c11be103f028a8d10985139f1da10c2f08e
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/midas/LICENSE
@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2019 Intel ISL (Intel Intelligent Systems Lab)
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
diff --git a/ControlNet-v1-1-nightly-main/annotator/midas/__init__.py b/ControlNet-v1-1-nightly-main/annotator/midas/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..237e1c07a6a0a1d830bac75aa8e1af4f3c8d59b0
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/midas/__init__.py
@@ -0,0 +1,31 @@
+# Midas Depth Estimation
+# From https://github.com/isl-org/MiDaS
+# MIT LICENSE
+
+import cv2
+import numpy as np
+import torch
+
+from einops import rearrange
+from .api import MiDaSInference
+
+
+class MidasDetector:
+ def __init__(self):
+ self.model = MiDaSInference(model_type="dpt_hybrid").cuda()
+
+ def __call__(self, input_image):
+ assert input_image.ndim == 3
+ image_depth = input_image
+ with torch.no_grad():
+ image_depth = torch.from_numpy(image_depth).float().cuda()
+ image_depth = image_depth / 127.5 - 1.0
+ image_depth = rearrange(image_depth, 'h w c -> 1 c h w')
+ depth = self.model(image_depth)[0]
+
+ depth -= torch.min(depth)
+ depth /= torch.max(depth)
+ depth = depth.cpu().numpy()
+ depth_image = (depth * 255.0).clip(0, 255).astype(np.uint8)
+
+ return depth_image
diff --git a/ControlNet-v1-1-nightly-main/annotator/midas/api.py b/ControlNet-v1-1-nightly-main/annotator/midas/api.py
new file mode 100644
index 0000000000000000000000000000000000000000..0ddcf47aba1db4d533c7540fa984f4757f4fbd80
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/midas/api.py
@@ -0,0 +1,169 @@
+# based on https://github.com/isl-org/MiDaS
+
+import cv2
+import os
+import torch
+import torch.nn as nn
+from torchvision.transforms import Compose
+
+from .midas.dpt_depth import DPTDepthModel
+from .midas.midas_net import MidasNet
+from .midas.midas_net_custom import MidasNet_small
+from .midas.transforms import Resize, NormalizeImage, PrepareForNet
+from annotator.util import annotator_ckpts_path
+
+
+ISL_PATHS = {
+ "dpt_large": os.path.join(annotator_ckpts_path, "dpt_large-midas-2f21e586.pt"),
+ "dpt_hybrid": os.path.join(annotator_ckpts_path, "dpt_hybrid-midas-501f0c75.pt"),
+ "midas_v21": "",
+ "midas_v21_small": "",
+}
+
+remote_model_path = "https://huggingface.co/lllyasviel/Annotators/resolve/main/dpt_hybrid-midas-501f0c75.pt"
+
+
+def disabled_train(self, mode=True):
+ """Overwrite model.train with this function to make sure train/eval mode
+ does not change anymore."""
+ return self
+
+
+def load_midas_transform(model_type):
+ # https://github.com/isl-org/MiDaS/blob/master/run.py
+ # load transform only
+ if model_type == "dpt_large": # DPT-Large
+ net_w, net_h = 384, 384
+ resize_mode = "minimal"
+ normalization = NormalizeImage(mean=[0.5, 0.5, 0.5], std=[0.5, 0.5, 0.5])
+
+ elif model_type == "dpt_hybrid": # DPT-Hybrid
+ net_w, net_h = 384, 384
+ resize_mode = "minimal"
+ normalization = NormalizeImage(mean=[0.5, 0.5, 0.5], std=[0.5, 0.5, 0.5])
+
+ elif model_type == "midas_v21":
+ net_w, net_h = 384, 384
+ resize_mode = "upper_bound"
+ normalization = NormalizeImage(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])
+
+ elif model_type == "midas_v21_small":
+ net_w, net_h = 256, 256
+ resize_mode = "upper_bound"
+ normalization = NormalizeImage(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])
+
+ else:
+ assert False, f"model_type '{model_type}' not implemented, use: --model_type large"
+
+ transform = Compose(
+ [
+ Resize(
+ net_w,
+ net_h,
+ resize_target=None,
+ keep_aspect_ratio=True,
+ ensure_multiple_of=32,
+ resize_method=resize_mode,
+ image_interpolation_method=cv2.INTER_CUBIC,
+ ),
+ normalization,
+ PrepareForNet(),
+ ]
+ )
+
+ return transform
+
+
+def load_model(model_type):
+ # https://github.com/isl-org/MiDaS/blob/master/run.py
+ # load network
+ model_path = ISL_PATHS[model_type]
+ if model_type == "dpt_large": # DPT-Large
+ model = DPTDepthModel(
+ path=model_path,
+ backbone="vitl16_384",
+ non_negative=True,
+ )
+ net_w, net_h = 384, 384
+ resize_mode = "minimal"
+ normalization = NormalizeImage(mean=[0.5, 0.5, 0.5], std=[0.5, 0.5, 0.5])
+
+ elif model_type == "dpt_hybrid": # DPT-Hybrid
+ if not os.path.exists(model_path):
+ from basicsr.utils.download_util import load_file_from_url
+ load_file_from_url(remote_model_path, model_dir=annotator_ckpts_path)
+
+ model = DPTDepthModel(
+ path=model_path,
+ backbone="vitb_rn50_384",
+ non_negative=True,
+ )
+ net_w, net_h = 384, 384
+ resize_mode = "minimal"
+ normalization = NormalizeImage(mean=[0.5, 0.5, 0.5], std=[0.5, 0.5, 0.5])
+
+ elif model_type == "midas_v21":
+ model = MidasNet(model_path, non_negative=True)
+ net_w, net_h = 384, 384
+ resize_mode = "upper_bound"
+ normalization = NormalizeImage(
+ mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]
+ )
+
+ elif model_type == "midas_v21_small":
+ model = MidasNet_small(model_path, features=64, backbone="efficientnet_lite3", exportable=True,
+ non_negative=True, blocks={'expand': True})
+ net_w, net_h = 256, 256
+ resize_mode = "upper_bound"
+ normalization = NormalizeImage(
+ mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]
+ )
+
+ else:
+ print(f"model_type '{model_type}' not implemented, use: --model_type large")
+ assert False
+
+ transform = Compose(
+ [
+ Resize(
+ net_w,
+ net_h,
+ resize_target=None,
+ keep_aspect_ratio=True,
+ ensure_multiple_of=32,
+ resize_method=resize_mode,
+ image_interpolation_method=cv2.INTER_CUBIC,
+ ),
+ normalization,
+ PrepareForNet(),
+ ]
+ )
+
+ return model.eval(), transform
+
+
+class MiDaSInference(nn.Module):
+ MODEL_TYPES_TORCH_HUB = [
+ "DPT_Large",
+ "DPT_Hybrid",
+ "MiDaS_small"
+ ]
+ MODEL_TYPES_ISL = [
+ "dpt_large",
+ "dpt_hybrid",
+ "midas_v21",
+ "midas_v21_small",
+ ]
+
+ def __init__(self, model_type):
+ super().__init__()
+ assert (model_type in self.MODEL_TYPES_ISL)
+ model, _ = load_model(model_type)
+ self.model = model
+ self.model.train = disabled_train
+
+ def forward(self, x):
+ with torch.no_grad():
+ prediction = self.model(x)
+ return prediction
+
diff --git a/ControlNet-v1-1-nightly-main/annotator/midas/midas/__init__.py b/ControlNet-v1-1-nightly-main/annotator/midas/midas/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/ControlNet-v1-1-nightly-main/annotator/midas/midas/base_model.py b/ControlNet-v1-1-nightly-main/annotator/midas/midas/base_model.py
new file mode 100644
index 0000000000000000000000000000000000000000..5cf430239b47ec5ec07531263f26f5c24a2311cd
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/midas/midas/base_model.py
@@ -0,0 +1,16 @@
+import torch
+
+
+class BaseModel(torch.nn.Module):
+ def load(self, path):
+ """Load model from file.
+
+ Args:
+ path (str): file path
+ """
+ parameters = torch.load(path, map_location=torch.device('cpu'))
+
+ if "optimizer" in parameters:
+ parameters = parameters["model"]
+
+ self.load_state_dict(parameters)
diff --git a/ControlNet-v1-1-nightly-main/annotator/midas/midas/blocks.py b/ControlNet-v1-1-nightly-main/annotator/midas/midas/blocks.py
new file mode 100644
index 0000000000000000000000000000000000000000..2145d18fa98060a618536d9a64fe6589e9be4f78
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/midas/midas/blocks.py
@@ -0,0 +1,342 @@
+import torch
+import torch.nn as nn
+
+from .vit import (
+ _make_pretrained_vitb_rn50_384,
+ _make_pretrained_vitl16_384,
+ _make_pretrained_vitb16_384,
+ forward_vit,
+)
+
+def _make_encoder(backbone, features, use_pretrained, groups=1, expand=False, exportable=True, hooks=None, use_vit_only=False, use_readout="ignore",):
+ if backbone == "vitl16_384":
+ pretrained = _make_pretrained_vitl16_384(
+ use_pretrained, hooks=hooks, use_readout=use_readout
+ )
+ scratch = _make_scratch(
+ [256, 512, 1024, 1024], features, groups=groups, expand=expand
+ ) # ViT-L/16 - 85.0% Top1 (backbone)
+ elif backbone == "vitb_rn50_384":
+ pretrained = _make_pretrained_vitb_rn50_384(
+ use_pretrained,
+ hooks=hooks,
+ use_vit_only=use_vit_only,
+ use_readout=use_readout,
+ )
+ scratch = _make_scratch(
+ [256, 512, 768, 768], features, groups=groups, expand=expand
+ ) # ViT-H/16 - 85.0% Top1 (backbone)
+ elif backbone == "vitb16_384":
+ pretrained = _make_pretrained_vitb16_384(
+ use_pretrained, hooks=hooks, use_readout=use_readout
+ )
+ scratch = _make_scratch(
+ [96, 192, 384, 768], features, groups=groups, expand=expand
+ ) # ViT-B/16 - 84.6% Top1 (backbone)
+ elif backbone == "resnext101_wsl":
+ pretrained = _make_pretrained_resnext101_wsl(use_pretrained)
+ scratch = _make_scratch([256, 512, 1024, 2048], features, groups=groups, expand=expand) # efficientnet_lite3
+ elif backbone == "efficientnet_lite3":
+ pretrained = _make_pretrained_efficientnet_lite3(use_pretrained, exportable=exportable)
+ scratch = _make_scratch([32, 48, 136, 384], features, groups=groups, expand=expand) # efficientnet_lite3
+ else:
+ print(f"Backbone '{backbone}' not implemented")
+ assert False
+
+ return pretrained, scratch
+
+
+def _make_scratch(in_shape, out_shape, groups=1, expand=False):
+ scratch = nn.Module()
+
+ out_shape1 = out_shape
+ out_shape2 = out_shape
+ out_shape3 = out_shape
+ out_shape4 = out_shape
+ if expand==True:
+ out_shape1 = out_shape
+ out_shape2 = out_shape*2
+ out_shape3 = out_shape*4
+ out_shape4 = out_shape*8
+
+ scratch.layer1_rn = nn.Conv2d(
+ in_shape[0], out_shape1, kernel_size=3, stride=1, padding=1, bias=False, groups=groups
+ )
+ scratch.layer2_rn = nn.Conv2d(
+ in_shape[1], out_shape2, kernel_size=3, stride=1, padding=1, bias=False, groups=groups
+ )
+ scratch.layer3_rn = nn.Conv2d(
+ in_shape[2], out_shape3, kernel_size=3, stride=1, padding=1, bias=False, groups=groups
+ )
+ scratch.layer4_rn = nn.Conv2d(
+ in_shape[3], out_shape4, kernel_size=3, stride=1, padding=1, bias=False, groups=groups
+ )
+
+ return scratch
+
+
+def _make_pretrained_efficientnet_lite3(use_pretrained, exportable=False):
+ efficientnet = torch.hub.load(
+ "rwightman/gen-efficientnet-pytorch",
+ "tf_efficientnet_lite3",
+ pretrained=use_pretrained,
+ exportable=exportable
+ )
+ return _make_efficientnet_backbone(efficientnet)
+
+
+def _make_efficientnet_backbone(effnet):
+ pretrained = nn.Module()
+
+ pretrained.layer1 = nn.Sequential(
+ effnet.conv_stem, effnet.bn1, effnet.act1, *effnet.blocks[0:2]
+ )
+ pretrained.layer2 = nn.Sequential(*effnet.blocks[2:3])
+ pretrained.layer3 = nn.Sequential(*effnet.blocks[3:5])
+ pretrained.layer4 = nn.Sequential(*effnet.blocks[5:9])
+
+ return pretrained
+
+
+def _make_resnet_backbone(resnet):
+ pretrained = nn.Module()
+ pretrained.layer1 = nn.Sequential(
+ resnet.conv1, resnet.bn1, resnet.relu, resnet.maxpool, resnet.layer1
+ )
+
+ pretrained.layer2 = resnet.layer2
+ pretrained.layer3 = resnet.layer3
+ pretrained.layer4 = resnet.layer4
+
+ return pretrained
+
+
+def _make_pretrained_resnext101_wsl(use_pretrained):
+ resnet = torch.hub.load("facebookresearch/WSL-Images", "resnext101_32x8d_wsl")
+ return _make_resnet_backbone(resnet)
+
+
+
+class Interpolate(nn.Module):
+ """Interpolation module.
+ """
+
+ def __init__(self, scale_factor, mode, align_corners=False):
+ """Init.
+
+ Args:
+ scale_factor (float): scaling
+ mode (str): interpolation mode
+ """
+ super(Interpolate, self).__init__()
+
+ self.interp = nn.functional.interpolate
+ self.scale_factor = scale_factor
+ self.mode = mode
+ self.align_corners = align_corners
+
+ def forward(self, x):
+ """Forward pass.
+
+ Args:
+ x (tensor): input
+
+ Returns:
+ tensor: interpolated data
+ """
+
+ x = self.interp(
+ x, scale_factor=self.scale_factor, mode=self.mode, align_corners=self.align_corners
+ )
+
+ return x
+
+
+class ResidualConvUnit(nn.Module):
+ """Residual convolution module.
+ """
+
+ def __init__(self, features):
+ """Init.
+
+ Args:
+ features (int): number of features
+ """
+ super().__init__()
+
+ self.conv1 = nn.Conv2d(
+ features, features, kernel_size=3, stride=1, padding=1, bias=True
+ )
+
+ self.conv2 = nn.Conv2d(
+ features, features, kernel_size=3, stride=1, padding=1, bias=True
+ )
+
+ self.relu = nn.ReLU(inplace=True)
+
+ def forward(self, x):
+ """Forward pass.
+
+ Args:
+ x (tensor): input
+
+ Returns:
+ tensor: output
+ """
+ out = self.relu(x)
+ out = self.conv1(out)
+ out = self.relu(out)
+ out = self.conv2(out)
+
+ return out + x
+
+
+class FeatureFusionBlock(nn.Module):
+ """Feature fusion block.
+ """
+
+ def __init__(self, features):
+ """Init.
+
+ Args:
+ features (int): number of features
+ """
+ super(FeatureFusionBlock, self).__init__()
+
+ self.resConfUnit1 = ResidualConvUnit(features)
+ self.resConfUnit2 = ResidualConvUnit(features)
+
+ def forward(self, *xs):
+ """Forward pass.
+
+ Returns:
+ tensor: output
+ """
+ output = xs[0]
+
+ if len(xs) == 2:
+ output += self.resConfUnit1(xs[1])
+
+ output = self.resConfUnit2(output)
+
+ output = nn.functional.interpolate(
+ output, scale_factor=2, mode="bilinear", align_corners=True
+ )
+
+ return output
+
+
+
+
+class ResidualConvUnit_custom(nn.Module):
+ """Residual convolution module.
+ """
+
+ def __init__(self, features, activation, bn):
+ """Init.
+
+ Args:
+ features (int): number of features
+ """
+ super().__init__()
+
+ self.bn = bn
+
+ self.groups=1
+
+ self.conv1 = nn.Conv2d(
+ features, features, kernel_size=3, stride=1, padding=1, bias=True, groups=self.groups
+ )
+
+ self.conv2 = nn.Conv2d(
+ features, features, kernel_size=3, stride=1, padding=1, bias=True, groups=self.groups
+ )
+
+ if self.bn==True:
+ self.bn1 = nn.BatchNorm2d(features)
+ self.bn2 = nn.BatchNorm2d(features)
+
+ self.activation = activation
+
+ self.skip_add = nn.quantized.FloatFunctional()
+
+ def forward(self, x):
+ """Forward pass.
+
+ Args:
+ x (tensor): input
+
+ Returns:
+ tensor: output
+ """
+
+ out = self.activation(x)
+ out = self.conv1(out)
+ if self.bn==True:
+ out = self.bn1(out)
+
+ out = self.activation(out)
+ out = self.conv2(out)
+ if self.bn==True:
+ out = self.bn2(out)
+
+ if self.groups > 1:
+ out = self.conv_merge(out)
+
+ return self.skip_add.add(out, x)
+
+ # return out + x
+
+
+class FeatureFusionBlock_custom(nn.Module):
+ """Feature fusion block.
+ """
+
+ def __init__(self, features, activation, deconv=False, bn=False, expand=False, align_corners=True):
+ """Init.
+
+ Args:
+ features (int): number of features
+ """
+ super(FeatureFusionBlock_custom, self).__init__()
+
+ self.deconv = deconv
+ self.align_corners = align_corners
+
+ self.groups=1
+
+ self.expand = expand
+ out_features = features
+ if self.expand==True:
+ out_features = features//2
+
+ self.out_conv = nn.Conv2d(features, out_features, kernel_size=1, stride=1, padding=0, bias=True, groups=1)
+
+ self.resConfUnit1 = ResidualConvUnit_custom(features, activation, bn)
+ self.resConfUnit2 = ResidualConvUnit_custom(features, activation, bn)
+
+ self.skip_add = nn.quantized.FloatFunctional()
+
+ def forward(self, *xs):
+ """Forward pass.
+
+ Returns:
+ tensor: output
+ """
+ output = xs[0]
+
+ if len(xs) == 2:
+ res = self.resConfUnit1(xs[1])
+ output = self.skip_add.add(output, res)
+ # output += res
+
+ output = self.resConfUnit2(output)
+
+ output = nn.functional.interpolate(
+ output, scale_factor=2, mode="bilinear", align_corners=self.align_corners
+ )
+
+ output = self.out_conv(output)
+
+ return output
+
diff --git a/ControlNet-v1-1-nightly-main/annotator/midas/midas/dpt_depth.py b/ControlNet-v1-1-nightly-main/annotator/midas/midas/dpt_depth.py
new file mode 100644
index 0000000000000000000000000000000000000000..4e9aab5d2767dffea39da5b3f30e2798688216f1
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/midas/midas/dpt_depth.py
@@ -0,0 +1,109 @@
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+
+from .base_model import BaseModel
+from .blocks import (
+ FeatureFusionBlock,
+ FeatureFusionBlock_custom,
+ Interpolate,
+ _make_encoder,
+ forward_vit,
+)
+
+
+def _make_fusion_block(features, use_bn):
+ return FeatureFusionBlock_custom(
+ features,
+ nn.ReLU(False),
+ deconv=False,
+ bn=use_bn,
+ expand=False,
+ align_corners=True,
+ )
+
+
+class DPT(BaseModel):
+ def __init__(
+ self,
+ head,
+ features=256,
+ backbone="vitb_rn50_384",
+ readout="project",
+ channels_last=False,
+ use_bn=False,
+ ):
+
+ super(DPT, self).__init__()
+
+ self.channels_last = channels_last
+
+ hooks = {
+ "vitb_rn50_384": [0, 1, 8, 11],
+ "vitb16_384": [2, 5, 8, 11],
+ "vitl16_384": [5, 11, 17, 23],
+ }
+
+ # Instantiate backbone and reassemble blocks
+ self.pretrained, self.scratch = _make_encoder(
+ backbone,
+ features,
+ False, # Set to true of you want to train from scratch, uses ImageNet weights
+ groups=1,
+ expand=False,
+ exportable=False,
+ hooks=hooks[backbone],
+ use_readout=readout,
+ )
+
+ self.scratch.refinenet1 = _make_fusion_block(features, use_bn)
+ self.scratch.refinenet2 = _make_fusion_block(features, use_bn)
+ self.scratch.refinenet3 = _make_fusion_block(features, use_bn)
+ self.scratch.refinenet4 = _make_fusion_block(features, use_bn)
+
+ self.scratch.output_conv = head
+
+
+ def forward(self, x):
+ if self.channels_last == True:
+ x.contiguous(memory_format=torch.channels_last)
+
+ layer_1, layer_2, layer_3, layer_4 = forward_vit(self.pretrained, x)
+
+ layer_1_rn = self.scratch.layer1_rn(layer_1)
+ layer_2_rn = self.scratch.layer2_rn(layer_2)
+ layer_3_rn = self.scratch.layer3_rn(layer_3)
+ layer_4_rn = self.scratch.layer4_rn(layer_4)
+
+ path_4 = self.scratch.refinenet4(layer_4_rn)
+ path_3 = self.scratch.refinenet3(path_4, layer_3_rn)
+ path_2 = self.scratch.refinenet2(path_3, layer_2_rn)
+ path_1 = self.scratch.refinenet1(path_2, layer_1_rn)
+
+ out = self.scratch.output_conv(path_1)
+
+ return out
+
+
+class DPTDepthModel(DPT):
+ def __init__(self, path=None, non_negative=True, **kwargs):
+ features = kwargs["features"] if "features" in kwargs else 256
+
+ head = nn.Sequential(
+ nn.Conv2d(features, features // 2, kernel_size=3, stride=1, padding=1),
+ Interpolate(scale_factor=2, mode="bilinear", align_corners=True),
+ nn.Conv2d(features // 2, 32, kernel_size=3, stride=1, padding=1),
+ nn.ReLU(True),
+ nn.Conv2d(32, 1, kernel_size=1, stride=1, padding=0),
+ nn.ReLU(True) if non_negative else nn.Identity(),
+ nn.Identity(),
+ )
+
+ super().__init__(head, **kwargs)
+
+ if path is not None:
+ self.load(path)
+
+ def forward(self, x):
+ return super().forward(x).squeeze(dim=1)
+
diff --git a/ControlNet-v1-1-nightly-main/annotator/midas/midas/midas_net.py b/ControlNet-v1-1-nightly-main/annotator/midas/midas/midas_net.py
new file mode 100644
index 0000000000000000000000000000000000000000..8a954977800b0a0f48807e80fa63041910e33c1f
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/midas/midas/midas_net.py
@@ -0,0 +1,76 @@
+"""MidashNet: Network for monocular depth estimation trained by mixing several datasets.
+This file contains code that is adapted from
+https://github.com/thomasjpfan/pytorch_refinenet/blob/master/pytorch_refinenet/refinenet/refinenet_4cascade.py
+"""
+import torch
+import torch.nn as nn
+
+from .base_model import BaseModel
+from .blocks import FeatureFusionBlock, Interpolate, _make_encoder
+
+
+class MidasNet(BaseModel):
+ """Network for monocular depth estimation.
+ """
+
+ def __init__(self, path=None, features=256, non_negative=True):
+ """Init.
+
+ Args:
+ path (str, optional): Path to saved model. Defaults to None.
+ features (int, optional): Number of features. Defaults to 256.
+ backbone (str, optional): Backbone network for encoder. Defaults to resnet50
+ """
+ print("Loading weights: ", path)
+
+ super(MidasNet, self).__init__()
+
+ use_pretrained = False if path is None else True
+
+ self.pretrained, self.scratch = _make_encoder(backbone="resnext101_wsl", features=features, use_pretrained=use_pretrained)
+
+ self.scratch.refinenet4 = FeatureFusionBlock(features)
+ self.scratch.refinenet3 = FeatureFusionBlock(features)
+ self.scratch.refinenet2 = FeatureFusionBlock(features)
+ self.scratch.refinenet1 = FeatureFusionBlock(features)
+
+ self.scratch.output_conv = nn.Sequential(
+ nn.Conv2d(features, 128, kernel_size=3, stride=1, padding=1),
+ Interpolate(scale_factor=2, mode="bilinear"),
+ nn.Conv2d(128, 32, kernel_size=3, stride=1, padding=1),
+ nn.ReLU(True),
+ nn.Conv2d(32, 1, kernel_size=1, stride=1, padding=0),
+ nn.ReLU(True) if non_negative else nn.Identity(),
+ )
+
+ if path:
+ self.load(path)
+
+ def forward(self, x):
+ """Forward pass.
+
+ Args:
+ x (tensor): input data (image)
+
+ Returns:
+ tensor: depth
+ """
+
+ layer_1 = self.pretrained.layer1(x)
+ layer_2 = self.pretrained.layer2(layer_1)
+ layer_3 = self.pretrained.layer3(layer_2)
+ layer_4 = self.pretrained.layer4(layer_3)
+
+ layer_1_rn = self.scratch.layer1_rn(layer_1)
+ layer_2_rn = self.scratch.layer2_rn(layer_2)
+ layer_3_rn = self.scratch.layer3_rn(layer_3)
+ layer_4_rn = self.scratch.layer4_rn(layer_4)
+
+ path_4 = self.scratch.refinenet4(layer_4_rn)
+ path_3 = self.scratch.refinenet3(path_4, layer_3_rn)
+ path_2 = self.scratch.refinenet2(path_3, layer_2_rn)
+ path_1 = self.scratch.refinenet1(path_2, layer_1_rn)
+
+ out = self.scratch.output_conv(path_1)
+
+ return torch.squeeze(out, dim=1)
diff --git a/ControlNet-v1-1-nightly-main/annotator/midas/midas/midas_net_custom.py b/ControlNet-v1-1-nightly-main/annotator/midas/midas/midas_net_custom.py
new file mode 100644
index 0000000000000000000000000000000000000000..50e4acb5e53d5fabefe3dde16ab49c33c2b7797c
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/midas/midas/midas_net_custom.py
@@ -0,0 +1,128 @@
+"""MidashNet: Network for monocular depth estimation trained by mixing several datasets.
+This file contains code that is adapted from
+https://github.com/thomasjpfan/pytorch_refinenet/blob/master/pytorch_refinenet/refinenet/refinenet_4cascade.py
+"""
+import torch
+import torch.nn as nn
+
+from .base_model import BaseModel
+from .blocks import FeatureFusionBlock, FeatureFusionBlock_custom, Interpolate, _make_encoder
+
+
+class MidasNet_small(BaseModel):
+ """Network for monocular depth estimation.
+ """
+
+ def __init__(self, path=None, features=64, backbone="efficientnet_lite3", non_negative=True, exportable=True, channels_last=False, align_corners=True,
+ blocks={'expand': True}):
+ """Init.
+
+ Args:
+ path (str, optional): Path to saved model. Defaults to None.
+ features (int, optional): Number of features. Defaults to 256.
+ backbone (str, optional): Backbone network for encoder. Defaults to resnet50
+ """
+ print("Loading weights: ", path)
+
+ super(MidasNet_small, self).__init__()
+
+ use_pretrained = False if path else True
+
+ self.channels_last = channels_last
+ self.blocks = blocks
+ self.backbone = backbone
+
+ self.groups = 1
+
+ features1=features
+ features2=features
+ features3=features
+ features4=features
+ self.expand = False
+ if "expand" in self.blocks and self.blocks['expand'] == True:
+ self.expand = True
+ features1=features
+ features2=features*2
+ features3=features*4
+ features4=features*8
+
+ self.pretrained, self.scratch = _make_encoder(self.backbone, features, use_pretrained, groups=self.groups, expand=self.expand, exportable=exportable)
+
+ self.scratch.activation = nn.ReLU(False)
+
+ self.scratch.refinenet4 = FeatureFusionBlock_custom(features4, self.scratch.activation, deconv=False, bn=False, expand=self.expand, align_corners=align_corners)
+ self.scratch.refinenet3 = FeatureFusionBlock_custom(features3, self.scratch.activation, deconv=False, bn=False, expand=self.expand, align_corners=align_corners)
+ self.scratch.refinenet2 = FeatureFusionBlock_custom(features2, self.scratch.activation, deconv=False, bn=False, expand=self.expand, align_corners=align_corners)
+ self.scratch.refinenet1 = FeatureFusionBlock_custom(features1, self.scratch.activation, deconv=False, bn=False, align_corners=align_corners)
+
+
+ self.scratch.output_conv = nn.Sequential(
+ nn.Conv2d(features, features//2, kernel_size=3, stride=1, padding=1, groups=self.groups),
+ Interpolate(scale_factor=2, mode="bilinear"),
+ nn.Conv2d(features//2, 32, kernel_size=3, stride=1, padding=1),
+ self.scratch.activation,
+ nn.Conv2d(32, 1, kernel_size=1, stride=1, padding=0),
+ nn.ReLU(True) if non_negative else nn.Identity(),
+ nn.Identity(),
+ )
+
+ if path:
+ self.load(path)
+
+
+ def forward(self, x):
+ """Forward pass.
+
+ Args:
+ x (tensor): input data (image)
+
+ Returns:
+ tensor: depth
+ """
+ if self.channels_last==True:
+ print("self.channels_last = ", self.channels_last)
+ x.contiguous(memory_format=torch.channels_last)
+
+
+ layer_1 = self.pretrained.layer1(x)
+ layer_2 = self.pretrained.layer2(layer_1)
+ layer_3 = self.pretrained.layer3(layer_2)
+ layer_4 = self.pretrained.layer4(layer_3)
+
+ layer_1_rn = self.scratch.layer1_rn(layer_1)
+ layer_2_rn = self.scratch.layer2_rn(layer_2)
+ layer_3_rn = self.scratch.layer3_rn(layer_3)
+ layer_4_rn = self.scratch.layer4_rn(layer_4)
+
+
+ path_4 = self.scratch.refinenet4(layer_4_rn)
+ path_3 = self.scratch.refinenet3(path_4, layer_3_rn)
+ path_2 = self.scratch.refinenet2(path_3, layer_2_rn)
+ path_1 = self.scratch.refinenet1(path_2, layer_1_rn)
+
+ out = self.scratch.output_conv(path_1)
+
+ return torch.squeeze(out, dim=1)
+
+
+
+def fuse_model(m):
+ prev_previous_type = nn.Identity()
+ prev_previous_name = ''
+ previous_type = nn.Identity()
+ previous_name = ''
+ for name, module in m.named_modules():
+ if prev_previous_type == nn.Conv2d and previous_type == nn.BatchNorm2d and type(module) == nn.ReLU:
+ # print("FUSED ", prev_previous_name, previous_name, name)
+ torch.quantization.fuse_modules(m, [prev_previous_name, previous_name, name], inplace=True)
+ elif prev_previous_type == nn.Conv2d and previous_type == nn.BatchNorm2d:
+ # print("FUSED ", prev_previous_name, previous_name)
+ torch.quantization.fuse_modules(m, [prev_previous_name, previous_name], inplace=True)
+ # elif previous_type == nn.Conv2d and type(module) == nn.ReLU:
+ # print("FUSED ", previous_name, name)
+ # torch.quantization.fuse_modules(m, [previous_name, name], inplace=True)
+
+ prev_previous_type = previous_type
+ prev_previous_name = previous_name
+ previous_type = type(module)
+ previous_name = name
\ No newline at end of file
diff --git a/ControlNet-v1-1-nightly-main/annotator/midas/midas/transforms.py b/ControlNet-v1-1-nightly-main/annotator/midas/midas/transforms.py
new file mode 100644
index 0000000000000000000000000000000000000000..350cbc11662633ad7f8968eb10be2e7de6e384e9
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/midas/midas/transforms.py
@@ -0,0 +1,234 @@
+import numpy as np
+import cv2
+import math
+
+
+def apply_min_size(sample, size, image_interpolation_method=cv2.INTER_AREA):
+ """Rezise the sample to ensure the given size. Keeps aspect ratio.
+
+ Args:
+ sample (dict): sample
+ size (tuple): image size
+
+ Returns:
+ tuple: new size
+ """
+ shape = list(sample["disparity"].shape)
+
+ if shape[0] >= size[0] and shape[1] >= size[1]:
+ return sample
+
+ scale = [0, 0]
+ scale[0] = size[0] / shape[0]
+ scale[1] = size[1] / shape[1]
+
+ scale = max(scale)
+
+ shape[0] = math.ceil(scale * shape[0])
+ shape[1] = math.ceil(scale * shape[1])
+
+ # resize
+ sample["image"] = cv2.resize(
+ sample["image"], tuple(shape[::-1]), interpolation=image_interpolation_method
+ )
+
+ sample["disparity"] = cv2.resize(
+ sample["disparity"], tuple(shape[::-1]), interpolation=cv2.INTER_NEAREST
+ )
+ sample["mask"] = cv2.resize(
+ sample["mask"].astype(np.float32),
+ tuple(shape[::-1]),
+ interpolation=cv2.INTER_NEAREST,
+ )
+ sample["mask"] = sample["mask"].astype(bool)
+
+ return tuple(shape)
+
+
+class Resize(object):
+ """Resize sample to given size (width, height).
+ """
+
+ def __init__(
+ self,
+ width,
+ height,
+ resize_target=True,
+ keep_aspect_ratio=False,
+ ensure_multiple_of=1,
+ resize_method="lower_bound",
+ image_interpolation_method=cv2.INTER_AREA,
+ ):
+ """Init.
+
+ Args:
+ width (int): desired output width
+ height (int): desired output height
+ resize_target (bool, optional):
+ True: Resize the full sample (image, mask, target).
+ False: Resize image only.
+ Defaults to True.
+ keep_aspect_ratio (bool, optional):
+ True: Keep the aspect ratio of the input sample.
+ Output sample might not have the given width and height, and
+ resize behaviour depends on the parameter 'resize_method'.
+ Defaults to False.
+ ensure_multiple_of (int, optional):
+ Output width and height is constrained to be multiple of this parameter.
+ Defaults to 1.
+ resize_method (str, optional):
+ "lower_bound": Output will be at least as large as the given size.
+ "upper_bound": Output will be at max as large as the given size. (Output size might be smaller than given size.)
+ "minimal": Scale as least as possible. (Output size might be smaller than given size.)
+ Defaults to "lower_bound".
+ """
+ self.__width = width
+ self.__height = height
+
+ self.__resize_target = resize_target
+ self.__keep_aspect_ratio = keep_aspect_ratio
+ self.__multiple_of = ensure_multiple_of
+ self.__resize_method = resize_method
+ self.__image_interpolation_method = image_interpolation_method
+
+ def constrain_to_multiple_of(self, x, min_val=0, max_val=None):
+ y = (np.round(x / self.__multiple_of) * self.__multiple_of).astype(int)
+
+ if max_val is not None and y > max_val:
+ y = (np.floor(x / self.__multiple_of) * self.__multiple_of).astype(int)
+
+ if y < min_val:
+ y = (np.ceil(x / self.__multiple_of) * self.__multiple_of).astype(int)
+
+ return y
+
+ def get_size(self, width, height):
+ # determine new height and width
+ scale_height = self.__height / height
+ scale_width = self.__width / width
+
+ if self.__keep_aspect_ratio:
+ if self.__resize_method == "lower_bound":
+ # scale such that output size is lower bound
+ if scale_width > scale_height:
+ # fit width
+ scale_height = scale_width
+ else:
+ # fit height
+ scale_width = scale_height
+ elif self.__resize_method == "upper_bound":
+ # scale such that output size is upper bound
+ if scale_width < scale_height:
+ # fit width
+ scale_height = scale_width
+ else:
+ # fit height
+ scale_width = scale_height
+ elif self.__resize_method == "minimal":
+ # scale as least as possbile
+ if abs(1 - scale_width) < abs(1 - scale_height):
+ # fit width
+ scale_height = scale_width
+ else:
+ # fit height
+ scale_width = scale_height
+ else:
+ raise ValueError(
+ f"resize_method {self.__resize_method} not implemented"
+ )
+
+ if self.__resize_method == "lower_bound":
+ new_height = self.constrain_to_multiple_of(
+ scale_height * height, min_val=self.__height
+ )
+ new_width = self.constrain_to_multiple_of(
+ scale_width * width, min_val=self.__width
+ )
+ elif self.__resize_method == "upper_bound":
+ new_height = self.constrain_to_multiple_of(
+ scale_height * height, max_val=self.__height
+ )
+ new_width = self.constrain_to_multiple_of(
+ scale_width * width, max_val=self.__width
+ )
+ elif self.__resize_method == "minimal":
+ new_height = self.constrain_to_multiple_of(scale_height * height)
+ new_width = self.constrain_to_multiple_of(scale_width * width)
+ else:
+ raise ValueError(f"resize_method {self.__resize_method} not implemented")
+
+ return (new_width, new_height)
+
+ def __call__(self, sample):
+ width, height = self.get_size(
+ sample["image"].shape[1], sample["image"].shape[0]
+ )
+
+ # resize sample
+ sample["image"] = cv2.resize(
+ sample["image"],
+ (width, height),
+ interpolation=self.__image_interpolation_method,
+ )
+
+ if self.__resize_target:
+ if "disparity" in sample:
+ sample["disparity"] = cv2.resize(
+ sample["disparity"],
+ (width, height),
+ interpolation=cv2.INTER_NEAREST,
+ )
+
+ if "depth" in sample:
+ sample["depth"] = cv2.resize(
+ sample["depth"], (width, height), interpolation=cv2.INTER_NEAREST
+ )
+
+ sample["mask"] = cv2.resize(
+ sample["mask"].astype(np.float32),
+ (width, height),
+ interpolation=cv2.INTER_NEAREST,
+ )
+ sample["mask"] = sample["mask"].astype(bool)
+
+ return sample
+
+
+class NormalizeImage(object):
+ """Normlize image by given mean and std.
+ """
+
+ def __init__(self, mean, std):
+ self.__mean = mean
+ self.__std = std
+
+ def __call__(self, sample):
+ sample["image"] = (sample["image"] - self.__mean) / self.__std
+
+ return sample
+
+
+class PrepareForNet(object):
+ """Prepare sample for usage as network input.
+ """
+
+ def __init__(self):
+ pass
+
+ def __call__(self, sample):
+ image = np.transpose(sample["image"], (2, 0, 1))
+ sample["image"] = np.ascontiguousarray(image).astype(np.float32)
+
+ if "mask" in sample:
+ sample["mask"] = sample["mask"].astype(np.float32)
+ sample["mask"] = np.ascontiguousarray(sample["mask"])
+
+ if "disparity" in sample:
+ disparity = sample["disparity"].astype(np.float32)
+ sample["disparity"] = np.ascontiguousarray(disparity)
+
+ if "depth" in sample:
+ depth = sample["depth"].astype(np.float32)
+ sample["depth"] = np.ascontiguousarray(depth)
+
+ return sample
diff --git a/ControlNet-v1-1-nightly-main/annotator/midas/midas/vit.py b/ControlNet-v1-1-nightly-main/annotator/midas/midas/vit.py
new file mode 100644
index 0000000000000000000000000000000000000000..ea46b1be88b261b0dec04f3da0256f5f66f88a74
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/midas/midas/vit.py
@@ -0,0 +1,491 @@
+import torch
+import torch.nn as nn
+import timm
+import types
+import math
+import torch.nn.functional as F
+
+
+class Slice(nn.Module):
+ def __init__(self, start_index=1):
+ super(Slice, self).__init__()
+ self.start_index = start_index
+
+ def forward(self, x):
+ return x[:, self.start_index :]
+
+
+class AddReadout(nn.Module):
+ def __init__(self, start_index=1):
+ super(AddReadout, self).__init__()
+ self.start_index = start_index
+
+ def forward(self, x):
+ if self.start_index == 2:
+ readout = (x[:, 0] + x[:, 1]) / 2
+ else:
+ readout = x[:, 0]
+ return x[:, self.start_index :] + readout.unsqueeze(1)
+
+
+class ProjectReadout(nn.Module):
+ def __init__(self, in_features, start_index=1):
+ super(ProjectReadout, self).__init__()
+ self.start_index = start_index
+
+ self.project = nn.Sequential(nn.Linear(2 * in_features, in_features), nn.GELU())
+
+ def forward(self, x):
+ readout = x[:, 0].unsqueeze(1).expand_as(x[:, self.start_index :])
+ features = torch.cat((x[:, self.start_index :], readout), -1)
+
+ return self.project(features)
+
+
+class Transpose(nn.Module):
+ def __init__(self, dim0, dim1):
+ super(Transpose, self).__init__()
+ self.dim0 = dim0
+ self.dim1 = dim1
+
+ def forward(self, x):
+ x = x.transpose(self.dim0, self.dim1)
+ return x
+
+
+def forward_vit(pretrained, x):
+ b, c, h, w = x.shape
+
+ glob = pretrained.model.forward_flex(x)
+
+ layer_1 = pretrained.activations["1"]
+ layer_2 = pretrained.activations["2"]
+ layer_3 = pretrained.activations["3"]
+ layer_4 = pretrained.activations["4"]
+
+ layer_1 = pretrained.act_postprocess1[0:2](layer_1)
+ layer_2 = pretrained.act_postprocess2[0:2](layer_2)
+ layer_3 = pretrained.act_postprocess3[0:2](layer_3)
+ layer_4 = pretrained.act_postprocess4[0:2](layer_4)
+
+ unflatten = nn.Sequential(
+ nn.Unflatten(
+ 2,
+ torch.Size(
+ [
+ h // pretrained.model.patch_size[1],
+ w // pretrained.model.patch_size[0],
+ ]
+ ),
+ )
+ )
+
+ if layer_1.ndim == 3:
+ layer_1 = unflatten(layer_1)
+ if layer_2.ndim == 3:
+ layer_2 = unflatten(layer_2)
+ if layer_3.ndim == 3:
+ layer_3 = unflatten(layer_3)
+ if layer_4.ndim == 3:
+ layer_4 = unflatten(layer_4)
+
+ layer_1 = pretrained.act_postprocess1[3 : len(pretrained.act_postprocess1)](layer_1)
+ layer_2 = pretrained.act_postprocess2[3 : len(pretrained.act_postprocess2)](layer_2)
+ layer_3 = pretrained.act_postprocess3[3 : len(pretrained.act_postprocess3)](layer_3)
+ layer_4 = pretrained.act_postprocess4[3 : len(pretrained.act_postprocess4)](layer_4)
+
+ return layer_1, layer_2, layer_3, layer_4
+
+
+def _resize_pos_embed(self, posemb, gs_h, gs_w):
+ posemb_tok, posemb_grid = (
+ posemb[:, : self.start_index],
+ posemb[0, self.start_index :],
+ )
+
+ gs_old = int(math.sqrt(len(posemb_grid)))
+
+ posemb_grid = posemb_grid.reshape(1, gs_old, gs_old, -1).permute(0, 3, 1, 2)
+ posemb_grid = F.interpolate(posemb_grid, size=(gs_h, gs_w), mode="bilinear")
+ posemb_grid = posemb_grid.permute(0, 2, 3, 1).reshape(1, gs_h * gs_w, -1)
+
+ posemb = torch.cat([posemb_tok, posemb_grid], dim=1)
+
+ return posemb
+
+
+def forward_flex(self, x):
+ b, c, h, w = x.shape
+
+ pos_embed = self._resize_pos_embed(
+ self.pos_embed, h // self.patch_size[1], w // self.patch_size[0]
+ )
+
+ B = x.shape[0]
+
+ if hasattr(self.patch_embed, "backbone"):
+ x = self.patch_embed.backbone(x)
+ if isinstance(x, (list, tuple)):
+ x = x[-1] # last feature if backbone outputs list/tuple of features
+
+ x = self.patch_embed.proj(x).flatten(2).transpose(1, 2)
+
+ if getattr(self, "dist_token", None) is not None:
+ cls_tokens = self.cls_token.expand(
+ B, -1, -1
+ ) # stole cls_tokens impl from Phil Wang, thanks
+ dist_token = self.dist_token.expand(B, -1, -1)
+ x = torch.cat((cls_tokens, dist_token, x), dim=1)
+ else:
+ cls_tokens = self.cls_token.expand(
+ B, -1, -1
+ ) # stole cls_tokens impl from Phil Wang, thanks
+ x = torch.cat((cls_tokens, x), dim=1)
+
+ x = x + pos_embed
+ x = self.pos_drop(x)
+
+ for blk in self.blocks:
+ x = blk(x)
+
+ x = self.norm(x)
+
+ return x
+
+
+activations = {}
+
+
+def get_activation(name):
+ def hook(model, input, output):
+ activations[name] = output
+
+ return hook
+
+
+def get_readout_oper(vit_features, features, use_readout, start_index=1):
+ if use_readout == "ignore":
+ readout_oper = [Slice(start_index)] * len(features)
+ elif use_readout == "add":
+ readout_oper = [AddReadout(start_index)] * len(features)
+ elif use_readout == "project":
+ readout_oper = [
+ ProjectReadout(vit_features, start_index) for out_feat in features
+ ]
+ else:
+ assert (
+ False
+ ), "wrong operation for readout token, use_readout can be 'ignore', 'add', or 'project'"
+
+ return readout_oper
+
+
+def _make_vit_b16_backbone(
+ model,
+ features=[96, 192, 384, 768],
+ size=[384, 384],
+ hooks=[2, 5, 8, 11],
+ vit_features=768,
+ use_readout="ignore",
+ start_index=1,
+):
+ pretrained = nn.Module()
+
+ pretrained.model = model
+ pretrained.model.blocks[hooks[0]].register_forward_hook(get_activation("1"))
+ pretrained.model.blocks[hooks[1]].register_forward_hook(get_activation("2"))
+ pretrained.model.blocks[hooks[2]].register_forward_hook(get_activation("3"))
+ pretrained.model.blocks[hooks[3]].register_forward_hook(get_activation("4"))
+
+ pretrained.activations = activations
+
+ readout_oper = get_readout_oper(vit_features, features, use_readout, start_index)
+
+ # 32, 48, 136, 384
+ pretrained.act_postprocess1 = nn.Sequential(
+ readout_oper[0],
+ Transpose(1, 2),
+ nn.Unflatten(2, torch.Size([size[0] // 16, size[1] // 16])),
+ nn.Conv2d(
+ in_channels=vit_features,
+ out_channels=features[0],
+ kernel_size=1,
+ stride=1,
+ padding=0,
+ ),
+ nn.ConvTranspose2d(
+ in_channels=features[0],
+ out_channels=features[0],
+ kernel_size=4,
+ stride=4,
+ padding=0,
+ bias=True,
+ dilation=1,
+ groups=1,
+ ),
+ )
+
+ pretrained.act_postprocess2 = nn.Sequential(
+ readout_oper[1],
+ Transpose(1, 2),
+ nn.Unflatten(2, torch.Size([size[0] // 16, size[1] // 16])),
+ nn.Conv2d(
+ in_channels=vit_features,
+ out_channels=features[1],
+ kernel_size=1,
+ stride=1,
+ padding=0,
+ ),
+ nn.ConvTranspose2d(
+ in_channels=features[1],
+ out_channels=features[1],
+ kernel_size=2,
+ stride=2,
+ padding=0,
+ bias=True,
+ dilation=1,
+ groups=1,
+ ),
+ )
+
+ pretrained.act_postprocess3 = nn.Sequential(
+ readout_oper[2],
+ Transpose(1, 2),
+ nn.Unflatten(2, torch.Size([size[0] // 16, size[1] // 16])),
+ nn.Conv2d(
+ in_channels=vit_features,
+ out_channels=features[2],
+ kernel_size=1,
+ stride=1,
+ padding=0,
+ ),
+ )
+
+ pretrained.act_postprocess4 = nn.Sequential(
+ readout_oper[3],
+ Transpose(1, 2),
+ nn.Unflatten(2, torch.Size([size[0] // 16, size[1] // 16])),
+ nn.Conv2d(
+ in_channels=vit_features,
+ out_channels=features[3],
+ kernel_size=1,
+ stride=1,
+ padding=0,
+ ),
+ nn.Conv2d(
+ in_channels=features[3],
+ out_channels=features[3],
+ kernel_size=3,
+ stride=2,
+ padding=1,
+ ),
+ )
+
+ pretrained.model.start_index = start_index
+ pretrained.model.patch_size = [16, 16]
+
+ # We inject this function into the VisionTransformer instances so that
+ # we can use it with interpolated position embeddings without modifying the library source.
+ pretrained.model.forward_flex = types.MethodType(forward_flex, pretrained.model)
+ pretrained.model._resize_pos_embed = types.MethodType(
+ _resize_pos_embed, pretrained.model
+ )
+
+ return pretrained
+
+
+def _make_pretrained_vitl16_384(pretrained, use_readout="ignore", hooks=None):
+ model = timm.create_model("vit_large_patch16_384", pretrained=pretrained)
+
+ hooks = [5, 11, 17, 23] if hooks == None else hooks
+ return _make_vit_b16_backbone(
+ model,
+ features=[256, 512, 1024, 1024],
+ hooks=hooks,
+ vit_features=1024,
+ use_readout=use_readout,
+ )
+
+
+def _make_pretrained_vitb16_384(pretrained, use_readout="ignore", hooks=None):
+ model = timm.create_model("vit_base_patch16_384", pretrained=pretrained)
+
+ hooks = [2, 5, 8, 11] if hooks == None else hooks
+ return _make_vit_b16_backbone(
+ model, features=[96, 192, 384, 768], hooks=hooks, use_readout=use_readout
+ )
+
+
+def _make_pretrained_deitb16_384(pretrained, use_readout="ignore", hooks=None):
+ model = timm.create_model("vit_deit_base_patch16_384", pretrained=pretrained)
+
+ hooks = [2, 5, 8, 11] if hooks == None else hooks
+ return _make_vit_b16_backbone(
+ model, features=[96, 192, 384, 768], hooks=hooks, use_readout=use_readout
+ )
+
+
+def _make_pretrained_deitb16_distil_384(pretrained, use_readout="ignore", hooks=None):
+ model = timm.create_model(
+ "vit_deit_base_distilled_patch16_384", pretrained=pretrained
+ )
+
+ hooks = [2, 5, 8, 11] if hooks == None else hooks
+ return _make_vit_b16_backbone(
+ model,
+ features=[96, 192, 384, 768],
+ hooks=hooks,
+ use_readout=use_readout,
+ start_index=2,
+ )
+
+
+def _make_vit_b_rn50_backbone(
+ model,
+ features=[256, 512, 768, 768],
+ size=[384, 384],
+ hooks=[0, 1, 8, 11],
+ vit_features=768,
+ use_vit_only=False,
+ use_readout="ignore",
+ start_index=1,
+):
+ pretrained = nn.Module()
+
+ pretrained.model = model
+
+ if use_vit_only == True:
+ pretrained.model.blocks[hooks[0]].register_forward_hook(get_activation("1"))
+ pretrained.model.blocks[hooks[1]].register_forward_hook(get_activation("2"))
+ else:
+ pretrained.model.patch_embed.backbone.stages[0].register_forward_hook(
+ get_activation("1")
+ )
+ pretrained.model.patch_embed.backbone.stages[1].register_forward_hook(
+ get_activation("2")
+ )
+
+ pretrained.model.blocks[hooks[2]].register_forward_hook(get_activation("3"))
+ pretrained.model.blocks[hooks[3]].register_forward_hook(get_activation("4"))
+
+ pretrained.activations = activations
+
+ readout_oper = get_readout_oper(vit_features, features, use_readout, start_index)
+
+ if use_vit_only == True:
+ pretrained.act_postprocess1 = nn.Sequential(
+ readout_oper[0],
+ Transpose(1, 2),
+ nn.Unflatten(2, torch.Size([size[0] // 16, size[1] // 16])),
+ nn.Conv2d(
+ in_channels=vit_features,
+ out_channels=features[0],
+ kernel_size=1,
+ stride=1,
+ padding=0,
+ ),
+ nn.ConvTranspose2d(
+ in_channels=features[0],
+ out_channels=features[0],
+ kernel_size=4,
+ stride=4,
+ padding=0,
+ bias=True,
+ dilation=1,
+ groups=1,
+ ),
+ )
+
+ pretrained.act_postprocess2 = nn.Sequential(
+ readout_oper[1],
+ Transpose(1, 2),
+ nn.Unflatten(2, torch.Size([size[0] // 16, size[1] // 16])),
+ nn.Conv2d(
+ in_channels=vit_features,
+ out_channels=features[1],
+ kernel_size=1,
+ stride=1,
+ padding=0,
+ ),
+ nn.ConvTranspose2d(
+ in_channels=features[1],
+ out_channels=features[1],
+ kernel_size=2,
+ stride=2,
+ padding=0,
+ bias=True,
+ dilation=1,
+ groups=1,
+ ),
+ )
+ else:
+ pretrained.act_postprocess1 = nn.Sequential(
+ nn.Identity(), nn.Identity(), nn.Identity()
+ )
+ pretrained.act_postprocess2 = nn.Sequential(
+ nn.Identity(), nn.Identity(), nn.Identity()
+ )
+
+ pretrained.act_postprocess3 = nn.Sequential(
+ readout_oper[2],
+ Transpose(1, 2),
+ nn.Unflatten(2, torch.Size([size[0] // 16, size[1] // 16])),
+ nn.Conv2d(
+ in_channels=vit_features,
+ out_channels=features[2],
+ kernel_size=1,
+ stride=1,
+ padding=0,
+ ),
+ )
+
+ pretrained.act_postprocess4 = nn.Sequential(
+ readout_oper[3],
+ Transpose(1, 2),
+ nn.Unflatten(2, torch.Size([size[0] // 16, size[1] // 16])),
+ nn.Conv2d(
+ in_channels=vit_features,
+ out_channels=features[3],
+ kernel_size=1,
+ stride=1,
+ padding=0,
+ ),
+ nn.Conv2d(
+ in_channels=features[3],
+ out_channels=features[3],
+ kernel_size=3,
+ stride=2,
+ padding=1,
+ ),
+ )
+
+ pretrained.model.start_index = start_index
+ pretrained.model.patch_size = [16, 16]
+
+ # We inject this function into the VisionTransformer instances so that
+ # we can use it with interpolated position embeddings without modifying the library source.
+ pretrained.model.forward_flex = types.MethodType(forward_flex, pretrained.model)
+
+ # We inject this function into the VisionTransformer instances so that
+ # we can use it with interpolated position embeddings without modifying the library source.
+ pretrained.model._resize_pos_embed = types.MethodType(
+ _resize_pos_embed, pretrained.model
+ )
+
+ return pretrained
+
+
+def _make_pretrained_vitb_rn50_384(
+ pretrained, use_readout="ignore", hooks=None, use_vit_only=False
+):
+ model = timm.create_model("vit_base_resnet50_384", pretrained=pretrained)
+
+ hooks = [0, 1, 8, 11] if hooks == None else hooks
+ return _make_vit_b_rn50_backbone(
+ model,
+ features=[256, 512, 768, 768],
+ size=[384, 384],
+ hooks=hooks,
+ use_vit_only=use_vit_only,
+ use_readout=use_readout,
+ )
diff --git a/ControlNet-v1-1-nightly-main/annotator/midas/utils.py b/ControlNet-v1-1-nightly-main/annotator/midas/utils.py
new file mode 100644
index 0000000000000000000000000000000000000000..9a9d3b5b66370fa98da9e067ba53ead848ea9a59
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/midas/utils.py
@@ -0,0 +1,189 @@
+"""Utils for monoDepth."""
+import sys
+import re
+import numpy as np
+import cv2
+import torch
+
+
+def read_pfm(path):
+ """Read pfm file.
+
+ Args:
+ path (str): path to file
+
+ Returns:
+ tuple: (data, scale)
+ """
+ with open(path, "rb") as file:
+
+ color = None
+ width = None
+ height = None
+ scale = None
+ endian = None
+
+ header = file.readline().rstrip()
+ if header.decode("ascii") == "PF":
+ color = True
+ elif header.decode("ascii") == "Pf":
+ color = False
+ else:
+ raise Exception("Not a PFM file: " + path)
+
+ dim_match = re.match(r"^(\d+)\s(\d+)\s$", file.readline().decode("ascii"))
+ if dim_match:
+ width, height = list(map(int, dim_match.groups()))
+ else:
+ raise Exception("Malformed PFM header.")
+
+ scale = float(file.readline().decode("ascii").rstrip())
+ if scale < 0:
+ # little-endian
+ endian = "<"
+ scale = -scale
+ else:
+ # big-endian
+ endian = ">"
+
+ data = np.fromfile(file, endian + "f")
+ shape = (height, width, 3) if color else (height, width)
+
+ data = np.reshape(data, shape)
+ data = np.flipud(data)
+
+ return data, scale
+
+
+def write_pfm(path, image, scale=1):
+ """Write pfm file.
+
+ Args:
+ path (str): pathto file
+ image (array): data
+ scale (int, optional): Scale. Defaults to 1.
+ """
+
+ with open(path, "wb") as file:
+ color = None
+
+ if image.dtype.name != "float32":
+ raise Exception("Image dtype must be float32.")
+
+ image = np.flipud(image)
+
+ if len(image.shape) == 3 and image.shape[2] == 3: # color image
+ color = True
+ elif (
+ len(image.shape) == 2 or len(image.shape) == 3 and image.shape[2] == 1
+ ): # greyscale
+ color = False
+ else:
+ raise Exception("Image must have H x W x 3, H x W x 1 or H x W dimensions.")
+
+ file.write("PF\n" if color else "Pf\n".encode())
+ file.write("%d %d\n".encode() % (image.shape[1], image.shape[0]))
+
+ endian = image.dtype.byteorder
+
+ if endian == "<" or endian == "=" and sys.byteorder == "little":
+ scale = -scale
+
+ file.write("%f\n".encode() % scale)
+
+ image.tofile(file)
+
+
+def read_image(path):
+ """Read image and output RGB image (0-1).
+
+ Args:
+ path (str): path to file
+
+ Returns:
+ array: RGB image (0-1)
+ """
+ img = cv2.imread(path)
+
+ if img.ndim == 2:
+ img = cv2.cvtColor(img, cv2.COLOR_GRAY2BGR)
+
+ img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB) / 255.0
+
+ return img
+
+
+def resize_image(img):
+ """Resize image and make it fit for network.
+
+ Args:
+ img (array): image
+
+ Returns:
+ tensor: data ready for network
+ """
+ height_orig = img.shape[0]
+ width_orig = img.shape[1]
+
+ if width_orig > height_orig:
+ scale = width_orig / 384
+ else:
+ scale = height_orig / 384
+
+ height = (np.ceil(height_orig / scale / 32) * 32).astype(int)
+ width = (np.ceil(width_orig / scale / 32) * 32).astype(int)
+
+ img_resized = cv2.resize(img, (width, height), interpolation=cv2.INTER_AREA)
+
+ img_resized = (
+ torch.from_numpy(np.transpose(img_resized, (2, 0, 1))).contiguous().float()
+ )
+ img_resized = img_resized.unsqueeze(0)
+
+ return img_resized
+
+
+def resize_depth(depth, width, height):
+ """Resize depth map and bring to CPU (numpy).
+
+ Args:
+ depth (tensor): depth
+ width (int): image width
+ height (int): image height
+
+ Returns:
+ array: processed depth
+ """
+ depth = torch.squeeze(depth[0, :, :, :]).to("cpu")
+
+ depth_resized = cv2.resize(
+ depth.numpy(), (width, height), interpolation=cv2.INTER_CUBIC
+ )
+
+ return depth_resized
+
+def write_depth(path, depth, bits=1):
+ """Write depth map to pfm and png file.
+
+ Args:
+ path (str): filepath without extension
+ depth (array): depth
+ """
+ write_pfm(path + ".pfm", depth.astype(np.float32))
+
+ depth_min = depth.min()
+ depth_max = depth.max()
+
+ max_val = (2**(8*bits))-1
+
+ if depth_max - depth_min > np.finfo("float").eps:
+ out = max_val * (depth - depth_min) / (depth_max - depth_min)
+ else:
+ out = np.zeros(depth.shape, dtype=depth.type)
+
+ if bits == 1:
+ cv2.imwrite(path + ".png", out.astype("uint8"))
+ elif bits == 2:
+ cv2.imwrite(path + ".png", out.astype("uint16"))
+
+ return
diff --git a/ControlNet-v1-1-nightly-main/annotator/mlsd/LICENSE b/ControlNet-v1-1-nightly-main/annotator/mlsd/LICENSE
new file mode 100644
index 0000000000000000000000000000000000000000..d855c6db44b4e873eedd750d34fa2eaf22e22363
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/mlsd/LICENSE
@@ -0,0 +1,201 @@
+ Apache License
+ Version 2.0, January 2004
+ http://www.apache.org/licenses/
+
+ TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+ 1. Definitions.
+
+ "License" shall mean the terms and conditions for use, reproduction,
+ and distribution as defined by Sections 1 through 9 of this document.
+
+ "Licensor" shall mean the copyright owner or entity authorized by
+ the copyright owner that is granting the License.
+
+ "Legal Entity" shall mean the union of the acting entity and all
+ other entities that control, are controlled by, or are under common
+ control with that entity. For the purposes of this definition,
+ "control" means (i) the power, direct or indirect, to cause the
+ direction or management of such entity, whether by contract or
+ otherwise, or (ii) ownership of fifty percent (50%) or more of the
+ outstanding shares, or (iii) beneficial ownership of such entity.
+
+ "You" (or "Your") shall mean an individual or Legal Entity
+ exercising permissions granted by this License.
+
+ "Source" form shall mean the preferred form for making modifications,
+ including but not limited to software source code, documentation
+ source, and configuration files.
+
+ "Object" form shall mean any form resulting from mechanical
+ transformation or translation of a Source form, including but
+ not limited to compiled object code, generated documentation,
+ and conversions to other media types.
+
+ "Work" shall mean the work of authorship, whether in Source or
+ Object form, made available under the License, as indicated by a
+ copyright notice that is included in or attached to the work
+ (an example is provided in the Appendix below).
+
+ "Derivative Works" shall mean any work, whether in Source or Object
+ form, that is based on (or derived from) the Work and for which the
+ editorial revisions, annotations, elaborations, or other modifications
+ represent, as a whole, an original work of authorship. For the purposes
+ of this License, Derivative Works shall not include works that remain
+ separable from, or merely link (or bind by name) to the interfaces of,
+ the Work and Derivative Works thereof.
+
+ "Contribution" shall mean any work of authorship, including
+ the original version of the Work and any modifications or additions
+ to that Work or Derivative Works thereof, that is intentionally
+ submitted to Licensor for inclusion in the Work by the copyright owner
+ or by an individual or Legal Entity authorized to submit on behalf of
+ the copyright owner. For the purposes of this definition, "submitted"
+ means any form of electronic, verbal, or written communication sent
+ to the Licensor or its representatives, including but not limited to
+ communication on electronic mailing lists, source code control systems,
+ and issue tracking systems that are managed by, or on behalf of, the
+ Licensor for the purpose of discussing and improving the Work, but
+ excluding communication that is conspicuously marked or otherwise
+ designated in writing by the copyright owner as "Not a Contribution."
+
+ "Contributor" shall mean Licensor and any individual or Legal Entity
+ on behalf of whom a Contribution has been received by Licensor and
+ subsequently incorporated within the Work.
+
+ 2. Grant of Copyright License. Subject to the terms and conditions of
+ this License, each Contributor hereby grants to You a perpetual,
+ worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+ copyright license to reproduce, prepare Derivative Works of,
+ publicly display, publicly perform, sublicense, and distribute the
+ Work and such Derivative Works in Source or Object form.
+
+ 3. Grant of Patent License. Subject to the terms and conditions of
+ this License, each Contributor hereby grants to You a perpetual,
+ worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+ (except as stated in this section) patent license to make, have made,
+ use, offer to sell, sell, import, and otherwise transfer the Work,
+ where such license applies only to those patent claims licensable
+ by such Contributor that are necessarily infringed by their
+ Contribution(s) alone or by combination of their Contribution(s)
+ with the Work to which such Contribution(s) was submitted. If You
+ institute patent litigation against any entity (including a
+ cross-claim or counterclaim in a lawsuit) alleging that the Work
+ or a Contribution incorporated within the Work constitutes direct
+ or contributory patent infringement, then any patent licenses
+ granted to You under this License for that Work shall terminate
+ as of the date such litigation is filed.
+
+ 4. Redistribution. You may reproduce and distribute copies of the
+ Work or Derivative Works thereof in any medium, with or without
+ modifications, and in Source or Object form, provided that You
+ meet the following conditions:
+
+ (a) You must give any other recipients of the Work or
+ Derivative Works a copy of this License; and
+
+ (b) You must cause any modified files to carry prominent notices
+ stating that You changed the files; and
+
+ (c) You must retain, in the Source form of any Derivative Works
+ that You distribute, all copyright, patent, trademark, and
+ attribution notices from the Source form of the Work,
+ excluding those notices that do not pertain to any part of
+ the Derivative Works; and
+
+ (d) If the Work includes a "NOTICE" text file as part of its
+ distribution, then any Derivative Works that You distribute must
+ include a readable copy of the attribution notices contained
+ within such NOTICE file, excluding those notices that do not
+ pertain to any part of the Derivative Works, in at least one
+ of the following places: within a NOTICE text file distributed
+ as part of the Derivative Works; within the Source form or
+ documentation, if provided along with the Derivative Works; or,
+ within a display generated by the Derivative Works, if and
+ wherever such third-party notices normally appear. The contents
+ of the NOTICE file are for informational purposes only and
+ do not modify the License. You may add Your own attribution
+ notices within Derivative Works that You distribute, alongside
+ or as an addendum to the NOTICE text from the Work, provided
+ that such additional attribution notices cannot be construed
+ as modifying the License.
+
+ You may add Your own copyright statement to Your modifications and
+ may provide additional or different license terms and conditions
+ for use, reproduction, or distribution of Your modifications, or
+ for any such Derivative Works as a whole, provided Your use,
+ reproduction, and distribution of the Work otherwise complies with
+ the conditions stated in this License.
+
+ 5. Submission of Contributions. Unless You explicitly state otherwise,
+ any Contribution intentionally submitted for inclusion in the Work
+ by You to the Licensor shall be under the terms and conditions of
+ this License, without any additional terms or conditions.
+ Notwithstanding the above, nothing herein shall supersede or modify
+ the terms of any separate license agreement you may have executed
+ with Licensor regarding such Contributions.
+
+ 6. Trademarks. This License does not grant permission to use the trade
+ names, trademarks, service marks, or product names of the Licensor,
+ except as required for reasonable and customary use in describing the
+ origin of the Work and reproducing the content of the NOTICE file.
+
+ 7. Disclaimer of Warranty. Unless required by applicable law or
+ agreed to in writing, Licensor provides the Work (and each
+ Contributor provides its Contributions) on an "AS IS" BASIS,
+ WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+ implied, including, without limitation, any warranties or conditions
+ of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+ PARTICULAR PURPOSE. You are solely responsible for determining the
+ appropriateness of using or redistributing the Work and assume any
+ risks associated with Your exercise of permissions under this License.
+
+ 8. Limitation of Liability. In no event and under no legal theory,
+ whether in tort (including negligence), contract, or otherwise,
+ unless required by applicable law (such as deliberate and grossly
+ negligent acts) or agreed to in writing, shall any Contributor be
+ liable to You for damages, including any direct, indirect, special,
+ incidental, or consequential damages of any character arising as a
+ result of this License or out of the use or inability to use the
+ Work (including but not limited to damages for loss of goodwill,
+ work stoppage, computer failure or malfunction, or any and all
+ other commercial damages or losses), even if such Contributor
+ has been advised of the possibility of such damages.
+
+ 9. Accepting Warranty or Additional Liability. While redistributing
+ the Work or Derivative Works thereof, You may choose to offer,
+ and charge a fee for, acceptance of support, warranty, indemnity,
+ or other liability obligations and/or rights consistent with this
+ License. However, in accepting such obligations, You may act only
+ on Your own behalf and on Your sole responsibility, not on behalf
+ of any other Contributor, and only if You agree to indemnify,
+ defend, and hold each Contributor harmless for any liability
+ incurred by, or claims asserted against, such Contributor by reason
+ of your accepting any such warranty or additional liability.
+
+ END OF TERMS AND CONDITIONS
+
+ APPENDIX: How to apply the Apache License to your work.
+
+ To apply the Apache License to your work, attach the following
+ boilerplate notice, with the fields enclosed by brackets "{}"
+ replaced with your own identifying information. (Don't include
+ the brackets!) The text should be enclosed in the appropriate
+ comment syntax for the file format. We also recommend that a
+ file or class name and description of purpose be included on the
+ same "printed page" as the copyright notice for easier
+ identification within third-party archives.
+
+ Copyright 2021-present NAVER Corp.
+
+ Licensed under the Apache License, Version 2.0 (the "License");
+ you may not use this file except in compliance with the License.
+ You may obtain a copy of the License at
+
+ http://www.apache.org/licenses/LICENSE-2.0
+
+ Unless required by applicable law or agreed to in writing, software
+ distributed under the License is distributed on an "AS IS" BASIS,
+ WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ See the License for the specific language governing permissions and
+ limitations under the License.
\ No newline at end of file
diff --git a/ControlNet-v1-1-nightly-main/annotator/mlsd/__init__.py b/ControlNet-v1-1-nightly-main/annotator/mlsd/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..81ea5482fe3de54a7bb727332cc6e875a705a9f7
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/mlsd/__init__.py
@@ -0,0 +1,43 @@
+# MLSD Line Detection
+# From https://github.com/navervision/mlsd
+# Apache-2.0 license
+
+import cv2
+import numpy as np
+import torch
+import os
+
+from einops import rearrange
+from .models.mbv2_mlsd_tiny import MobileV2_MLSD_Tiny
+from .models.mbv2_mlsd_large import MobileV2_MLSD_Large
+from .utils import pred_lines
+
+from annotator.util import annotator_ckpts_path
+
+
+remote_model_path = "https://huggingface.co/lllyasviel/Annotators/resolve/main/mlsd_large_512_fp32.pth"
+
+
+class MLSDdetector:
+ def __init__(self):
+ model_path = os.path.join(annotator_ckpts_path, "mlsd_large_512_fp32.pth")
+ if not os.path.exists(model_path):
+ from basicsr.utils.download_util import load_file_from_url
+ load_file_from_url(remote_model_path, model_dir=annotator_ckpts_path)
+ model = MobileV2_MLSD_Large()
+ model.load_state_dict(torch.load(model_path), strict=True)
+ self.model = model.cuda().eval()
+
+ def __call__(self, input_image, thr_v, thr_d):
+ assert input_image.ndim == 3
+ img = input_image
+ img_output = np.zeros_like(img)
+ try:
+ with torch.no_grad():
+ lines = pred_lines(img, self.model, [img.shape[0], img.shape[1]], thr_v, thr_d)
+ for line in lines:
+ x_start, y_start, x_end, y_end = [int(val) for val in line]
+ cv2.line(img_output, (x_start, y_start), (x_end, y_end), [255, 255, 255], 1)
+ except Exception as e:
+ pass
+ return img_output[:, :, 0]
diff --git a/ControlNet-v1-1-nightly-main/annotator/mlsd/models/mbv2_mlsd_large.py b/ControlNet-v1-1-nightly-main/annotator/mlsd/models/mbv2_mlsd_large.py
new file mode 100644
index 0000000000000000000000000000000000000000..5b9799e7573ca41549b3c3b13ac47b906b369603
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/mlsd/models/mbv2_mlsd_large.py
@@ -0,0 +1,292 @@
+import os
+import sys
+import torch
+import torch.nn as nn
+import torch.utils.model_zoo as model_zoo
+from torch.nn import functional as F
+
+
+class BlockTypeA(nn.Module):
+ def __init__(self, in_c1, in_c2, out_c1, out_c2, upscale = True):
+ super(BlockTypeA, self).__init__()
+ self.conv1 = nn.Sequential(
+ nn.Conv2d(in_c2, out_c2, kernel_size=1),
+ nn.BatchNorm2d(out_c2),
+ nn.ReLU(inplace=True)
+ )
+ self.conv2 = nn.Sequential(
+ nn.Conv2d(in_c1, out_c1, kernel_size=1),
+ nn.BatchNorm2d(out_c1),
+ nn.ReLU(inplace=True)
+ )
+ self.upscale = upscale
+
+ def forward(self, a, b):
+ b = self.conv1(b)
+ a = self.conv2(a)
+ if self.upscale:
+ b = F.interpolate(b, scale_factor=2.0, mode='bilinear', align_corners=True)
+ return torch.cat((a, b), dim=1)
+
+
+class BlockTypeB(nn.Module):
+ def __init__(self, in_c, out_c):
+ super(BlockTypeB, self).__init__()
+ self.conv1 = nn.Sequential(
+ nn.Conv2d(in_c, in_c, kernel_size=3, padding=1),
+ nn.BatchNorm2d(in_c),
+ nn.ReLU()
+ )
+ self.conv2 = nn.Sequential(
+ nn.Conv2d(in_c, out_c, kernel_size=3, padding=1),
+ nn.BatchNorm2d(out_c),
+ nn.ReLU()
+ )
+
+ def forward(self, x):
+ x = self.conv1(x) + x
+ x = self.conv2(x)
+ return x
+
+class BlockTypeC(nn.Module):
+ def __init__(self, in_c, out_c):
+ super(BlockTypeC, self).__init__()
+ self.conv1 = nn.Sequential(
+ nn.Conv2d(in_c, in_c, kernel_size=3, padding=5, dilation=5),
+ nn.BatchNorm2d(in_c),
+ nn.ReLU()
+ )
+ self.conv2 = nn.Sequential(
+ nn.Conv2d(in_c, in_c, kernel_size=3, padding=1),
+ nn.BatchNorm2d(in_c),
+ nn.ReLU()
+ )
+ self.conv3 = nn.Conv2d(in_c, out_c, kernel_size=1)
+
+ def forward(self, x):
+ x = self.conv1(x)
+ x = self.conv2(x)
+ x = self.conv3(x)
+ return x
+
+def _make_divisible(v, divisor, min_value=None):
+ """
+ This function is taken from the original tf repo.
+ It ensures that all layers have a channel number that is divisible by 8
+ It can be seen here:
+ https://github.com/tensorflow/models/blob/master/research/slim/nets/mobilenet/mobilenet.py
+ :param v:
+ :param divisor:
+ :param min_value:
+ :return:
+ """
+ if min_value is None:
+ min_value = divisor
+ new_v = max(min_value, int(v + divisor / 2) // divisor * divisor)
+ # Make sure that round down does not go down by more than 10%.
+ if new_v < 0.9 * v:
+ new_v += divisor
+ return new_v
+
+
+class ConvBNReLU(nn.Sequential):
+ def __init__(self, in_planes, out_planes, kernel_size=3, stride=1, groups=1):
+ self.channel_pad = out_planes - in_planes
+ self.stride = stride
+ #padding = (kernel_size - 1) // 2
+
+ # TFLite uses slightly different padding than PyTorch
+ if stride == 2:
+ padding = 0
+ else:
+ padding = (kernel_size - 1) // 2
+
+ super(ConvBNReLU, self).__init__(
+ nn.Conv2d(in_planes, out_planes, kernel_size, stride, padding, groups=groups, bias=False),
+ nn.BatchNorm2d(out_planes),
+ nn.ReLU6(inplace=True)
+ )
+ self.max_pool = nn.MaxPool2d(kernel_size=stride, stride=stride)
+
+
+ def forward(self, x):
+ # TFLite uses different padding
+ if self.stride == 2:
+ x = F.pad(x, (0, 1, 0, 1), "constant", 0)
+ #print(x.shape)
+
+ for module in self:
+ if not isinstance(module, nn.MaxPool2d):
+ x = module(x)
+ return x
+
+
+class InvertedResidual(nn.Module):
+ def __init__(self, inp, oup, stride, expand_ratio):
+ super(InvertedResidual, self).__init__()
+ self.stride = stride
+ assert stride in [1, 2]
+
+ hidden_dim = int(round(inp * expand_ratio))
+ self.use_res_connect = self.stride == 1 and inp == oup
+
+ layers = []
+ if expand_ratio != 1:
+ # pw
+ layers.append(ConvBNReLU(inp, hidden_dim, kernel_size=1))
+ layers.extend([
+ # dw
+ ConvBNReLU(hidden_dim, hidden_dim, stride=stride, groups=hidden_dim),
+ # pw-linear
+ nn.Conv2d(hidden_dim, oup, 1, 1, 0, bias=False),
+ nn.BatchNorm2d(oup),
+ ])
+ self.conv = nn.Sequential(*layers)
+
+ def forward(self, x):
+ if self.use_res_connect:
+ return x + self.conv(x)
+ else:
+ return self.conv(x)
+
+
+class MobileNetV2(nn.Module):
+ def __init__(self, pretrained=True):
+ """
+ MobileNet V2 main class
+ Args:
+ num_classes (int): Number of classes
+ width_mult (float): Width multiplier - adjusts number of channels in each layer by this amount
+ inverted_residual_setting: Network structure
+ round_nearest (int): Round the number of channels in each layer to be a multiple of this number
+ Set to 1 to turn off rounding
+ block: Module specifying inverted residual building block for mobilenet
+ """
+ super(MobileNetV2, self).__init__()
+
+ block = InvertedResidual
+ input_channel = 32
+ last_channel = 1280
+ width_mult = 1.0
+ round_nearest = 8
+
+ inverted_residual_setting = [
+ # t, c, n, s
+ [1, 16, 1, 1],
+ [6, 24, 2, 2],
+ [6, 32, 3, 2],
+ [6, 64, 4, 2],
+ [6, 96, 3, 1],
+ #[6, 160, 3, 2],
+ #[6, 320, 1, 1],
+ ]
+
+ # only check the first element, assuming user knows t,c,n,s are required
+ if len(inverted_residual_setting) == 0 or len(inverted_residual_setting[0]) != 4:
+ raise ValueError("inverted_residual_setting should be non-empty "
+ "or a 4-element list, got {}".format(inverted_residual_setting))
+
+ # building first layer
+ input_channel = _make_divisible(input_channel * width_mult, round_nearest)
+ self.last_channel = _make_divisible(last_channel * max(1.0, width_mult), round_nearest)
+ features = [ConvBNReLU(4, input_channel, stride=2)]
+ # building inverted residual blocks
+ for t, c, n, s in inverted_residual_setting:
+ output_channel = _make_divisible(c * width_mult, round_nearest)
+ for i in range(n):
+ stride = s if i == 0 else 1
+ features.append(block(input_channel, output_channel, stride, expand_ratio=t))
+ input_channel = output_channel
+
+ self.features = nn.Sequential(*features)
+ self.fpn_selected = [1, 3, 6, 10, 13]
+ # weight initialization
+ for m in self.modules():
+ if isinstance(m, nn.Conv2d):
+ nn.init.kaiming_normal_(m.weight, mode='fan_out')
+ if m.bias is not None:
+ nn.init.zeros_(m.bias)
+ elif isinstance(m, nn.BatchNorm2d):
+ nn.init.ones_(m.weight)
+ nn.init.zeros_(m.bias)
+ elif isinstance(m, nn.Linear):
+ nn.init.normal_(m.weight, 0, 0.01)
+ nn.init.zeros_(m.bias)
+ if pretrained:
+ self._load_pretrained_model()
+
+ def _forward_impl(self, x):
+ # This exists since TorchScript doesn't support inheritance, so the superclass method
+ # (this one) needs to have a name other than `forward` that can be accessed in a subclass
+ fpn_features = []
+ for i, f in enumerate(self.features):
+ if i > self.fpn_selected[-1]:
+ break
+ x = f(x)
+ if i in self.fpn_selected:
+ fpn_features.append(x)
+
+ c1, c2, c3, c4, c5 = fpn_features
+ return c1, c2, c3, c4, c5
+
+
+ def forward(self, x):
+ return self._forward_impl(x)
+
+ def _load_pretrained_model(self):
+ pretrain_dict = model_zoo.load_url('https://download.pytorch.org/models/mobilenet_v2-b0353104.pth')
+ model_dict = {}
+ state_dict = self.state_dict()
+ for k, v in pretrain_dict.items():
+ if k in state_dict:
+ model_dict[k] = v
+ state_dict.update(model_dict)
+ self.load_state_dict(state_dict)
+
+
+class MobileV2_MLSD_Large(nn.Module):
+ def __init__(self):
+ super(MobileV2_MLSD_Large, self).__init__()
+
+ self.backbone = MobileNetV2(pretrained=False)
+ ## A, B
+ self.block15 = BlockTypeA(in_c1= 64, in_c2= 96,
+ out_c1= 64, out_c2=64,
+ upscale=False)
+ self.block16 = BlockTypeB(128, 64)
+
+ ## A, B
+ self.block17 = BlockTypeA(in_c1 = 32, in_c2 = 64,
+ out_c1= 64, out_c2= 64)
+ self.block18 = BlockTypeB(128, 64)
+
+ ## A, B
+ self.block19 = BlockTypeA(in_c1=24, in_c2=64,
+ out_c1=64, out_c2=64)
+ self.block20 = BlockTypeB(128, 64)
+
+ ## A, B, C
+ self.block21 = BlockTypeA(in_c1=16, in_c2=64,
+ out_c1=64, out_c2=64)
+ self.block22 = BlockTypeB(128, 64)
+
+ self.block23 = BlockTypeC(64, 16)
+
+ def forward(self, x):
+ c1, c2, c3, c4, c5 = self.backbone(x)
+
+ x = self.block15(c4, c5)
+ x = self.block16(x)
+
+ x = self.block17(c3, x)
+ x = self.block18(x)
+
+ x = self.block19(c2, x)
+ x = self.block20(x)
+
+ x = self.block21(c1, x)
+ x = self.block22(x)
+ x = self.block23(x)
+ x = x[:, 7:, :, :]
+
+ return x
\ No newline at end of file
diff --git a/ControlNet-v1-1-nightly-main/annotator/mlsd/models/mbv2_mlsd_tiny.py b/ControlNet-v1-1-nightly-main/annotator/mlsd/models/mbv2_mlsd_tiny.py
new file mode 100644
index 0000000000000000000000000000000000000000..e3ed633f2cc23ea1829a627fdb879ab39f641f83
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/mlsd/models/mbv2_mlsd_tiny.py
@@ -0,0 +1,275 @@
+import os
+import sys
+import torch
+import torch.nn as nn
+import torch.utils.model_zoo as model_zoo
+from torch.nn import functional as F
+
+
+class BlockTypeA(nn.Module):
+ def __init__(self, in_c1, in_c2, out_c1, out_c2, upscale = True):
+ super(BlockTypeA, self).__init__()
+ self.conv1 = nn.Sequential(
+ nn.Conv2d(in_c2, out_c2, kernel_size=1),
+ nn.BatchNorm2d(out_c2),
+ nn.ReLU(inplace=True)
+ )
+ self.conv2 = nn.Sequential(
+ nn.Conv2d(in_c1, out_c1, kernel_size=1),
+ nn.BatchNorm2d(out_c1),
+ nn.ReLU(inplace=True)
+ )
+ self.upscale = upscale
+
+ def forward(self, a, b):
+ b = self.conv1(b)
+ a = self.conv2(a)
+ b = F.interpolate(b, scale_factor=2.0, mode='bilinear', align_corners=True)
+ return torch.cat((a, b), dim=1)
+
+
+class BlockTypeB(nn.Module):
+ def __init__(self, in_c, out_c):
+ super(BlockTypeB, self).__init__()
+ self.conv1 = nn.Sequential(
+ nn.Conv2d(in_c, in_c, kernel_size=3, padding=1),
+ nn.BatchNorm2d(in_c),
+ nn.ReLU()
+ )
+ self.conv2 = nn.Sequential(
+ nn.Conv2d(in_c, out_c, kernel_size=3, padding=1),
+ nn.BatchNorm2d(out_c),
+ nn.ReLU()
+ )
+
+ def forward(self, x):
+ x = self.conv1(x) + x
+ x = self.conv2(x)
+ return x
+
+class BlockTypeC(nn.Module):
+ def __init__(self, in_c, out_c):
+ super(BlockTypeC, self).__init__()
+ self.conv1 = nn.Sequential(
+ nn.Conv2d(in_c, in_c, kernel_size=3, padding=5, dilation=5),
+ nn.BatchNorm2d(in_c),
+ nn.ReLU()
+ )
+ self.conv2 = nn.Sequential(
+ nn.Conv2d(in_c, in_c, kernel_size=3, padding=1),
+ nn.BatchNorm2d(in_c),
+ nn.ReLU()
+ )
+ self.conv3 = nn.Conv2d(in_c, out_c, kernel_size=1)
+
+ def forward(self, x):
+ x = self.conv1(x)
+ x = self.conv2(x)
+ x = self.conv3(x)
+ return x
+
+def _make_divisible(v, divisor, min_value=None):
+ """
+ This function is taken from the original tf repo.
+ It ensures that all layers have a channel number that is divisible by 8
+ It can be seen here:
+ https://github.com/tensorflow/models/blob/master/research/slim/nets/mobilenet/mobilenet.py
+ :param v:
+ :param divisor:
+ :param min_value:
+ :return:
+ """
+ if min_value is None:
+ min_value = divisor
+ new_v = max(min_value, int(v + divisor / 2) // divisor * divisor)
+ # Make sure that round down does not go down by more than 10%.
+ if new_v < 0.9 * v:
+ new_v += divisor
+ return new_v
+
+
+class ConvBNReLU(nn.Sequential):
+ def __init__(self, in_planes, out_planes, kernel_size=3, stride=1, groups=1):
+ self.channel_pad = out_planes - in_planes
+ self.stride = stride
+ #padding = (kernel_size - 1) // 2
+
+ # TFLite uses slightly different padding than PyTorch
+ if stride == 2:
+ padding = 0
+ else:
+ padding = (kernel_size - 1) // 2
+
+ super(ConvBNReLU, self).__init__(
+ nn.Conv2d(in_planes, out_planes, kernel_size, stride, padding, groups=groups, bias=False),
+ nn.BatchNorm2d(out_planes),
+ nn.ReLU6(inplace=True)
+ )
+ self.max_pool = nn.MaxPool2d(kernel_size=stride, stride=stride)
+
+
+ def forward(self, x):
+ # TFLite uses different padding
+ if self.stride == 2:
+ x = F.pad(x, (0, 1, 0, 1), "constant", 0)
+ #print(x.shape)
+
+ for module in self:
+ if not isinstance(module, nn.MaxPool2d):
+ x = module(x)
+ return x
+
+
+class InvertedResidual(nn.Module):
+ def __init__(self, inp, oup, stride, expand_ratio):
+ super(InvertedResidual, self).__init__()
+ self.stride = stride
+ assert stride in [1, 2]
+
+ hidden_dim = int(round(inp * expand_ratio))
+ self.use_res_connect = self.stride == 1 and inp == oup
+
+ layers = []
+ if expand_ratio != 1:
+ # pw
+ layers.append(ConvBNReLU(inp, hidden_dim, kernel_size=1))
+ layers.extend([
+ # dw
+ ConvBNReLU(hidden_dim, hidden_dim, stride=stride, groups=hidden_dim),
+ # pw-linear
+ nn.Conv2d(hidden_dim, oup, 1, 1, 0, bias=False),
+ nn.BatchNorm2d(oup),
+ ])
+ self.conv = nn.Sequential(*layers)
+
+ def forward(self, x):
+ if self.use_res_connect:
+ return x + self.conv(x)
+ else:
+ return self.conv(x)
+
+
+class MobileNetV2(nn.Module):
+ def __init__(self, pretrained=True):
+ """
+ MobileNet V2 main class
+ Args:
+ num_classes (int): Number of classes
+ width_mult (float): Width multiplier - adjusts number of channels in each layer by this amount
+ inverted_residual_setting: Network structure
+ round_nearest (int): Round the number of channels in each layer to be a multiple of this number
+ Set to 1 to turn off rounding
+ block: Module specifying inverted residual building block for mobilenet
+ """
+ super(MobileNetV2, self).__init__()
+
+ block = InvertedResidual
+ input_channel = 32
+ last_channel = 1280
+ width_mult = 1.0
+ round_nearest = 8
+
+ inverted_residual_setting = [
+ # t, c, n, s
+ [1, 16, 1, 1],
+ [6, 24, 2, 2],
+ [6, 32, 3, 2],
+ [6, 64, 4, 2],
+ #[6, 96, 3, 1],
+ #[6, 160, 3, 2],
+ #[6, 320, 1, 1],
+ ]
+
+ # only check the first element, assuming user knows t,c,n,s are required
+ if len(inverted_residual_setting) == 0 or len(inverted_residual_setting[0]) != 4:
+ raise ValueError("inverted_residual_setting should be non-empty "
+ "or a 4-element list, got {}".format(inverted_residual_setting))
+
+ # building first layer
+ input_channel = _make_divisible(input_channel * width_mult, round_nearest)
+ self.last_channel = _make_divisible(last_channel * max(1.0, width_mult), round_nearest)
+ features = [ConvBNReLU(4, input_channel, stride=2)]
+ # building inverted residual blocks
+ for t, c, n, s in inverted_residual_setting:
+ output_channel = _make_divisible(c * width_mult, round_nearest)
+ for i in range(n):
+ stride = s if i == 0 else 1
+ features.append(block(input_channel, output_channel, stride, expand_ratio=t))
+ input_channel = output_channel
+ self.features = nn.Sequential(*features)
+
+ self.fpn_selected = [3, 6, 10]
+ # weight initialization
+ for m in self.modules():
+ if isinstance(m, nn.Conv2d):
+ nn.init.kaiming_normal_(m.weight, mode='fan_out')
+ if m.bias is not None:
+ nn.init.zeros_(m.bias)
+ elif isinstance(m, nn.BatchNorm2d):
+ nn.init.ones_(m.weight)
+ nn.init.zeros_(m.bias)
+ elif isinstance(m, nn.Linear):
+ nn.init.normal_(m.weight, 0, 0.01)
+ nn.init.zeros_(m.bias)
+
+ #if pretrained:
+ # self._load_pretrained_model()
+
+ def _forward_impl(self, x):
+ # This exists since TorchScript doesn't support inheritance, so the superclass method
+ # (this one) needs to have a name other than `forward` that can be accessed in a subclass
+ fpn_features = []
+ for i, f in enumerate(self.features):
+ if i > self.fpn_selected[-1]:
+ break
+ x = f(x)
+ if i in self.fpn_selected:
+ fpn_features.append(x)
+
+ c2, c3, c4 = fpn_features
+ return c2, c3, c4
+
+
+ def forward(self, x):
+ return self._forward_impl(x)
+
+ def _load_pretrained_model(self):
+ pretrain_dict = model_zoo.load_url('https://download.pytorch.org/models/mobilenet_v2-b0353104.pth')
+ model_dict = {}
+ state_dict = self.state_dict()
+ for k, v in pretrain_dict.items():
+ if k in state_dict:
+ model_dict[k] = v
+ state_dict.update(model_dict)
+ self.load_state_dict(state_dict)
+
+
+class MobileV2_MLSD_Tiny(nn.Module):
+ def __init__(self):
+ super(MobileV2_MLSD_Tiny, self).__init__()
+
+ self.backbone = MobileNetV2(pretrained=True)
+
+ self.block12 = BlockTypeA(in_c1= 32, in_c2= 64,
+ out_c1= 64, out_c2=64)
+ self.block13 = BlockTypeB(128, 64)
+
+ self.block14 = BlockTypeA(in_c1 = 24, in_c2 = 64,
+ out_c1= 32, out_c2= 32)
+ self.block15 = BlockTypeB(64, 64)
+
+ self.block16 = BlockTypeC(64, 16)
+
+ def forward(self, x):
+ c2, c3, c4 = self.backbone(x)
+
+ x = self.block12(c3, c4)
+ x = self.block13(x)
+ x = self.block14(c2, x)
+ x = self.block15(x)
+ x = self.block16(x)
+ x = x[:, 7:, :, :]
+ #print(x.shape)
+ x = F.interpolate(x, scale_factor=2.0, mode='bilinear', align_corners=True)
+
+ return x
\ No newline at end of file
diff --git a/ControlNet-v1-1-nightly-main/annotator/mlsd/utils.py b/ControlNet-v1-1-nightly-main/annotator/mlsd/utils.py
new file mode 100644
index 0000000000000000000000000000000000000000..ae3cf9420a33a4abae27c48ac4b90938c7d63cc3
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/mlsd/utils.py
@@ -0,0 +1,580 @@
+'''
+modified by lihaoweicv
+pytorch version
+'''
+
+'''
+M-LSD
+Copyright 2021-present NAVER Corp.
+Apache License v2.0
+'''
+
+import os
+import numpy as np
+import cv2
+import torch
+from torch.nn import functional as F
+
+
+def deccode_output_score_and_ptss(tpMap, topk_n = 200, ksize = 5):
+ '''
+ tpMap:
+ center: tpMap[1, 0, :, :]
+ displacement: tpMap[1, 1:5, :, :]
+ '''
+ b, c, h, w = tpMap.shape
+ assert b==1, 'only support bsize==1'
+ displacement = tpMap[:, 1:5, :, :][0]
+ center = tpMap[:, 0, :, :]
+ heat = torch.sigmoid(center)
+ hmax = F.max_pool2d( heat, (ksize, ksize), stride=1, padding=(ksize-1)//2)
+ keep = (hmax == heat).float()
+ heat = heat * keep
+ heat = heat.reshape(-1, )
+
+ scores, indices = torch.topk(heat, topk_n, dim=-1, largest=True)
+ yy = torch.floor_divide(indices, w).unsqueeze(-1)
+ xx = torch.fmod(indices, w).unsqueeze(-1)
+ ptss = torch.cat((yy, xx),dim=-1)
+
+ ptss = ptss.detach().cpu().numpy()
+ scores = scores.detach().cpu().numpy()
+ displacement = displacement.detach().cpu().numpy()
+ displacement = displacement.transpose((1,2,0))
+ return ptss, scores, displacement
+
+
+def pred_lines(image, model,
+ input_shape=[512, 512],
+ score_thr=0.10,
+ dist_thr=20.0):
+ h, w, _ = image.shape
+ h_ratio, w_ratio = [h / input_shape[0], w / input_shape[1]]
+
+ resized_image = np.concatenate([cv2.resize(image, (input_shape[1], input_shape[0]), interpolation=cv2.INTER_AREA),
+ np.ones([input_shape[0], input_shape[1], 1])], axis=-1)
+
+ resized_image = resized_image.transpose((2,0,1))
+ batch_image = np.expand_dims(resized_image, axis=0).astype('float32')
+ batch_image = (batch_image / 127.5) - 1.0
+
+ batch_image = torch.from_numpy(batch_image).float().cuda()
+ outputs = model(batch_image)
+ pts, pts_score, vmap = deccode_output_score_and_ptss(outputs, 200, 3)
+ start = vmap[:, :, :2]
+ end = vmap[:, :, 2:]
+ dist_map = np.sqrt(np.sum((start - end) ** 2, axis=-1))
+
+ segments_list = []
+ for center, score in zip(pts, pts_score):
+ y, x = center
+ distance = dist_map[y, x]
+ if score > score_thr and distance > dist_thr:
+ disp_x_start, disp_y_start, disp_x_end, disp_y_end = vmap[y, x, :]
+ x_start = x + disp_x_start
+ y_start = y + disp_y_start
+ x_end = x + disp_x_end
+ y_end = y + disp_y_end
+ segments_list.append([x_start, y_start, x_end, y_end])
+
+ lines = 2 * np.array(segments_list) # 256 > 512
+ lines[:, 0] = lines[:, 0] * w_ratio
+ lines[:, 1] = lines[:, 1] * h_ratio
+ lines[:, 2] = lines[:, 2] * w_ratio
+ lines[:, 3] = lines[:, 3] * h_ratio
+
+ return lines
+
+
+def pred_squares(image,
+ model,
+ input_shape=[512, 512],
+ params={'score': 0.06,
+ 'outside_ratio': 0.28,
+ 'inside_ratio': 0.45,
+ 'w_overlap': 0.0,
+ 'w_degree': 1.95,
+ 'w_length': 0.0,
+ 'w_area': 1.86,
+ 'w_center': 0.14}):
+ '''
+ shape = [height, width]
+ '''
+ h, w, _ = image.shape
+ original_shape = [h, w]
+
+ resized_image = np.concatenate([cv2.resize(image, (input_shape[0], input_shape[1]), interpolation=cv2.INTER_AREA),
+ np.ones([input_shape[0], input_shape[1], 1])], axis=-1)
+ resized_image = resized_image.transpose((2, 0, 1))
+ batch_image = np.expand_dims(resized_image, axis=0).astype('float32')
+ batch_image = (batch_image / 127.5) - 1.0
+
+ batch_image = torch.from_numpy(batch_image).float().cuda()
+ outputs = model(batch_image)
+
+ pts, pts_score, vmap = deccode_output_score_and_ptss(outputs, 200, 3)
+ start = vmap[:, :, :2] # (x, y)
+ end = vmap[:, :, 2:] # (x, y)
+ dist_map = np.sqrt(np.sum((start - end) ** 2, axis=-1))
+
+ junc_list = []
+ segments_list = []
+ for junc, score in zip(pts, pts_score):
+ y, x = junc
+ distance = dist_map[y, x]
+ if score > params['score'] and distance > 20.0:
+ junc_list.append([x, y])
+ disp_x_start, disp_y_start, disp_x_end, disp_y_end = vmap[y, x, :]
+ d_arrow = 1.0
+ x_start = x + d_arrow * disp_x_start
+ y_start = y + d_arrow * disp_y_start
+ x_end = x + d_arrow * disp_x_end
+ y_end = y + d_arrow * disp_y_end
+ segments_list.append([x_start, y_start, x_end, y_end])
+
+ segments = np.array(segments_list)
+
+ ####### post processing for squares
+ # 1. get unique lines
+ point = np.array([[0, 0]])
+ point = point[0]
+ start = segments[:, :2]
+ end = segments[:, 2:]
+ diff = start - end
+ a = diff[:, 1]
+ b = -diff[:, 0]
+ c = a * start[:, 0] + b * start[:, 1]
+
+ d = np.abs(a * point[0] + b * point[1] - c) / np.sqrt(a ** 2 + b ** 2 + 1e-10)
+ theta = np.arctan2(diff[:, 0], diff[:, 1]) * 180 / np.pi
+ theta[theta < 0.0] += 180
+ hough = np.concatenate([d[:, None], theta[:, None]], axis=-1)
+
+ d_quant = 1
+ theta_quant = 2
+ hough[:, 0] //= d_quant
+ hough[:, 1] //= theta_quant
+ _, indices, counts = np.unique(hough, axis=0, return_index=True, return_counts=True)
+
+ acc_map = np.zeros([512 // d_quant + 1, 360 // theta_quant + 1], dtype='float32')
+ idx_map = np.zeros([512 // d_quant + 1, 360 // theta_quant + 1], dtype='int32') - 1
+ yx_indices = hough[indices, :].astype('int32')
+ acc_map[yx_indices[:, 0], yx_indices[:, 1]] = counts
+ idx_map[yx_indices[:, 0], yx_indices[:, 1]] = indices
+
+ acc_map_np = acc_map
+ # acc_map = acc_map[None, :, :, None]
+ #
+ # ### fast suppression using tensorflow op
+ # acc_map = tf.constant(acc_map, dtype=tf.float32)
+ # max_acc_map = tf.keras.layers.MaxPool2D(pool_size=(5, 5), strides=1, padding='same')(acc_map)
+ # acc_map = acc_map * tf.cast(tf.math.equal(acc_map, max_acc_map), tf.float32)
+ # flatten_acc_map = tf.reshape(acc_map, [1, -1])
+ # topk_values, topk_indices = tf.math.top_k(flatten_acc_map, k=len(pts))
+ # _, h, w, _ = acc_map.shape
+ # y = tf.expand_dims(topk_indices // w, axis=-1)
+ # x = tf.expand_dims(topk_indices % w, axis=-1)
+ # yx = tf.concat([y, x], axis=-1)
+
+ ### fast suppression using pytorch op
+ acc_map = torch.from_numpy(acc_map_np).unsqueeze(0).unsqueeze(0)
+ _,_, h, w = acc_map.shape
+ max_acc_map = F.max_pool2d(acc_map,kernel_size=5, stride=1, padding=2)
+ acc_map = acc_map * ( (acc_map == max_acc_map).float() )
+ flatten_acc_map = acc_map.reshape([-1, ])
+
+ scores, indices = torch.topk(flatten_acc_map, len(pts), dim=-1, largest=True)
+ yy = torch.div(indices, w, rounding_mode='floor').unsqueeze(-1)
+ xx = torch.fmod(indices, w).unsqueeze(-1)
+ yx = torch.cat((yy, xx), dim=-1)
+
+ yx = yx.detach().cpu().numpy()
+
+ topk_values = scores.detach().cpu().numpy()
+ indices = idx_map[yx[:, 0], yx[:, 1]]
+ basis = 5 // 2
+
+ merged_segments = []
+ for yx_pt, max_indice, value in zip(yx, indices, topk_values):
+ y, x = yx_pt
+ if max_indice == -1 or value == 0:
+ continue
+ segment_list = []
+ for y_offset in range(-basis, basis + 1):
+ for x_offset in range(-basis, basis + 1):
+ indice = idx_map[y + y_offset, x + x_offset]
+ cnt = int(acc_map_np[y + y_offset, x + x_offset])
+ if indice != -1:
+ segment_list.append(segments[indice])
+ if cnt > 1:
+ check_cnt = 1
+ current_hough = hough[indice]
+ for new_indice, new_hough in enumerate(hough):
+ if (current_hough == new_hough).all() and indice != new_indice:
+ segment_list.append(segments[new_indice])
+ check_cnt += 1
+ if check_cnt == cnt:
+ break
+ group_segments = np.array(segment_list).reshape([-1, 2])
+ sorted_group_segments = np.sort(group_segments, axis=0)
+ x_min, y_min = sorted_group_segments[0, :]
+ x_max, y_max = sorted_group_segments[-1, :]
+
+ deg = theta[max_indice]
+ if deg >= 90:
+ merged_segments.append([x_min, y_max, x_max, y_min])
+ else:
+ merged_segments.append([x_min, y_min, x_max, y_max])
+
+ # 2. get intersections
+ new_segments = np.array(merged_segments) # (x1, y1, x2, y2)
+ start = new_segments[:, :2] # (x1, y1)
+ end = new_segments[:, 2:] # (x2, y2)
+ new_centers = (start + end) / 2.0
+ diff = start - end
+ dist_segments = np.sqrt(np.sum(diff ** 2, axis=-1))
+
+ # ax + by = c
+ a = diff[:, 1]
+ b = -diff[:, 0]
+ c = a * start[:, 0] + b * start[:, 1]
+ pre_det = a[:, None] * b[None, :]
+ det = pre_det - np.transpose(pre_det)
+
+ pre_inter_y = a[:, None] * c[None, :]
+ inter_y = (pre_inter_y - np.transpose(pre_inter_y)) / (det + 1e-10)
+ pre_inter_x = c[:, None] * b[None, :]
+ inter_x = (pre_inter_x - np.transpose(pre_inter_x)) / (det + 1e-10)
+ inter_pts = np.concatenate([inter_x[:, :, None], inter_y[:, :, None]], axis=-1).astype('int32')
+
+ # 3. get corner information
+ # 3.1 get distance
+ '''
+ dist_segments:
+ | dist(0), dist(1), dist(2), ...|
+ dist_inter_to_segment1:
+ | dist(inter,0), dist(inter,0), dist(inter,0), ... |
+ | dist(inter,1), dist(inter,1), dist(inter,1), ... |
+ ...
+ dist_inter_to_semgnet2:
+ | dist(inter,0), dist(inter,1), dist(inter,2), ... |
+ | dist(inter,0), dist(inter,1), dist(inter,2), ... |
+ ...
+ '''
+
+ dist_inter_to_segment1_start = np.sqrt(
+ np.sum(((inter_pts - start[:, None, :]) ** 2), axis=-1, keepdims=True)) # [n_batch, n_batch, 1]
+ dist_inter_to_segment1_end = np.sqrt(
+ np.sum(((inter_pts - end[:, None, :]) ** 2), axis=-1, keepdims=True)) # [n_batch, n_batch, 1]
+ dist_inter_to_segment2_start = np.sqrt(
+ np.sum(((inter_pts - start[None, :, :]) ** 2), axis=-1, keepdims=True)) # [n_batch, n_batch, 1]
+ dist_inter_to_segment2_end = np.sqrt(
+ np.sum(((inter_pts - end[None, :, :]) ** 2), axis=-1, keepdims=True)) # [n_batch, n_batch, 1]
+
+ # sort ascending
+ dist_inter_to_segment1 = np.sort(
+ np.concatenate([dist_inter_to_segment1_start, dist_inter_to_segment1_end], axis=-1),
+ axis=-1) # [n_batch, n_batch, 2]
+ dist_inter_to_segment2 = np.sort(
+ np.concatenate([dist_inter_to_segment2_start, dist_inter_to_segment2_end], axis=-1),
+ axis=-1) # [n_batch, n_batch, 2]
+
+ # 3.2 get degree
+ inter_to_start = new_centers[:, None, :] - inter_pts
+ deg_inter_to_start = np.arctan2(inter_to_start[:, :, 1], inter_to_start[:, :, 0]) * 180 / np.pi
+ deg_inter_to_start[deg_inter_to_start < 0.0] += 360
+ inter_to_end = new_centers[None, :, :] - inter_pts
+ deg_inter_to_end = np.arctan2(inter_to_end[:, :, 1], inter_to_end[:, :, 0]) * 180 / np.pi
+ deg_inter_to_end[deg_inter_to_end < 0.0] += 360
+
+ '''
+ B -- G
+ | |
+ C -- R
+ B : blue / G: green / C: cyan / R: red
+
+ 0 -- 1
+ | |
+ 3 -- 2
+ '''
+ # rename variables
+ deg1_map, deg2_map = deg_inter_to_start, deg_inter_to_end
+ # sort deg ascending
+ deg_sort = np.sort(np.concatenate([deg1_map[:, :, None], deg2_map[:, :, None]], axis=-1), axis=-1)
+
+ deg_diff_map = np.abs(deg1_map - deg2_map)
+ # we only consider the smallest degree of intersect
+ deg_diff_map[deg_diff_map > 180] = 360 - deg_diff_map[deg_diff_map > 180]
+
+ # define available degree range
+ deg_range = [60, 120]
+
+ corner_dict = {corner_info: [] for corner_info in range(4)}
+ inter_points = []
+ for i in range(inter_pts.shape[0]):
+ for j in range(i + 1, inter_pts.shape[1]):
+ # i, j > line index, always i < j
+ x, y = inter_pts[i, j, :]
+ deg1, deg2 = deg_sort[i, j, :]
+ deg_diff = deg_diff_map[i, j]
+
+ check_degree = deg_diff > deg_range[0] and deg_diff < deg_range[1]
+
+ outside_ratio = params['outside_ratio'] # over ratio >>> drop it!
+ inside_ratio = params['inside_ratio'] # over ratio >>> drop it!
+ check_distance = ((dist_inter_to_segment1[i, j, 1] >= dist_segments[i] and \
+ dist_inter_to_segment1[i, j, 0] <= dist_segments[i] * outside_ratio) or \
+ (dist_inter_to_segment1[i, j, 1] <= dist_segments[i] and \
+ dist_inter_to_segment1[i, j, 0] <= dist_segments[i] * inside_ratio)) and \
+ ((dist_inter_to_segment2[i, j, 1] >= dist_segments[j] and \
+ dist_inter_to_segment2[i, j, 0] <= dist_segments[j] * outside_ratio) or \
+ (dist_inter_to_segment2[i, j, 1] <= dist_segments[j] and \
+ dist_inter_to_segment2[i, j, 0] <= dist_segments[j] * inside_ratio))
+
+ if check_degree and check_distance:
+ corner_info = None
+
+ if (deg1 >= 0 and deg1 <= 45 and deg2 >= 45 and deg2 <= 120) or \
+ (deg2 >= 315 and deg1 >= 45 and deg1 <= 120):
+ corner_info, color_info = 0, 'blue'
+ elif (deg1 >= 45 and deg1 <= 125 and deg2 >= 125 and deg2 <= 225):
+ corner_info, color_info = 1, 'green'
+ elif (deg1 >= 125 and deg1 <= 225 and deg2 >= 225 and deg2 <= 315):
+ corner_info, color_info = 2, 'black'
+ elif (deg1 >= 0 and deg1 <= 45 and deg2 >= 225 and deg2 <= 315) or \
+ (deg2 >= 315 and deg1 >= 225 and deg1 <= 315):
+ corner_info, color_info = 3, 'cyan'
+ else:
+ corner_info, color_info = 4, 'red' # we don't use it
+ continue
+
+ corner_dict[corner_info].append([x, y, i, j])
+ inter_points.append([x, y])
+
+ square_list = []
+ connect_list = []
+ segments_list = []
+ for corner0 in corner_dict[0]:
+ for corner1 in corner_dict[1]:
+ connect01 = False
+ for corner0_line in corner0[2:]:
+ if corner0_line in corner1[2:]:
+ connect01 = True
+ break
+ if connect01:
+ for corner2 in corner_dict[2]:
+ connect12 = False
+ for corner1_line in corner1[2:]:
+ if corner1_line in corner2[2:]:
+ connect12 = True
+ break
+ if connect12:
+ for corner3 in corner_dict[3]:
+ connect23 = False
+ for corner2_line in corner2[2:]:
+ if corner2_line in corner3[2:]:
+ connect23 = True
+ break
+ if connect23:
+ for corner3_line in corner3[2:]:
+ if corner3_line in corner0[2:]:
+ # SQUARE!!!
+ '''
+ 0 -- 1
+ | |
+ 3 -- 2
+ square_list:
+ order: 0 > 1 > 2 > 3
+ | x0, y0, x1, y1, x2, y2, x3, y3 |
+ | x0, y0, x1, y1, x2, y2, x3, y3 |
+ ...
+ connect_list:
+ order: 01 > 12 > 23 > 30
+ | line_idx01, line_idx12, line_idx23, line_idx30 |
+ | line_idx01, line_idx12, line_idx23, line_idx30 |
+ ...
+ segments_list:
+ order: 0 > 1 > 2 > 3
+ | line_idx0_i, line_idx0_j, line_idx1_i, line_idx1_j, line_idx2_i, line_idx2_j, line_idx3_i, line_idx3_j |
+ | line_idx0_i, line_idx0_j, line_idx1_i, line_idx1_j, line_idx2_i, line_idx2_j, line_idx3_i, line_idx3_j |
+ ...
+ '''
+ square_list.append(corner0[:2] + corner1[:2] + corner2[:2] + corner3[:2])
+ connect_list.append([corner0_line, corner1_line, corner2_line, corner3_line])
+ segments_list.append(corner0[2:] + corner1[2:] + corner2[2:] + corner3[2:])
+
+ def check_outside_inside(segments_info, connect_idx):
+ # return 'outside or inside', min distance, cover_param, peri_param
+ if connect_idx == segments_info[0]:
+ check_dist_mat = dist_inter_to_segment1
+ else:
+ check_dist_mat = dist_inter_to_segment2
+
+ i, j = segments_info
+ min_dist, max_dist = check_dist_mat[i, j, :]
+ connect_dist = dist_segments[connect_idx]
+ if max_dist > connect_dist:
+ return 'outside', min_dist, 0, 1
+ else:
+ return 'inside', min_dist, -1, -1
+
+ top_square = None
+
+ try:
+ map_size = input_shape[0] / 2
+ squares = np.array(square_list).reshape([-1, 4, 2])
+ score_array = []
+ connect_array = np.array(connect_list)
+ segments_array = np.array(segments_list).reshape([-1, 4, 2])
+
+ # get degree of corners:
+ squares_rollup = np.roll(squares, 1, axis=1)
+ squares_rolldown = np.roll(squares, -1, axis=1)
+ vec1 = squares_rollup - squares
+ normalized_vec1 = vec1 / (np.linalg.norm(vec1, axis=-1, keepdims=True) + 1e-10)
+ vec2 = squares_rolldown - squares
+ normalized_vec2 = vec2 / (np.linalg.norm(vec2, axis=-1, keepdims=True) + 1e-10)
+ inner_products = np.sum(normalized_vec1 * normalized_vec2, axis=-1) # [n_squares, 4]
+ squares_degree = np.arccos(inner_products) * 180 / np.pi # [n_squares, 4]
+
+ # get square score
+ overlap_scores = []
+ degree_scores = []
+ length_scores = []
+
+ for connects, segments, square, degree in zip(connect_array, segments_array, squares, squares_degree):
+ '''
+ 0 -- 1
+ | |
+ 3 -- 2
+
+ # segments: [4, 2]
+ # connects: [4]
+ '''
+
+ ###################################### OVERLAP SCORES
+ cover = 0
+ perimeter = 0
+ # check 0 > 1 > 2 > 3
+ square_length = []
+
+ for start_idx in range(4):
+ end_idx = (start_idx + 1) % 4
+
+ connect_idx = connects[start_idx] # segment idx of segment01
+ start_segments = segments[start_idx]
+ end_segments = segments[end_idx]
+
+ start_point = square[start_idx]
+ end_point = square[end_idx]
+
+ # check whether outside or inside
+ start_position, start_min, start_cover_param, start_peri_param = check_outside_inside(start_segments,
+ connect_idx)
+ end_position, end_min, end_cover_param, end_peri_param = check_outside_inside(end_segments, connect_idx)
+
+ cover += dist_segments[connect_idx] + start_cover_param * start_min + end_cover_param * end_min
+ perimeter += dist_segments[connect_idx] + start_peri_param * start_min + end_peri_param * end_min
+
+ square_length.append(
+ dist_segments[connect_idx] + start_peri_param * start_min + end_peri_param * end_min)
+
+ overlap_scores.append(cover / perimeter)
+ ######################################
+ ###################################### DEGREE SCORES
+ '''
+ deg0 vs deg2
+ deg1 vs deg3
+ '''
+ deg0, deg1, deg2, deg3 = degree
+ deg_ratio1 = deg0 / deg2
+ if deg_ratio1 > 1.0:
+ deg_ratio1 = 1 / deg_ratio1
+ deg_ratio2 = deg1 / deg3
+ if deg_ratio2 > 1.0:
+ deg_ratio2 = 1 / deg_ratio2
+ degree_scores.append((deg_ratio1 + deg_ratio2) / 2)
+ ######################################
+ ###################################### LENGTH SCORES
+ '''
+ len0 vs len2
+ len1 vs len3
+ '''
+ len0, len1, len2, len3 = square_length
+ len_ratio1 = len0 / len2 if len2 > len0 else len2 / len0
+ len_ratio2 = len1 / len3 if len3 > len1 else len3 / len1
+ length_scores.append((len_ratio1 + len_ratio2) / 2)
+
+ ######################################
+
+ overlap_scores = np.array(overlap_scores)
+ overlap_scores /= np.max(overlap_scores)
+
+ degree_scores = np.array(degree_scores)
+ # degree_scores /= np.max(degree_scores)
+
+ length_scores = np.array(length_scores)
+
+ ###################################### AREA SCORES
+ area_scores = np.reshape(squares, [-1, 4, 2])
+ area_x = area_scores[:, :, 0]
+ area_y = area_scores[:, :, 1]
+ correction = area_x[:, -1] * area_y[:, 0] - area_y[:, -1] * area_x[:, 0]
+ area_scores = np.sum(area_x[:, :-1] * area_y[:, 1:], axis=-1) - np.sum(area_y[:, :-1] * area_x[:, 1:], axis=-1)
+ area_scores = 0.5 * np.abs(area_scores + correction)
+ area_scores /= (map_size * map_size) # np.max(area_scores)
+ ######################################
+
+ ###################################### CENTER SCORES
+ centers = np.array([[256 // 2, 256 // 2]], dtype='float32') # [1, 2]
+ # squares: [n, 4, 2]
+ square_centers = np.mean(squares, axis=1) # [n, 2]
+ center2center = np.sqrt(np.sum((centers - square_centers) ** 2))
+ center_scores = center2center / (map_size / np.sqrt(2.0))
+
+ '''
+ score_w = [overlap, degree, area, center, length]
+ '''
+ score_w = [0.0, 1.0, 10.0, 0.5, 1.0]
+ score_array = params['w_overlap'] * overlap_scores \
+ + params['w_degree'] * degree_scores \
+ + params['w_area'] * area_scores \
+ - params['w_center'] * center_scores \
+ + params['w_length'] * length_scores
+
+ best_square = []
+
+ sorted_idx = np.argsort(score_array)[::-1]
+ score_array = score_array[sorted_idx]
+ squares = squares[sorted_idx]
+
+ except Exception as e:
+ pass
+
+ '''return list
+ merged_lines, squares, scores
+ '''
+
+ try:
+ new_segments[:, 0] = new_segments[:, 0] * 2 / input_shape[1] * original_shape[1]
+ new_segments[:, 1] = new_segments[:, 1] * 2 / input_shape[0] * original_shape[0]
+ new_segments[:, 2] = new_segments[:, 2] * 2 / input_shape[1] * original_shape[1]
+ new_segments[:, 3] = new_segments[:, 3] * 2 / input_shape[0] * original_shape[0]
+ except:
+ new_segments = []
+
+ try:
+ squares[:, :, 0] = squares[:, :, 0] * 2 / input_shape[1] * original_shape[1]
+ squares[:, :, 1] = squares[:, :, 1] * 2 / input_shape[0] * original_shape[0]
+ except:
+ squares = []
+ score_array = []
+
+ try:
+ inter_points = np.array(inter_points)
+ inter_points[:, 0] = inter_points[:, 0] * 2 / input_shape[1] * original_shape[1]
+ inter_points[:, 1] = inter_points[:, 1] * 2 / input_shape[0] * original_shape[0]
+ except:
+ inter_points = []
+
+ return new_segments, squares, score_array, inter_points
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/LICENSE b/ControlNet-v1-1-nightly-main/annotator/normalbae/LICENSE
new file mode 100644
index 0000000000000000000000000000000000000000..16a9d56a3d4c15e4f34ac5426459c58487b01520
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/LICENSE
@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2022 Caroline Chan
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
\ No newline at end of file
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/__init__.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..1d8cf903ce1e2d0e9166e9a0a0626c79f475bb75
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/__init__.py
@@ -0,0 +1,55 @@
+# Estimating and Exploiting the Aleatoric Uncertainty in Surface Normal Estimation
+# https://github.com/baegwangbin/surface_normal_uncertainty
+
+import os
+import types
+import torch
+import numpy as np
+
+from einops import rearrange
+from .models.NNET import NNET
+from .utils import utils
+from annotator.util import annotator_ckpts_path
+import torchvision.transforms as transforms
+
+
+class NormalBaeDetector:
+ def __init__(self):
+ remote_model_path = "https://huggingface.co/lllyasviel/Annotators/resolve/main/scannet.pt"
+ modelpath = os.path.join(annotator_ckpts_path, "scannet.pt")
+ if not os.path.exists(modelpath):
+ from basicsr.utils.download_util import load_file_from_url
+ load_file_from_url(remote_model_path, model_dir=annotator_ckpts_path)
+ args = types.SimpleNamespace()
+ args.mode = 'client'
+ args.architecture = 'BN'
+ args.pretrained = 'scannet'
+ args.sampling_ratio = 0.4
+ args.importance_ratio = 0.7
+ model = NNET(args)
+ model = utils.load_checkpoint(modelpath, model)
+ model = model.cuda()
+ model.eval()
+ self.model = model
+ self.norm = transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])
+
+ def __call__(self, input_image):
+ assert input_image.ndim == 3
+ image_normal = input_image
+ with torch.no_grad():
+ image_normal = torch.from_numpy(image_normal).float().cuda()
+ image_normal = image_normal / 255.0
+ image_normal = rearrange(image_normal, 'h w c -> 1 c h w')
+ image_normal = self.norm(image_normal)
+
+ normal = self.model(image_normal)
+ normal = normal[0][-1][:, :3]
+ # d = torch.sum(normal ** 2.0, dim=1, keepdim=True) ** 0.5
+ # d = torch.maximum(d, torch.ones_like(d) * 1e-5)
+ # normal /= d
+ normal = ((normal + 1) * 0.5).clip(0, 1)
+
+ normal = rearrange(normal[0], 'c h w -> h w c').cpu().numpy()
+ normal_image = (normal * 255.0).clip(0, 255).astype(np.uint8)
+
+ return normal_image
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/NNET.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/NNET.py
new file mode 100644
index 0000000000000000000000000000000000000000..3ddbc50c3ac18aa4b7f16779fe3c0133981ecc7a
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/NNET.py
@@ -0,0 +1,22 @@
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+
+from .submodules.encoder import Encoder
+from .submodules.decoder import Decoder
+
+
+class NNET(nn.Module):
+ def __init__(self, args):
+ super(NNET, self).__init__()
+ self.encoder = Encoder()
+ self.decoder = Decoder(args)
+
+ def get_1x_lr_params(self): # lr/10 learning rate
+ return self.encoder.parameters()
+
+ def get_10x_lr_params(self): # lr learning rate
+ return self.decoder.parameters()
+
+ def forward(self, img, **kwargs):
+ return self.decoder(self.encoder(img), **kwargs)
\ No newline at end of file
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/baseline.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/baseline.py
new file mode 100644
index 0000000000000000000000000000000000000000..602d0fbdac1acc9ede9bc1f2e10a5df78831ce9d
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/baseline.py
@@ -0,0 +1,85 @@
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+
+from .submodules.submodules import UpSampleBN, norm_normalize
+
+
+# This is the baseline encoder-decoder we used in the ablation study
+class NNET(nn.Module):
+ def __init__(self, args=None):
+ super(NNET, self).__init__()
+ self.encoder = Encoder()
+ self.decoder = Decoder(num_classes=4)
+
+ def forward(self, x, **kwargs):
+ out = self.decoder(self.encoder(x), **kwargs)
+
+ # Bilinearly upsample the output to match the input resolution
+ up_out = F.interpolate(out, size=[x.size(2), x.size(3)], mode='bilinear', align_corners=False)
+
+ # L2-normalize the first three channels / ensure positive value for concentration parameters (kappa)
+ up_out = norm_normalize(up_out)
+ return up_out
+
+ def get_1x_lr_params(self): # lr/10 learning rate
+ return self.encoder.parameters()
+
+ def get_10x_lr_params(self): # lr learning rate
+ modules = [self.decoder]
+ for m in modules:
+ yield from m.parameters()
+
+
+# Encoder
+class Encoder(nn.Module):
+ def __init__(self):
+ super(Encoder, self).__init__()
+
+ basemodel_name = 'tf_efficientnet_b5_ap'
+ basemodel = torch.hub.load('rwightman/gen-efficientnet-pytorch', basemodel_name, pretrained=True)
+
+ # Remove last layer
+ basemodel.global_pool = nn.Identity()
+ basemodel.classifier = nn.Identity()
+
+ self.original_model = basemodel
+
+ def forward(self, x):
+ features = [x]
+ for k, v in self.original_model._modules.items():
+ if (k == 'blocks'):
+ for ki, vi in v._modules.items():
+ features.append(vi(features[-1]))
+ else:
+ features.append(v(features[-1]))
+ return features
+
+
+# Decoder (no pixel-wise MLP, no uncertainty-guided sampling)
+class Decoder(nn.Module):
+ def __init__(self, num_classes=4):
+ super(Decoder, self).__init__()
+ self.conv2 = nn.Conv2d(2048, 2048, kernel_size=1, stride=1, padding=0)
+ self.up1 = UpSampleBN(skip_input=2048 + 176, output_features=1024)
+ self.up2 = UpSampleBN(skip_input=1024 + 64, output_features=512)
+ self.up3 = UpSampleBN(skip_input=512 + 40, output_features=256)
+ self.up4 = UpSampleBN(skip_input=256 + 24, output_features=128)
+ self.conv3 = nn.Conv2d(128, num_classes, kernel_size=3, stride=1, padding=1)
+
+ def forward(self, features):
+ x_block0, x_block1, x_block2, x_block3, x_block4 = features[4], features[5], features[6], features[8], features[11]
+ x_d0 = self.conv2(x_block4)
+ x_d1 = self.up1(x_d0, x_block3)
+ x_d2 = self.up2(x_d1, x_block2)
+ x_d3 = self.up3(x_d2, x_block1)
+ x_d4 = self.up4(x_d3, x_block0)
+ out = self.conv3(x_d4)
+ return out
+
+
+if __name__ == '__main__':
+ model = Baseline()
+ x = torch.rand(2, 3, 480, 640)
+ out = model(x)
+ print(out.shape)
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/decoder.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/decoder.py
new file mode 100644
index 0000000000000000000000000000000000000000..993203d1792311f1c492091eaea3c1ac9088187f
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/decoder.py
@@ -0,0 +1,202 @@
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from .submodules import UpSampleBN, UpSampleGN, norm_normalize, sample_points
+
+
+class Decoder(nn.Module):
+ def __init__(self, args):
+ super(Decoder, self).__init__()
+
+ # hyper-parameter for sampling
+ self.sampling_ratio = args.sampling_ratio
+ self.importance_ratio = args.importance_ratio
+
+ # feature-map
+ self.conv2 = nn.Conv2d(2048, 2048, kernel_size=1, stride=1, padding=0)
+ if args.architecture == 'BN':
+ self.up1 = UpSampleBN(skip_input=2048 + 176, output_features=1024)
+ self.up2 = UpSampleBN(skip_input=1024 + 64, output_features=512)
+ self.up3 = UpSampleBN(skip_input=512 + 40, output_features=256)
+ self.up4 = UpSampleBN(skip_input=256 + 24, output_features=128)
+
+ elif args.architecture == 'GN':
+ self.up1 = UpSampleGN(skip_input=2048 + 176, output_features=1024)
+ self.up2 = UpSampleGN(skip_input=1024 + 64, output_features=512)
+ self.up3 = UpSampleGN(skip_input=512 + 40, output_features=256)
+ self.up4 = UpSampleGN(skip_input=256 + 24, output_features=128)
+
+ else:
+ raise Exception('invalid architecture')
+
+ # produces 1/8 res output
+ self.out_conv_res8 = nn.Conv2d(512, 4, kernel_size=3, stride=1, padding=1)
+
+ # produces 1/4 res output
+ self.out_conv_res4 = nn.Sequential(
+ nn.Conv1d(512 + 4, 128, kernel_size=1), nn.ReLU(),
+ nn.Conv1d(128, 128, kernel_size=1), nn.ReLU(),
+ nn.Conv1d(128, 128, kernel_size=1), nn.ReLU(),
+ nn.Conv1d(128, 4, kernel_size=1),
+ )
+
+ # produces 1/2 res output
+ self.out_conv_res2 = nn.Sequential(
+ nn.Conv1d(256 + 4, 128, kernel_size=1), nn.ReLU(),
+ nn.Conv1d(128, 128, kernel_size=1), nn.ReLU(),
+ nn.Conv1d(128, 128, kernel_size=1), nn.ReLU(),
+ nn.Conv1d(128, 4, kernel_size=1),
+ )
+
+ # produces 1/1 res output
+ self.out_conv_res1 = nn.Sequential(
+ nn.Conv1d(128 + 4, 128, kernel_size=1), nn.ReLU(),
+ nn.Conv1d(128, 128, kernel_size=1), nn.ReLU(),
+ nn.Conv1d(128, 128, kernel_size=1), nn.ReLU(),
+ nn.Conv1d(128, 4, kernel_size=1),
+ )
+
+ def forward(self, features, gt_norm_mask=None, mode='test'):
+ x_block0, x_block1, x_block2, x_block3, x_block4 = features[4], features[5], features[6], features[8], features[11]
+
+ # generate feature-map
+
+ x_d0 = self.conv2(x_block4) # x_d0 : [2, 2048, 15, 20] 1/32 res
+ x_d1 = self.up1(x_d0, x_block3) # x_d1 : [2, 1024, 30, 40] 1/16 res
+ x_d2 = self.up2(x_d1, x_block2) # x_d2 : [2, 512, 60, 80] 1/8 res
+ x_d3 = self.up3(x_d2, x_block1) # x_d3: [2, 256, 120, 160] 1/4 res
+ x_d4 = self.up4(x_d3, x_block0) # x_d4: [2, 128, 240, 320] 1/2 res
+
+ # 1/8 res output
+ out_res8 = self.out_conv_res8(x_d2) # out_res8: [2, 4, 60, 80] 1/8 res output
+ out_res8 = norm_normalize(out_res8) # out_res8: [2, 4, 60, 80] 1/8 res output
+
+ ################################################################################################################
+ # out_res4
+ ################################################################################################################
+
+ if mode == 'train':
+ # upsampling ... out_res8: [2, 4, 60, 80] -> out_res8_res4: [2, 4, 120, 160]
+ out_res8_res4 = F.interpolate(out_res8, scale_factor=2, mode='bilinear', align_corners=True)
+ B, _, H, W = out_res8_res4.shape
+
+ # samples: [B, 1, N, 2]
+ point_coords_res4, rows_int, cols_int = sample_points(out_res8_res4.detach(), gt_norm_mask,
+ sampling_ratio=self.sampling_ratio,
+ beta=self.importance_ratio)
+
+ # output (needed for evaluation / visualization)
+ out_res4 = out_res8_res4
+
+ # grid_sample feature-map
+ feat_res4 = F.grid_sample(x_d2, point_coords_res4, mode='bilinear', align_corners=True) # (B, 512, 1, N)
+ init_pred = F.grid_sample(out_res8, point_coords_res4, mode='bilinear', align_corners=True) # (B, 4, 1, N)
+ feat_res4 = torch.cat([feat_res4, init_pred], dim=1) # (B, 512+4, 1, N)
+
+ # prediction (needed to compute loss)
+ samples_pred_res4 = self.out_conv_res4(feat_res4[:, :, 0, :]) # (B, 4, N)
+ samples_pred_res4 = norm_normalize(samples_pred_res4) # (B, 4, N) - normalized
+
+ for i in range(B):
+ out_res4[i, :, rows_int[i, :], cols_int[i, :]] = samples_pred_res4[i, :, :]
+
+ else:
+ # grid_sample feature-map
+ feat_map = F.interpolate(x_d2, scale_factor=2, mode='bilinear', align_corners=True)
+ init_pred = F.interpolate(out_res8, scale_factor=2, mode='bilinear', align_corners=True)
+ feat_map = torch.cat([feat_map, init_pred], dim=1) # (B, 512+4, H, W)
+ B, _, H, W = feat_map.shape
+
+ # try all pixels
+ out_res4 = self.out_conv_res4(feat_map.view(B, 512 + 4, -1)) # (B, 4, N)
+ out_res4 = norm_normalize(out_res4) # (B, 4, N) - normalized
+ out_res4 = out_res4.view(B, 4, H, W)
+ samples_pred_res4 = point_coords_res4 = None
+
+ ################################################################################################################
+ # out_res2
+ ################################################################################################################
+
+ if mode == 'train':
+
+ # upsampling ... out_res4: [2, 4, 120, 160] -> out_res4_res2: [2, 4, 240, 320]
+ out_res4_res2 = F.interpolate(out_res4, scale_factor=2, mode='bilinear', align_corners=True)
+ B, _, H, W = out_res4_res2.shape
+
+ # samples: [B, 1, N, 2]
+ point_coords_res2, rows_int, cols_int = sample_points(out_res4_res2.detach(), gt_norm_mask,
+ sampling_ratio=self.sampling_ratio,
+ beta=self.importance_ratio)
+
+ # output (needed for evaluation / visualization)
+ out_res2 = out_res4_res2
+
+ # grid_sample feature-map
+ feat_res2 = F.grid_sample(x_d3, point_coords_res2, mode='bilinear', align_corners=True) # (B, 256, 1, N)
+ init_pred = F.grid_sample(out_res4, point_coords_res2, mode='bilinear', align_corners=True) # (B, 4, 1, N)
+ feat_res2 = torch.cat([feat_res2, init_pred], dim=1) # (B, 256+4, 1, N)
+
+ # prediction (needed to compute loss)
+ samples_pred_res2 = self.out_conv_res2(feat_res2[:, :, 0, :]) # (B, 4, N)
+ samples_pred_res2 = norm_normalize(samples_pred_res2) # (B, 4, N) - normalized
+
+ for i in range(B):
+ out_res2[i, :, rows_int[i, :], cols_int[i, :]] = samples_pred_res2[i, :, :]
+
+ else:
+ # grid_sample feature-map
+ feat_map = F.interpolate(x_d3, scale_factor=2, mode='bilinear', align_corners=True)
+ init_pred = F.interpolate(out_res4, scale_factor=2, mode='bilinear', align_corners=True)
+ feat_map = torch.cat([feat_map, init_pred], dim=1) # (B, 512+4, H, W)
+ B, _, H, W = feat_map.shape
+
+ out_res2 = self.out_conv_res2(feat_map.view(B, 256 + 4, -1)) # (B, 4, N)
+ out_res2 = norm_normalize(out_res2) # (B, 4, N) - normalized
+ out_res2 = out_res2.view(B, 4, H, W)
+ samples_pred_res2 = point_coords_res2 = None
+
+ ################################################################################################################
+ # out_res1
+ ################################################################################################################
+
+ if mode == 'train':
+ # upsampling ... out_res4: [2, 4, 120, 160] -> out_res4_res2: [2, 4, 240, 320]
+ out_res2_res1 = F.interpolate(out_res2, scale_factor=2, mode='bilinear', align_corners=True)
+ B, _, H, W = out_res2_res1.shape
+
+ # samples: [B, 1, N, 2]
+ point_coords_res1, rows_int, cols_int = sample_points(out_res2_res1.detach(), gt_norm_mask,
+ sampling_ratio=self.sampling_ratio,
+ beta=self.importance_ratio)
+
+ # output (needed for evaluation / visualization)
+ out_res1 = out_res2_res1
+
+ # grid_sample feature-map
+ feat_res1 = F.grid_sample(x_d4, point_coords_res1, mode='bilinear', align_corners=True) # (B, 128, 1, N)
+ init_pred = F.grid_sample(out_res2, point_coords_res1, mode='bilinear', align_corners=True) # (B, 4, 1, N)
+ feat_res1 = torch.cat([feat_res1, init_pred], dim=1) # (B, 128+4, 1, N)
+
+ # prediction (needed to compute loss)
+ samples_pred_res1 = self.out_conv_res1(feat_res1[:, :, 0, :]) # (B, 4, N)
+ samples_pred_res1 = norm_normalize(samples_pred_res1) # (B, 4, N) - normalized
+
+ for i in range(B):
+ out_res1[i, :, rows_int[i, :], cols_int[i, :]] = samples_pred_res1[i, :, :]
+
+ else:
+ # grid_sample feature-map
+ feat_map = F.interpolate(x_d4, scale_factor=2, mode='bilinear', align_corners=True)
+ init_pred = F.interpolate(out_res2, scale_factor=2, mode='bilinear', align_corners=True)
+ feat_map = torch.cat([feat_map, init_pred], dim=1) # (B, 512+4, H, W)
+ B, _, H, W = feat_map.shape
+
+ out_res1 = self.out_conv_res1(feat_map.view(B, 128 + 4, -1)) # (B, 4, N)
+ out_res1 = norm_normalize(out_res1) # (B, 4, N) - normalized
+ out_res1 = out_res1.view(B, 4, H, W)
+ samples_pred_res1 = point_coords_res1 = None
+
+ return [out_res8, out_res4, out_res2, out_res1], \
+ [out_res8, samples_pred_res4, samples_pred_res2, samples_pred_res1], \
+ [None, point_coords_res4, point_coords_res2, point_coords_res1]
+
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/.gitignore b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/.gitignore
new file mode 100644
index 0000000000000000000000000000000000000000..f04e5fff91094d9b9c662bba977d762bf71516ac
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/.gitignore
@@ -0,0 +1,109 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+# Usually these files are written by a python script from a template
+# before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+.hypothesis/
+.pytest_cache/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# pyenv
+.python-version
+
+# celery beat schedule file
+celerybeat-schedule
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# pytorch stuff
+*.pth
+*.onnx
+*.pb
+
+trained_models/
+.fuse_hidden*
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/BENCHMARK.md b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/BENCHMARK.md
new file mode 100644
index 0000000000000000000000000000000000000000..6ead7171ce5a5bbd2702f6b5c825dc9808ba5658
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/BENCHMARK.md
@@ -0,0 +1,555 @@
+# Model Performance Benchmarks
+
+All benchmarks run as per:
+
+```
+python onnx_export.py --model mobilenetv3_100 ./mobilenetv3_100.onnx
+python onnx_optimize.py ./mobilenetv3_100.onnx --output mobilenetv3_100-opt.onnx
+python onnx_to_caffe.py ./mobilenetv3_100.onnx --c2-prefix mobilenetv3
+python onnx_to_caffe.py ./mobilenetv3_100-opt.onnx --c2-prefix mobilenetv3-opt
+python caffe2_benchmark.py --c2-init ./mobilenetv3.init.pb --c2-predict ./mobilenetv3.predict.pb
+python caffe2_benchmark.py --c2-init ./mobilenetv3-opt.init.pb --c2-predict ./mobilenetv3-opt.predict.pb
+```
+
+## EfficientNet-B0
+
+### Unoptimized
+```
+Main run finished. Milliseconds per iter: 49.2862. Iters per second: 20.2897
+Time per operator type:
+ 29.7378 ms. 60.5145%. Conv
+ 12.1785 ms. 24.7824%. Sigmoid
+ 3.62811 ms. 7.38297%. SpatialBN
+ 2.98444 ms. 6.07314%. Mul
+ 0.326902 ms. 0.665225%. AveragePool
+ 0.197317 ms. 0.401528%. FC
+ 0.0852877 ms. 0.173555%. Add
+ 0.0032607 ms. 0.00663532%. Squeeze
+ 49.1416 ms in Total
+FLOP per operator type:
+ 0.76907 GFLOP. 95.2696%. Conv
+ 0.0269508 GFLOP. 3.33857%. SpatialBN
+ 0.00846444 GFLOP. 1.04855%. Mul
+ 0.002561 GFLOP. 0.317248%. FC
+ 0.000210112 GFLOP. 0.0260279%. Add
+ 0.807256 GFLOP in Total
+Feature Memory Read per operator type:
+ 58.5253 MB. 43.0891%. Mul
+ 43.2015 MB. 31.807%. Conv
+ 27.2869 MB. 20.0899%. SpatialBN
+ 5.12912 MB. 3.77631%. FC
+ 1.6809 MB. 1.23756%. Add
+ 135.824 MB in Total
+Feature Memory Written per operator type:
+ 33.8578 MB. 38.1965%. Mul
+ 26.9881 MB. 30.4465%. Conv
+ 26.9508 MB. 30.4044%. SpatialBN
+ 0.840448 MB. 0.948147%. Add
+ 0.004 MB. 0.00451258%. FC
+ 88.6412 MB in Total
+Parameter Memory per operator type:
+ 15.8248 MB. 74.9391%. Conv
+ 5.124 MB. 24.265%. FC
+ 0.168064 MB. 0.795877%. SpatialBN
+ 0 MB. 0%. Add
+ 0 MB. 0%. Mul
+ 21.1168 MB in Total
+```
+### Optimized
+```
+Main run finished. Milliseconds per iter: 46.0838. Iters per second: 21.6996
+Time per operator type:
+ 29.776 ms. 65.002%. Conv
+ 12.2803 ms. 26.8084%. Sigmoid
+ 3.15073 ms. 6.87815%. Mul
+ 0.328651 ms. 0.717456%. AveragePool
+ 0.186237 ms. 0.406563%. FC
+ 0.0832429 ms. 0.181722%. Add
+ 0.0026184 ms. 0.00571606%. Squeeze
+ 45.8078 ms in Total
+FLOP per operator type:
+ 0.76907 GFLOP. 98.5601%. Conv
+ 0.00846444 GFLOP. 1.08476%. Mul
+ 0.002561 GFLOP. 0.328205%. FC
+ 0.000210112 GFLOP. 0.0269269%. Add
+ 0.780305 GFLOP in Total
+Feature Memory Read per operator type:
+ 58.5253 MB. 53.8803%. Mul
+ 43.2855 MB. 39.8501%. Conv
+ 5.12912 MB. 4.72204%. FC
+ 1.6809 MB. 1.54749%. Add
+ 108.621 MB in Total
+Feature Memory Written per operator type:
+ 33.8578 MB. 54.8834%. Mul
+ 26.9881 MB. 43.7477%. Conv
+ 0.840448 MB. 1.36237%. Add
+ 0.004 MB. 0.00648399%. FC
+ 61.6904 MB in Total
+Parameter Memory per operator type:
+ 15.8248 MB. 75.5403%. Conv
+ 5.124 MB. 24.4597%. FC
+ 0 MB. 0%. Add
+ 0 MB. 0%. Mul
+ 20.9488 MB in Total
+```
+
+## EfficientNet-B1
+### Optimized
+```
+Main run finished. Milliseconds per iter: 71.8102. Iters per second: 13.9256
+Time per operator type:
+ 45.7915 ms. 66.3206%. Conv
+ 17.8718 ms. 25.8841%. Sigmoid
+ 4.44132 ms. 6.43244%. Mul
+ 0.51001 ms. 0.738658%. AveragePool
+ 0.233283 ms. 0.337868%. Add
+ 0.194986 ms. 0.282402%. FC
+ 0.00268255 ms. 0.00388519%. Squeeze
+ 69.0456 ms in Total
+FLOP per operator type:
+ 1.37105 GFLOP. 98.7673%. Conv
+ 0.0138759 GFLOP. 0.99959%. Mul
+ 0.002561 GFLOP. 0.184489%. FC
+ 0.000674432 GFLOP. 0.0485847%. Add
+ 1.38816 GFLOP in Total
+Feature Memory Read per operator type:
+ 94.624 MB. 54.0789%. Mul
+ 69.8255 MB. 39.9062%. Conv
+ 5.39546 MB. 3.08357%. Add
+ 5.12912 MB. 2.93136%. FC
+ 174.974 MB in Total
+Feature Memory Written per operator type:
+ 55.5035 MB. 54.555%. Mul
+ 43.5333 MB. 42.7894%. Conv
+ 2.69773 MB. 2.65163%. Add
+ 0.004 MB. 0.00393165%. FC
+ 101.739 MB in Total
+Parameter Memory per operator type:
+ 25.7479 MB. 83.4024%. Conv
+ 5.124 MB. 16.5976%. FC
+ 0 MB. 0%. Add
+ 0 MB. 0%. Mul
+ 30.8719 MB in Total
+```
+
+## EfficientNet-B2
+### Optimized
+```
+Main run finished. Milliseconds per iter: 92.28. Iters per second: 10.8366
+Time per operator type:
+ 61.4627 ms. 67.5845%. Conv
+ 22.7458 ms. 25.0113%. Sigmoid
+ 5.59931 ms. 6.15701%. Mul
+ 0.642567 ms. 0.706568%. AveragePool
+ 0.272795 ms. 0.299965%. Add
+ 0.216178 ms. 0.237709%. FC
+ 0.00268895 ms. 0.00295677%. Squeeze
+ 90.942 ms in Total
+FLOP per operator type:
+ 1.98431 GFLOP. 98.9343%. Conv
+ 0.0177039 GFLOP. 0.882686%. Mul
+ 0.002817 GFLOP. 0.140451%. FC
+ 0.000853984 GFLOP. 0.0425782%. Add
+ 2.00568 GFLOP in Total
+Feature Memory Read per operator type:
+ 120.609 MB. 54.9637%. Mul
+ 86.3512 MB. 39.3519%. Conv
+ 6.83187 MB. 3.11341%. Add
+ 5.64163 MB. 2.571%. FC
+ 219.433 MB in Total
+Feature Memory Written per operator type:
+ 70.8155 MB. 54.6573%. Mul
+ 55.3273 MB. 42.7031%. Conv
+ 3.41594 MB. 2.63651%. Add
+ 0.004 MB. 0.00308731%. FC
+ 129.563 MB in Total
+Parameter Memory per operator type:
+ 30.4721 MB. 84.3913%. Conv
+ 5.636 MB. 15.6087%. FC
+ 0 MB. 0%. Add
+ 0 MB. 0%. Mul
+ 36.1081 MB in Total
+```
+
+## MixNet-M
+### Optimized
+```
+Main run finished. Milliseconds per iter: 63.1122. Iters per second: 15.8448
+Time per operator type:
+ 48.1139 ms. 75.2052%. Conv
+ 7.1341 ms. 11.1511%. Sigmoid
+ 2.63706 ms. 4.12189%. SpatialBN
+ 1.73186 ms. 2.70701%. Mul
+ 1.38707 ms. 2.16809%. Split
+ 1.29322 ms. 2.02139%. Concat
+ 1.00093 ms. 1.56452%. Relu
+ 0.235309 ms. 0.367803%. Add
+ 0.221579 ms. 0.346343%. FC
+ 0.219315 ms. 0.342803%. AveragePool
+ 0.00250145 ms. 0.00390993%. Squeeze
+ 63.9768 ms in Total
+FLOP per operator type:
+ 0.675273 GFLOP. 95.5827%. Conv
+ 0.0221072 GFLOP. 3.12921%. SpatialBN
+ 0.00538445 GFLOP. 0.762152%. Mul
+ 0.003073 GFLOP. 0.434973%. FC
+ 0.000642488 GFLOP. 0.0909421%. Add
+ 0 GFLOP. 0%. Concat
+ 0 GFLOP. 0%. Relu
+ 0.70648 GFLOP in Total
+Feature Memory Read per operator type:
+ 46.8424 MB. 30.502%. Conv
+ 36.8626 MB. 24.0036%. Mul
+ 22.3152 MB. 14.5309%. SpatialBN
+ 22.1074 MB. 14.3955%. Concat
+ 14.1496 MB. 9.21372%. Relu
+ 6.15414 MB. 4.00735%. FC
+ 5.1399 MB. 3.34692%. Add
+ 153.571 MB in Total
+Feature Memory Written per operator type:
+ 32.7672 MB. 28.4331%. Conv
+ 22.1072 MB. 19.1831%. Concat
+ 22.1072 MB. 19.1831%. SpatialBN
+ 21.5378 MB. 18.689%. Mul
+ 14.1496 MB. 12.2781%. Relu
+ 2.56995 MB. 2.23003%. Add
+ 0.004 MB. 0.00347092%. FC
+ 115.243 MB in Total
+Parameter Memory per operator type:
+ 13.7059 MB. 68.674%. Conv
+ 6.148 MB. 30.8049%. FC
+ 0.104 MB. 0.521097%. SpatialBN
+ 0 MB. 0%. Add
+ 0 MB. 0%. Concat
+ 0 MB. 0%. Mul
+ 0 MB. 0%. Relu
+ 19.9579 MB in Total
+```
+
+## TF MobileNet-V3 Large 1.0
+
+### Optimized
+```
+Main run finished. Milliseconds per iter: 22.0495. Iters per second: 45.3525
+Time per operator type:
+ 17.437 ms. 80.0087%. Conv
+ 1.27662 ms. 5.8577%. Add
+ 1.12759 ms. 5.17387%. Div
+ 0.701155 ms. 3.21721%. Mul
+ 0.562654 ms. 2.58171%. Relu
+ 0.431144 ms. 1.97828%. Clip
+ 0.156902 ms. 0.719936%. FC
+ 0.0996858 ms. 0.457402%. AveragePool
+ 0.00112455 ms. 0.00515993%. Flatten
+ 21.7939 ms in Total
+FLOP per operator type:
+ 0.43062 GFLOP. 98.1484%. Conv
+ 0.002561 GFLOP. 0.583713%. FC
+ 0.00210867 GFLOP. 0.480616%. Mul
+ 0.00193868 GFLOP. 0.441871%. Add
+ 0.00151532 GFLOP. 0.345377%. Div
+ 0 GFLOP. 0%. Relu
+ 0.438743 GFLOP in Total
+Feature Memory Read per operator type:
+ 34.7967 MB. 43.9391%. Conv
+ 14.496 MB. 18.3046%. Mul
+ 9.44828 MB. 11.9307%. Add
+ 9.26157 MB. 11.6949%. Relu
+ 6.0614 MB. 7.65395%. Div
+ 5.12912 MB. 6.47673%. FC
+ 79.193 MB in Total
+Feature Memory Written per operator type:
+ 17.6247 MB. 35.8656%. Conv
+ 9.26157 MB. 18.847%. Relu
+ 8.43469 MB. 17.1643%. Mul
+ 7.75472 MB. 15.7806%. Add
+ 6.06128 MB. 12.3345%. Div
+ 0.004 MB. 0.00813985%. FC
+ 49.1409 MB in Total
+Parameter Memory per operator type:
+ 16.6851 MB. 76.5052%. Conv
+ 5.124 MB. 23.4948%. FC
+ 0 MB. 0%. Add
+ 0 MB. 0%. Div
+ 0 MB. 0%. Mul
+ 0 MB. 0%. Relu
+ 21.8091 MB in Total
+```
+
+## MobileNet-V3 (RW)
+
+### Unoptimized
+```
+Main run finished. Milliseconds per iter: 24.8316. Iters per second: 40.2712
+Time per operator type:
+ 15.9266 ms. 69.2624%. Conv
+ 2.36551 ms. 10.2873%. SpatialBN
+ 1.39102 ms. 6.04936%. Add
+ 1.30327 ms. 5.66773%. Div
+ 0.737014 ms. 3.20517%. Mul
+ 0.639697 ms. 2.78195%. Relu
+ 0.375681 ms. 1.63378%. Clip
+ 0.153126 ms. 0.665921%. FC
+ 0.0993787 ms. 0.432184%. AveragePool
+ 0.0032632 ms. 0.0141912%. Squeeze
+ 22.9946 ms in Total
+FLOP per operator type:
+ 0.430616 GFLOP. 94.4041%. Conv
+ 0.0175992 GFLOP. 3.85829%. SpatialBN
+ 0.002561 GFLOP. 0.561449%. FC
+ 0.00210961 GFLOP. 0.46249%. Mul
+ 0.00173891 GFLOP. 0.381223%. Add
+ 0.00151626 GFLOP. 0.33241%. Div
+ 0 GFLOP. 0%. Relu
+ 0.456141 GFLOP in Total
+Feature Memory Read per operator type:
+ 34.7354 MB. 36.4363%. Conv
+ 17.7944 MB. 18.6658%. SpatialBN
+ 14.5035 MB. 15.2137%. Mul
+ 9.25778 MB. 9.71113%. Relu
+ 7.84641 MB. 8.23064%. Add
+ 6.06516 MB. 6.36216%. Div
+ 5.12912 MB. 5.38029%. FC
+ 95.3317 MB in Total
+Feature Memory Written per operator type:
+ 17.6246 MB. 26.7264%. Conv
+ 17.5992 MB. 26.6878%. SpatialBN
+ 9.25778 MB. 14.0387%. Relu
+ 8.43843 MB. 12.7962%. Mul
+ 6.95565 MB. 10.5477%. Add
+ 6.06502 MB. 9.19713%. Div
+ 0.004 MB. 0.00606568%. FC
+ 65.9447 MB in Total
+Parameter Memory per operator type:
+ 16.6778 MB. 76.1564%. Conv
+ 5.124 MB. 23.3979%. FC
+ 0.0976 MB. 0.445674%. SpatialBN
+ 0 MB. 0%. Add
+ 0 MB. 0%. Div
+ 0 MB. 0%. Mul
+ 0 MB. 0%. Relu
+ 21.8994 MB in Total
+
+```
+### Optimized
+
+```
+Main run finished. Milliseconds per iter: 22.0981. Iters per second: 45.2527
+Time per operator type:
+ 17.146 ms. 78.8965%. Conv
+ 1.38453 ms. 6.37084%. Add
+ 1.30991 ms. 6.02749%. Div
+ 0.685417 ms. 3.15391%. Mul
+ 0.532589 ms. 2.45068%. Relu
+ 0.418263 ms. 1.92461%. Clip
+ 0.15128 ms. 0.696106%. FC
+ 0.102065 ms. 0.469648%. AveragePool
+ 0.0022143 ms. 0.010189%. Squeeze
+ 21.7323 ms in Total
+FLOP per operator type:
+ 0.430616 GFLOP. 98.1927%. Conv
+ 0.002561 GFLOP. 0.583981%. FC
+ 0.00210961 GFLOP. 0.481051%. Mul
+ 0.00173891 GFLOP. 0.396522%. Add
+ 0.00151626 GFLOP. 0.34575%. Div
+ 0 GFLOP. 0%. Relu
+ 0.438542 GFLOP in Total
+Feature Memory Read per operator type:
+ 34.7842 MB. 44.833%. Conv
+ 14.5035 MB. 18.6934%. Mul
+ 9.25778 MB. 11.9323%. Relu
+ 7.84641 MB. 10.1132%. Add
+ 6.06516 MB. 7.81733%. Div
+ 5.12912 MB. 6.61087%. FC
+ 77.5861 MB in Total
+Feature Memory Written per operator type:
+ 17.6246 MB. 36.4556%. Conv
+ 9.25778 MB. 19.1492%. Relu
+ 8.43843 MB. 17.4544%. Mul
+ 6.95565 MB. 14.3874%. Add
+ 6.06502 MB. 12.5452%. Div
+ 0.004 MB. 0.00827378%. FC
+ 48.3455 MB in Total
+Parameter Memory per operator type:
+ 16.6778 MB. 76.4973%. Conv
+ 5.124 MB. 23.5027%. FC
+ 0 MB. 0%. Add
+ 0 MB. 0%. Div
+ 0 MB. 0%. Mul
+ 0 MB. 0%. Relu
+ 21.8018 MB in Total
+
+```
+
+## MnasNet-A1
+
+### Unoptimized
+```
+Main run finished. Milliseconds per iter: 30.0892. Iters per second: 33.2345
+Time per operator type:
+ 24.4656 ms. 79.0905%. Conv
+ 4.14958 ms. 13.4144%. SpatialBN
+ 1.60598 ms. 5.19169%. Relu
+ 0.295219 ms. 0.95436%. Mul
+ 0.187609 ms. 0.606486%. FC
+ 0.120556 ms. 0.389724%. AveragePool
+ 0.09036 ms. 0.292109%. Add
+ 0.015727 ms. 0.050841%. Sigmoid
+ 0.00306205 ms. 0.00989875%. Squeeze
+ 30.9337 ms in Total
+FLOP per operator type:
+ 0.620598 GFLOP. 95.6434%. Conv
+ 0.0248873 GFLOP. 3.8355%. SpatialBN
+ 0.002561 GFLOP. 0.394688%. FC
+ 0.000597408 GFLOP. 0.0920695%. Mul
+ 0.000222656 GFLOP. 0.0343146%. Add
+ 0 GFLOP. 0%. Relu
+ 0.648867 GFLOP in Total
+Feature Memory Read per operator type:
+ 35.5457 MB. 38.4109%. Conv
+ 25.1552 MB. 27.1829%. SpatialBN
+ 22.5235 MB. 24.339%. Relu
+ 5.12912 MB. 5.54256%. FC
+ 2.40586 MB. 2.59978%. Mul
+ 1.78125 MB. 1.92483%. Add
+ 92.5406 MB in Total
+Feature Memory Written per operator type:
+ 24.9042 MB. 32.9424%. Conv
+ 24.8873 MB. 32.92%. SpatialBN
+ 22.5235 MB. 29.7932%. Relu
+ 2.38963 MB. 3.16092%. Mul
+ 0.890624 MB. 1.17809%. Add
+ 0.004 MB. 0.00529106%. FC
+ 75.5993 MB in Total
+Parameter Memory per operator type:
+ 10.2732 MB. 66.1459%. Conv
+ 5.124 MB. 32.9917%. FC
+ 0.133952 MB. 0.86247%. SpatialBN
+ 0 MB. 0%. Add
+ 0 MB. 0%. Mul
+ 0 MB. 0%. Relu
+ 15.5312 MB in Total
+```
+
+### Optimized
+```
+Main run finished. Milliseconds per iter: 24.2367. Iters per second: 41.2597
+Time per operator type:
+ 22.0547 ms. 91.1375%. Conv
+ 1.49096 ms. 6.16116%. Relu
+ 0.253417 ms. 1.0472%. Mul
+ 0.18506 ms. 0.76473%. FC
+ 0.112942 ms. 0.466717%. AveragePool
+ 0.086769 ms. 0.358559%. Add
+ 0.0127889 ms. 0.0528479%. Sigmoid
+ 0.0027346 ms. 0.0113003%. Squeeze
+ 24.1994 ms in Total
+FLOP per operator type:
+ 0.620598 GFLOP. 99.4581%. Conv
+ 0.002561 GFLOP. 0.41043%. FC
+ 0.000597408 GFLOP. 0.0957417%. Mul
+ 0.000222656 GFLOP. 0.0356832%. Add
+ 0 GFLOP. 0%. Relu
+ 0.623979 GFLOP in Total
+Feature Memory Read per operator type:
+ 35.6127 MB. 52.7968%. Conv
+ 22.5235 MB. 33.3917%. Relu
+ 5.12912 MB. 7.60406%. FC
+ 2.40586 MB. 3.56675%. Mul
+ 1.78125 MB. 2.64075%. Add
+ 67.4524 MB in Total
+Feature Memory Written per operator type:
+ 24.9042 MB. 49.1092%. Conv
+ 22.5235 MB. 44.4145%. Relu
+ 2.38963 MB. 4.71216%. Mul
+ 0.890624 MB. 1.75624%. Add
+ 0.004 MB. 0.00788768%. FC
+ 50.712 MB in Total
+Parameter Memory per operator type:
+ 10.2732 MB. 66.7213%. Conv
+ 5.124 MB. 33.2787%. FC
+ 0 MB. 0%. Add
+ 0 MB. 0%. Mul
+ 0 MB. 0%. Relu
+ 15.3972 MB in Total
+```
+## MnasNet-B1
+
+### Unoptimized
+```
+Main run finished. Milliseconds per iter: 28.3109. Iters per second: 35.322
+Time per operator type:
+ 29.1121 ms. 83.3081%. Conv
+ 4.14959 ms. 11.8746%. SpatialBN
+ 1.35823 ms. 3.88675%. Relu
+ 0.186188 ms. 0.532802%. FC
+ 0.116244 ms. 0.332647%. Add
+ 0.018641 ms. 0.0533437%. AveragePool
+ 0.0040904 ms. 0.0117052%. Squeeze
+ 34.9451 ms in Total
+FLOP per operator type:
+ 0.626272 GFLOP. 96.2088%. Conv
+ 0.0218266 GFLOP. 3.35303%. SpatialBN
+ 0.002561 GFLOP. 0.393424%. FC
+ 0.000291648 GFLOP. 0.0448034%. Add
+ 0 GFLOP. 0%. Relu
+ 0.650951 GFLOP in Total
+Feature Memory Read per operator type:
+ 34.4354 MB. 41.3788%. Conv
+ 22.1299 MB. 26.5921%. SpatialBN
+ 19.1923 MB. 23.0622%. Relu
+ 5.12912 MB. 6.16333%. FC
+ 2.33318 MB. 2.80364%. Add
+ 83.2199 MB in Total
+Feature Memory Written per operator type:
+ 21.8266 MB. 34.0955%. Conv
+ 21.8266 MB. 34.0955%. SpatialBN
+ 19.1923 MB. 29.9805%. Relu
+ 1.16659 MB. 1.82234%. Add
+ 0.004 MB. 0.00624844%. FC
+ 64.016 MB in Total
+Parameter Memory per operator type:
+ 12.2576 MB. 69.9104%. Conv
+ 5.124 MB. 29.2245%. FC
+ 0.15168 MB. 0.865099%. SpatialBN
+ 0 MB. 0%. Add
+ 0 MB. 0%. Relu
+ 17.5332 MB in Total
+```
+
+### Optimized
+```
+Main run finished. Milliseconds per iter: 26.6364. Iters per second: 37.5426
+Time per operator type:
+ 24.9888 ms. 94.0962%. Conv
+ 1.26147 ms. 4.75011%. Relu
+ 0.176234 ms. 0.663619%. FC
+ 0.113309 ms. 0.426672%. Add
+ 0.0138708 ms. 0.0522311%. AveragePool
+ 0.00295685 ms. 0.0111341%. Squeeze
+ 26.5566 ms in Total
+FLOP per operator type:
+ 0.626272 GFLOP. 99.5466%. Conv
+ 0.002561 GFLOP. 0.407074%. FC
+ 0.000291648 GFLOP. 0.0463578%. Add
+ 0 GFLOP. 0%. Relu
+ 0.629124 GFLOP in Total
+Feature Memory Read per operator type:
+ 34.5112 MB. 56.4224%. Conv
+ 19.1923 MB. 31.3775%. Relu
+ 5.12912 MB. 8.3856%. FC
+ 2.33318 MB. 3.81452%. Add
+ 61.1658 MB in Total
+Feature Memory Written per operator type:
+ 21.8266 MB. 51.7346%. Conv
+ 19.1923 MB. 45.4908%. Relu
+ 1.16659 MB. 2.76513%. Add
+ 0.004 MB. 0.00948104%. FC
+ 42.1895 MB in Total
+Parameter Memory per operator type:
+ 12.2576 MB. 70.5205%. Conv
+ 5.124 MB. 29.4795%. FC
+ 0 MB. 0%. Add
+ 0 MB. 0%. Relu
+ 17.3816 MB in Total
+```
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/LICENSE b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/LICENSE
new file mode 100644
index 0000000000000000000000000000000000000000..80e7d15508202f3262a50db27f5198460d7f509f
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/LICENSE
@@ -0,0 +1,201 @@
+ Apache License
+ Version 2.0, January 2004
+ http://www.apache.org/licenses/
+
+ TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+ 1. Definitions.
+
+ "License" shall mean the terms and conditions for use, reproduction,
+ and distribution as defined by Sections 1 through 9 of this document.
+
+ "Licensor" shall mean the copyright owner or entity authorized by
+ the copyright owner that is granting the License.
+
+ "Legal Entity" shall mean the union of the acting entity and all
+ other entities that control, are controlled by, or are under common
+ control with that entity. For the purposes of this definition,
+ "control" means (i) the power, direct or indirect, to cause the
+ direction or management of such entity, whether by contract or
+ otherwise, or (ii) ownership of fifty percent (50%) or more of the
+ outstanding shares, or (iii) beneficial ownership of such entity.
+
+ "You" (or "Your") shall mean an individual or Legal Entity
+ exercising permissions granted by this License.
+
+ "Source" form shall mean the preferred form for making modifications,
+ including but not limited to software source code, documentation
+ source, and configuration files.
+
+ "Object" form shall mean any form resulting from mechanical
+ transformation or translation of a Source form, including but
+ not limited to compiled object code, generated documentation,
+ and conversions to other media types.
+
+ "Work" shall mean the work of authorship, whether in Source or
+ Object form, made available under the License, as indicated by a
+ copyright notice that is included in or attached to the work
+ (an example is provided in the Appendix below).
+
+ "Derivative Works" shall mean any work, whether in Source or Object
+ form, that is based on (or derived from) the Work and for which the
+ editorial revisions, annotations, elaborations, or other modifications
+ represent, as a whole, an original work of authorship. For the purposes
+ of this License, Derivative Works shall not include works that remain
+ separable from, or merely link (or bind by name) to the interfaces of,
+ the Work and Derivative Works thereof.
+
+ "Contribution" shall mean any work of authorship, including
+ the original version of the Work and any modifications or additions
+ to that Work or Derivative Works thereof, that is intentionally
+ submitted to Licensor for inclusion in the Work by the copyright owner
+ or by an individual or Legal Entity authorized to submit on behalf of
+ the copyright owner. For the purposes of this definition, "submitted"
+ means any form of electronic, verbal, or written communication sent
+ to the Licensor or its representatives, including but not limited to
+ communication on electronic mailing lists, source code control systems,
+ and issue tracking systems that are managed by, or on behalf of, the
+ Licensor for the purpose of discussing and improving the Work, but
+ excluding communication that is conspicuously marked or otherwise
+ designated in writing by the copyright owner as "Not a Contribution."
+
+ "Contributor" shall mean Licensor and any individual or Legal Entity
+ on behalf of whom a Contribution has been received by Licensor and
+ subsequently incorporated within the Work.
+
+ 2. Grant of Copyright License. Subject to the terms and conditions of
+ this License, each Contributor hereby grants to You a perpetual,
+ worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+ copyright license to reproduce, prepare Derivative Works of,
+ publicly display, publicly perform, sublicense, and distribute the
+ Work and such Derivative Works in Source or Object form.
+
+ 3. Grant of Patent License. Subject to the terms and conditions of
+ this License, each Contributor hereby grants to You a perpetual,
+ worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+ (except as stated in this section) patent license to make, have made,
+ use, offer to sell, sell, import, and otherwise transfer the Work,
+ where such license applies only to those patent claims licensable
+ by such Contributor that are necessarily infringed by their
+ Contribution(s) alone or by combination of their Contribution(s)
+ with the Work to which such Contribution(s) was submitted. If You
+ institute patent litigation against any entity (including a
+ cross-claim or counterclaim in a lawsuit) alleging that the Work
+ or a Contribution incorporated within the Work constitutes direct
+ or contributory patent infringement, then any patent licenses
+ granted to You under this License for that Work shall terminate
+ as of the date such litigation is filed.
+
+ 4. Redistribution. You may reproduce and distribute copies of the
+ Work or Derivative Works thereof in any medium, with or without
+ modifications, and in Source or Object form, provided that You
+ meet the following conditions:
+
+ (a) You must give any other recipients of the Work or
+ Derivative Works a copy of this License; and
+
+ (b) You must cause any modified files to carry prominent notices
+ stating that You changed the files; and
+
+ (c) You must retain, in the Source form of any Derivative Works
+ that You distribute, all copyright, patent, trademark, and
+ attribution notices from the Source form of the Work,
+ excluding those notices that do not pertain to any part of
+ the Derivative Works; and
+
+ (d) If the Work includes a "NOTICE" text file as part of its
+ distribution, then any Derivative Works that You distribute must
+ include a readable copy of the attribution notices contained
+ within such NOTICE file, excluding those notices that do not
+ pertain to any part of the Derivative Works, in at least one
+ of the following places: within a NOTICE text file distributed
+ as part of the Derivative Works; within the Source form or
+ documentation, if provided along with the Derivative Works; or,
+ within a display generated by the Derivative Works, if and
+ wherever such third-party notices normally appear. The contents
+ of the NOTICE file are for informational purposes only and
+ do not modify the License. You may add Your own attribution
+ notices within Derivative Works that You distribute, alongside
+ or as an addendum to the NOTICE text from the Work, provided
+ that such additional attribution notices cannot be construed
+ as modifying the License.
+
+ You may add Your own copyright statement to Your modifications and
+ may provide additional or different license terms and conditions
+ for use, reproduction, or distribution of Your modifications, or
+ for any such Derivative Works as a whole, provided Your use,
+ reproduction, and distribution of the Work otherwise complies with
+ the conditions stated in this License.
+
+ 5. Submission of Contributions. Unless You explicitly state otherwise,
+ any Contribution intentionally submitted for inclusion in the Work
+ by You to the Licensor shall be under the terms and conditions of
+ this License, without any additional terms or conditions.
+ Notwithstanding the above, nothing herein shall supersede or modify
+ the terms of any separate license agreement you may have executed
+ with Licensor regarding such Contributions.
+
+ 6. Trademarks. This License does not grant permission to use the trade
+ names, trademarks, service marks, or product names of the Licensor,
+ except as required for reasonable and customary use in describing the
+ origin of the Work and reproducing the content of the NOTICE file.
+
+ 7. Disclaimer of Warranty. Unless required by applicable law or
+ agreed to in writing, Licensor provides the Work (and each
+ Contributor provides its Contributions) on an "AS IS" BASIS,
+ WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+ implied, including, without limitation, any warranties or conditions
+ of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+ PARTICULAR PURPOSE. You are solely responsible for determining the
+ appropriateness of using or redistributing the Work and assume any
+ risks associated with Your exercise of permissions under this License.
+
+ 8. Limitation of Liability. In no event and under no legal theory,
+ whether in tort (including negligence), contract, or otherwise,
+ unless required by applicable law (such as deliberate and grossly
+ negligent acts) or agreed to in writing, shall any Contributor be
+ liable to You for damages, including any direct, indirect, special,
+ incidental, or consequential damages of any character arising as a
+ result of this License or out of the use or inability to use the
+ Work (including but not limited to damages for loss of goodwill,
+ work stoppage, computer failure or malfunction, or any and all
+ other commercial damages or losses), even if such Contributor
+ has been advised of the possibility of such damages.
+
+ 9. Accepting Warranty or Additional Liability. While redistributing
+ the Work or Derivative Works thereof, You may choose to offer,
+ and charge a fee for, acceptance of support, warranty, indemnity,
+ or other liability obligations and/or rights consistent with this
+ License. However, in accepting such obligations, You may act only
+ on Your own behalf and on Your sole responsibility, not on behalf
+ of any other Contributor, and only if You agree to indemnify,
+ defend, and hold each Contributor harmless for any liability
+ incurred by, or claims asserted against, such Contributor by reason
+ of your accepting any such warranty or additional liability.
+
+ END OF TERMS AND CONDITIONS
+
+ APPENDIX: How to apply the Apache License to your work.
+
+ To apply the Apache License to your work, attach the following
+ boilerplate notice, with the fields enclosed by brackets "{}"
+ replaced with your own identifying information. (Don't include
+ the brackets!) The text should be enclosed in the appropriate
+ comment syntax for the file format. We also recommend that a
+ file or class name and description of purpose be included on the
+ same "printed page" as the copyright notice for easier
+ identification within third-party archives.
+
+ Copyright 2020 Ross Wightman
+
+ Licensed under the Apache License, Version 2.0 (the "License");
+ you may not use this file except in compliance with the License.
+ You may obtain a copy of the License at
+
+ http://www.apache.org/licenses/LICENSE-2.0
+
+ Unless required by applicable law or agreed to in writing, software
+ distributed under the License is distributed on an "AS IS" BASIS,
+ WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ See the License for the specific language governing permissions and
+ limitations under the License.
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/README.md b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..463368280d6a5015060eb73d20fe6512f8e04c50
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/README.md
@@ -0,0 +1,323 @@
+# (Generic) EfficientNets for PyTorch
+
+A 'generic' implementation of EfficientNet, MixNet, MobileNetV3, etc. that covers most of the compute/parameter efficient architectures derived from the MobileNet V1/V2 block sequence, including those found via automated neural architecture search.
+
+All models are implemented by GenEfficientNet or MobileNetV3 classes, with string based architecture definitions to configure the block layouts (idea from [here](https://github.com/tensorflow/tpu/blob/master/models/official/mnasnet/mnasnet_models.py))
+
+## What's New
+
+### Aug 19, 2020
+* Add updated PyTorch trained EfficientNet-B3 weights trained by myself with `timm` (82.1 top-1)
+* Add PyTorch trained EfficientNet-Lite0 contributed by [@hal-314](https://github.com/hal-314) (75.5 top-1)
+* Update ONNX and Caffe2 export / utility scripts to work with latest PyTorch / ONNX
+* ONNX runtime based validation script added
+* activations (mostly) brought in sync with `timm` equivalents
+
+
+### April 5, 2020
+* Add some newly trained MobileNet-V2 models trained with latest h-params, rand augment. They compare quite favourably to EfficientNet-Lite
+ * 3.5M param MobileNet-V2 100 @ 73%
+ * 4.5M param MobileNet-V2 110d @ 75%
+ * 6.1M param MobileNet-V2 140 @ 76.5%
+ * 5.8M param MobileNet-V2 120d @ 77.3%
+
+### March 23, 2020
+ * Add EfficientNet-Lite models w/ weights ported from [Tensorflow TPU](https://github.com/tensorflow/tpu/tree/master/models/official/efficientnet/lite)
+ * Add PyTorch trained MobileNet-V3 Large weights with 75.77% top-1
+ * IMPORTANT CHANGE (if training from scratch) - weight init changed to better match Tensorflow impl, set `fix_group_fanout=False` in `initialize_weight_goog` for old behavior
+
+### Feb 12, 2020
+ * Add EfficientNet-L2 and B0-B7 NoisyStudent weights ported from [Tensorflow TPU](https://github.com/tensorflow/tpu/tree/master/models/official/efficientnet)
+ * Port new EfficientNet-B8 (RandAugment) weights from TF TPU, these are different than the B8 AdvProp, different input normalization.
+ * Add RandAugment PyTorch trained EfficientNet-ES (EdgeTPU-Small) weights with 78.1 top-1. Trained by [Andrew Lavin](https://github.com/andravin)
+
+### Jan 22, 2020
+ * Update weights for EfficientNet B0, B2, B3 and MixNet-XL with latest RandAugment trained weights. Trained with (https://github.com/rwightman/pytorch-image-models)
+ * Fix torchscript compatibility for PyTorch 1.4, add torchscript support for MixedConv2d using ModuleDict
+ * Test models, torchscript, onnx export with PyTorch 1.4 -- no issues
+
+### Nov 22, 2019
+ * New top-1 high! Ported official TF EfficientNet AdvProp (https://arxiv.org/abs/1911.09665) weights and B8 model spec. Created a new set of `ap` models since they use a different
+ preprocessing (Inception mean/std) from the original EfficientNet base/AA/RA weights.
+
+### Nov 15, 2019
+ * Ported official TF MobileNet-V3 float32 large/small/minimalistic weights
+ * Modifications to MobileNet-V3 model and components to support some additional config needed for differences between TF MobileNet-V3 and mine
+
+### Oct 30, 2019
+ * Many of the models will now work with torch.jit.script, MixNet being the biggest exception
+ * Improved interface for enabling torchscript or ONNX export compatible modes (via config)
+ * Add JIT optimized mem-efficient Swish/Mish autograd.fn in addition to memory-efficient autgrad.fn
+ * Activation factory to select best version of activation by name or override one globally
+ * Add pretrained checkpoint load helper that handles input conv and classifier changes
+
+### Oct 27, 2019
+ * Add CondConv EfficientNet variants ported from https://github.com/tensorflow/tpu/tree/master/models/official/efficientnet/condconv
+ * Add RandAug weights for TF EfficientNet B5 and B7 from https://github.com/tensorflow/tpu/tree/master/models/official/efficientnet
+ * Bring over MixNet-XL model and depth scaling algo from my pytorch-image-models code base
+ * Switch activations and global pooling to modules
+ * Add memory-efficient Swish/Mish impl
+ * Add as_sequential() method to all models and allow as an argument in entrypoint fns
+ * Move MobileNetV3 into own file since it has a different head
+ * Remove ChamNet, MobileNet V2/V1 since they will likely never be used here
+
+## Models
+
+Implemented models include:
+ * EfficientNet NoisyStudent (B0-B7, L2) (https://arxiv.org/abs/1911.04252)
+ * EfficientNet AdvProp (B0-B8) (https://arxiv.org/abs/1911.09665)
+ * EfficientNet (B0-B8) (https://arxiv.org/abs/1905.11946)
+ * EfficientNet-EdgeTPU (S, M, L) (https://ai.googleblog.com/2019/08/efficientnet-edgetpu-creating.html)
+ * EfficientNet-CondConv (https://arxiv.org/abs/1904.04971)
+ * EfficientNet-Lite (https://github.com/tensorflow/tpu/tree/master/models/official/efficientnet/lite)
+ * MixNet (https://arxiv.org/abs/1907.09595)
+ * MNASNet B1, A1 (Squeeze-Excite), and Small (https://arxiv.org/abs/1807.11626)
+ * MobileNet-V3 (https://arxiv.org/abs/1905.02244)
+ * FBNet-C (https://arxiv.org/abs/1812.03443)
+ * Single-Path NAS (https://arxiv.org/abs/1904.02877)
+
+I originally implemented and trained some these models with code [here](https://github.com/rwightman/pytorch-image-models), this repository contains just the GenEfficientNet models, validation, and associated ONNX/Caffe2 export code.
+
+## Pretrained
+
+I've managed to train several of the models to accuracies close to or above the originating papers and official impl. My training code is here: https://github.com/rwightman/pytorch-image-models
+
+
+|Model | Prec@1 (Err) | Prec@5 (Err) | Param#(M) | MAdds(M) | Image Scaling | Resolution | Crop |
+|---|---|---|---|---|---|---|---|
+| efficientnet_b3 | 82.240 (17.760) | 96.116 (3.884) | 12.23 | TBD | bicubic | 320 | 1.0 |
+| efficientnet_b3 | 82.076 (17.924) | 96.020 (3.980) | 12.23 | TBD | bicubic | 300 | 0.904 |
+| mixnet_xl | 81.074 (18.926) | 95.282 (4.718) | 11.90 | TBD | bicubic | 256 | 1.0 |
+| efficientnet_b2 | 80.612 (19.388) | 95.318 (4.682) | 9.1 | TBD | bicubic | 288 | 1.0 |
+| mixnet_xl | 80.476 (19.524) | 94.936 (5.064) | 11.90 | TBD | bicubic | 224 | 0.875 |
+| efficientnet_b2 | 80.288 (19.712) | 95.166 (4.834) | 9.1 | 1003 | bicubic | 260 | 0.890 |
+| mixnet_l | 78.976 (21.024 | 94.184 (5.816) | 7.33 | TBD | bicubic | 224 | 0.875 |
+| efficientnet_b1 | 78.692 (21.308) | 94.086 (5.914) | 7.8 | 694 | bicubic | 240 | 0.882 |
+| efficientnet_es | 78.066 (21.934) | 93.926 (6.074) | 5.44 | TBD | bicubic | 224 | 0.875 |
+| efficientnet_b0 | 77.698 (22.302) | 93.532 (6.468) | 5.3 | 390 | bicubic | 224 | 0.875 |
+| mobilenetv2_120d | 77.294 (22.706 | 93.502 (6.498) | 5.8 | TBD | bicubic | 224 | 0.875 |
+| mixnet_m | 77.256 (22.744) | 93.418 (6.582) | 5.01 | 353 | bicubic | 224 | 0.875 |
+| mobilenetv2_140 | 76.524 (23.476) | 92.990 (7.010) | 6.1 | TBD | bicubic | 224 | 0.875 |
+| mixnet_s | 75.988 (24.012) | 92.794 (7.206) | 4.13 | TBD | bicubic | 224 | 0.875 |
+| mobilenetv3_large_100 | 75.766 (24.234) | 92.542 (7.458) | 5.5 | TBD | bicubic | 224 | 0.875 |
+| mobilenetv3_rw | 75.634 (24.366) | 92.708 (7.292) | 5.5 | 219 | bicubic | 224 | 0.875 |
+| efficientnet_lite0 | 75.472 (24.528) | 92.520 (7.480) | 4.65 | TBD | bicubic | 224 | 0.875 |
+| mnasnet_a1 | 75.448 (24.552) | 92.604 (7.396) | 3.9 | 312 | bicubic | 224 | 0.875 |
+| fbnetc_100 | 75.124 (24.876) | 92.386 (7.614) | 5.6 | 385 | bilinear | 224 | 0.875 |
+| mobilenetv2_110d | 75.052 (24.948) | 92.180 (7.820) | 4.5 | TBD | bicubic | 224 | 0.875 |
+| mnasnet_b1 | 74.658 (25.342) | 92.114 (7.886) | 4.4 | 315 | bicubic | 224 | 0.875 |
+| spnasnet_100 | 74.084 (25.916) | 91.818 (8.182) | 4.4 | TBD | bilinear | 224 | 0.875 |
+| mobilenetv2_100 | 72.978 (27.022) | 91.016 (8.984) | 3.5 | TBD | bicubic | 224 | 0.875 |
+
+
+More pretrained models to come...
+
+
+## Ported Weights
+
+The weights ported from Tensorflow checkpoints for the EfficientNet models do pretty much match accuracy in Tensorflow once a SAME convolution padding equivalent is added, and the same crop factors, image scaling, etc (see table) are used via cmd line args.
+
+**IMPORTANT:**
+* Tensorflow ported weights for EfficientNet AdvProp (AP), EfficientNet EdgeTPU, EfficientNet-CondConv, EfficientNet-Lite, and MobileNet-V3 models use Inception style (0.5, 0.5, 0.5) for mean and std.
+* Enabling the Tensorflow preprocessing pipeline with `--tf-preprocessing` at validation time will improve scores by 0.1-0.5%, very close to original TF impl.
+
+To run validation for tf_efficientnet_b5:
+`python validate.py /path/to/imagenet/validation/ --model tf_efficientnet_b5 -b 64 --img-size 456 --crop-pct 0.934 --interpolation bicubic`
+
+To run validation w/ TF preprocessing for tf_efficientnet_b5:
+`python validate.py /path/to/imagenet/validation/ --model tf_efficientnet_b5 -b 64 --img-size 456 --tf-preprocessing`
+
+To run validation for a model with Inception preprocessing, ie EfficientNet-B8 AdvProp:
+`python validate.py /path/to/imagenet/validation/ --model tf_efficientnet_b8_ap -b 48 --num-gpu 2 --img-size 672 --crop-pct 0.954 --mean 0.5 --std 0.5`
+
+|Model | Prec@1 (Err) | Prec@5 (Err) | Param # | Image Scaling | Image Size | Crop |
+|---|---|---|---|---|---|---|
+| tf_efficientnet_l2_ns *tfp | 88.352 (11.648) | 98.652 (1.348) | 480 | bicubic | 800 | N/A |
+| tf_efficientnet_l2_ns | TBD | TBD | 480 | bicubic | 800 | 0.961 |
+| tf_efficientnet_l2_ns_475 | 88.234 (11.766) | 98.546 (1.454) | 480 | bicubic | 475 | 0.936 |
+| tf_efficientnet_l2_ns_475 *tfp | 88.172 (11.828) | 98.566 (1.434) | 480 | bicubic | 475 | N/A |
+| tf_efficientnet_b7_ns *tfp | 86.844 (13.156) | 98.084 (1.916) | 66.35 | bicubic | 600 | N/A |
+| tf_efficientnet_b7_ns | 86.840 (13.160) | 98.094 (1.906) | 66.35 | bicubic | 600 | N/A |
+| tf_efficientnet_b6_ns | 86.452 (13.548) | 97.882 (2.118) | 43.04 | bicubic | 528 | N/A |
+| tf_efficientnet_b6_ns *tfp | 86.444 (13.556) | 97.880 (2.120) | 43.04 | bicubic | 528 | N/A |
+| tf_efficientnet_b5_ns *tfp | 86.064 (13.936) | 97.746 (2.254) | 30.39 | bicubic | 456 | N/A |
+| tf_efficientnet_b5_ns | 86.088 (13.912) | 97.752 (2.248) | 30.39 | bicubic | 456 | N/A |
+| tf_efficientnet_b8_ap *tfp | 85.436 (14.564) | 97.272 (2.728) | 87.4 | bicubic | 672 | N/A |
+| tf_efficientnet_b8 *tfp | 85.384 (14.616) | 97.394 (2.606) | 87.4 | bicubic | 672 | N/A |
+| tf_efficientnet_b8 | 85.370 (14.630) | 97.390 (2.610) | 87.4 | bicubic | 672 | 0.954 |
+| tf_efficientnet_b8_ap | 85.368 (14.632) | 97.294 (2.706) | 87.4 | bicubic | 672 | 0.954 |
+| tf_efficientnet_b4_ns *tfp | 85.298 (14.702) | 97.504 (2.496) | 19.34 | bicubic | 380 | N/A |
+| tf_efficientnet_b4_ns | 85.162 (14.838) | 97.470 (2.530) | 19.34 | bicubic | 380 | 0.922 |
+| tf_efficientnet_b7_ap *tfp | 85.154 (14.846) | 97.244 (2.756) | 66.35 | bicubic | 600 | N/A |
+| tf_efficientnet_b7_ap | 85.118 (14.882) | 97.252 (2.748) | 66.35 | bicubic | 600 | 0.949 |
+| tf_efficientnet_b7 *tfp | 84.940 (15.060) | 97.214 (2.786) | 66.35 | bicubic | 600 | N/A |
+| tf_efficientnet_b7 | 84.932 (15.068) | 97.208 (2.792) | 66.35 | bicubic | 600 | 0.949 |
+| tf_efficientnet_b6_ap | 84.786 (15.214) | 97.138 (2.862) | 43.04 | bicubic | 528 | 0.942 |
+| tf_efficientnet_b6_ap *tfp | 84.760 (15.240) | 97.124 (2.876) | 43.04 | bicubic | 528 | N/A |
+| tf_efficientnet_b5_ap *tfp | 84.276 (15.724) | 96.932 (3.068) | 30.39 | bicubic | 456 | N/A |
+| tf_efficientnet_b5_ap | 84.254 (15.746) | 96.976 (3.024) | 30.39 | bicubic | 456 | 0.934 |
+| tf_efficientnet_b6 *tfp | 84.140 (15.860) | 96.852 (3.148) | 43.04 | bicubic | 528 | N/A |
+| tf_efficientnet_b6 | 84.110 (15.890) | 96.886 (3.114) | 43.04 | bicubic | 528 | 0.942 |
+| tf_efficientnet_b3_ns *tfp | 84.054 (15.946) | 96.918 (3.082) | 12.23 | bicubic | 300 | N/A |
+| tf_efficientnet_b3_ns | 84.048 (15.952) | 96.910 (3.090) | 12.23 | bicubic | 300 | .904 |
+| tf_efficientnet_b5 *tfp | 83.822 (16.178) | 96.756 (3.244) | 30.39 | bicubic | 456 | N/A |
+| tf_efficientnet_b5 | 83.812 (16.188) | 96.748 (3.252) | 30.39 | bicubic | 456 | 0.934 |
+| tf_efficientnet_b4_ap *tfp | 83.278 (16.722) | 96.376 (3.624) | 19.34 | bicubic | 380 | N/A |
+| tf_efficientnet_b4_ap | 83.248 (16.752) | 96.388 (3.612) | 19.34 | bicubic | 380 | 0.922 |
+| tf_efficientnet_b4 | 83.022 (16.978) | 96.300 (3.700) | 19.34 | bicubic | 380 | 0.922 |
+| tf_efficientnet_b4 *tfp | 82.948 (17.052) | 96.308 (3.692) | 19.34 | bicubic | 380 | N/A |
+| tf_efficientnet_b2_ns *tfp | 82.436 (17.564) | 96.268 (3.732) | 9.11 | bicubic | 260 | N/A |
+| tf_efficientnet_b2_ns | 82.380 (17.620) | 96.248 (3.752) | 9.11 | bicubic | 260 | 0.89 |
+| tf_efficientnet_b3_ap *tfp | 81.882 (18.118) | 95.662 (4.338) | 12.23 | bicubic | 300 | N/A |
+| tf_efficientnet_b3_ap | 81.828 (18.172) | 95.624 (4.376) | 12.23 | bicubic | 300 | 0.904 |
+| tf_efficientnet_b3 | 81.636 (18.364) | 95.718 (4.282) | 12.23 | bicubic | 300 | 0.904 |
+| tf_efficientnet_b3 *tfp | 81.576 (18.424) | 95.662 (4.338) | 12.23 | bicubic | 300 | N/A |
+| tf_efficientnet_lite4 | 81.528 (18.472) | 95.668 (4.332) | 13.00 | bilinear | 380 | 0.92 |
+| tf_efficientnet_b1_ns *tfp | 81.514 (18.486) | 95.776 (4.224) | 7.79 | bicubic | 240 | N/A |
+| tf_efficientnet_lite4 *tfp | 81.502 (18.498) | 95.676 (4.324) | 13.00 | bilinear | 380 | N/A |
+| tf_efficientnet_b1_ns | 81.388 (18.612) | 95.738 (4.262) | 7.79 | bicubic | 240 | 0.88 |
+| tf_efficientnet_el | 80.534 (19.466) | 95.190 (4.810) | 10.59 | bicubic | 300 | 0.904 |
+| tf_efficientnet_el *tfp | 80.476 (19.524) | 95.200 (4.800) | 10.59 | bicubic | 300 | N/A |
+| tf_efficientnet_b2_ap *tfp | 80.420 (19.580) | 95.040 (4.960) | 9.11 | bicubic | 260 | N/A |
+| tf_efficientnet_b2_ap | 80.306 (19.694) | 95.028 (4.972) | 9.11 | bicubic | 260 | 0.890 |
+| tf_efficientnet_b2 *tfp | 80.188 (19.812) | 94.974 (5.026) | 9.11 | bicubic | 260 | N/A |
+| tf_efficientnet_b2 | 80.086 (19.914) | 94.908 (5.092) | 9.11 | bicubic | 260 | 0.890 |
+| tf_efficientnet_lite3 | 79.812 (20.188) | 94.914 (5.086) | 8.20 | bilinear | 300 | 0.904 |
+| tf_efficientnet_lite3 *tfp | 79.734 (20.266) | 94.838 (5.162) | 8.20 | bilinear | 300 | N/A |
+| tf_efficientnet_b1_ap *tfp | 79.532 (20.468) | 94.378 (5.622) | 7.79 | bicubic | 240 | N/A |
+| tf_efficientnet_cc_b1_8e *tfp | 79.464 (20.536)| 94.492 (5.508) | 39.7 | bicubic | 240 | 0.88 |
+| tf_efficientnet_cc_b1_8e | 79.298 (20.702) | 94.364 (5.636) | 39.7 | bicubic | 240 | 0.88 |
+| tf_efficientnet_b1_ap | 79.278 (20.722) | 94.308 (5.692) | 7.79 | bicubic | 240 | 0.88 |
+| tf_efficientnet_b1 *tfp | 79.172 (20.828) | 94.450 (5.550) | 7.79 | bicubic | 240 | N/A |
+| tf_efficientnet_em *tfp | 78.958 (21.042) | 94.458 (5.542) | 6.90 | bicubic | 240 | N/A |
+| tf_efficientnet_b0_ns *tfp | 78.806 (21.194) | 94.496 (5.504) | 5.29 | bicubic | 224 | N/A |
+| tf_mixnet_l *tfp | 78.846 (21.154) | 94.212 (5.788) | 7.33 | bilinear | 224 | N/A |
+| tf_efficientnet_b1 | 78.826 (21.174) | 94.198 (5.802) | 7.79 | bicubic | 240 | 0.88 |
+| tf_mixnet_l | 78.770 (21.230) | 94.004 (5.996) | 7.33 | bicubic | 224 | 0.875 |
+| tf_efficientnet_em | 78.742 (21.258) | 94.332 (5.668) | 6.90 | bicubic | 240 | 0.875 |
+| tf_efficientnet_b0_ns | 78.658 (21.342) | 94.376 (5.624) | 5.29 | bicubic | 224 | 0.875 |
+| tf_efficientnet_cc_b0_8e *tfp | 78.314 (21.686) | 93.790 (6.210) | 24.0 | bicubic | 224 | 0.875 |
+| tf_efficientnet_cc_b0_8e | 77.908 (22.092) | 93.656 (6.344) | 24.0 | bicubic | 224 | 0.875 |
+| tf_efficientnet_cc_b0_4e *tfp | 77.746 (22.254) | 93.552 (6.448) | 13.3 | bicubic | 224 | 0.875 |
+| tf_efficientnet_cc_b0_4e | 77.304 (22.696) | 93.332 (6.668) | 13.3 | bicubic | 224 | 0.875 |
+| tf_efficientnet_es *tfp | 77.616 (22.384) | 93.750 (6.250) | 5.44 | bicubic | 224 | N/A |
+| tf_efficientnet_lite2 *tfp | 77.544 (22.456) | 93.800 (6.200) | 6.09 | bilinear | 260 | N/A |
+| tf_efficientnet_lite2 | 77.460 (22.540) | 93.746 (6.254) | 6.09 | bicubic | 260 | 0.89 |
+| tf_efficientnet_b0_ap *tfp | 77.514 (22.486) | 93.576 (6.424) | 5.29 | bicubic | 224 | N/A |
+| tf_efficientnet_es | 77.264 (22.736) | 93.600 (6.400) | 5.44 | bicubic | 224 | N/A |
+| tf_efficientnet_b0 *tfp | 77.258 (22.742) | 93.478 (6.522) | 5.29 | bicubic | 224 | N/A |
+| tf_efficientnet_b0_ap | 77.084 (22.916) | 93.254 (6.746) | 5.29 | bicubic | 224 | 0.875 |
+| tf_mixnet_m *tfp | 77.072 (22.928) | 93.368 (6.632) | 5.01 | bilinear | 224 | N/A |
+| tf_mixnet_m | 76.950 (23.050) | 93.156 (6.844) | 5.01 | bicubic | 224 | 0.875 |
+| tf_efficientnet_b0 | 76.848 (23.152) | 93.228 (6.772) | 5.29 | bicubic | 224 | 0.875 |
+| tf_efficientnet_lite1 *tfp | 76.764 (23.236) | 93.326 (6.674) | 5.42 | bilinear | 240 | N/A |
+| tf_efficientnet_lite1 | 76.638 (23.362) | 93.232 (6.768) | 5.42 | bicubic | 240 | 0.882 |
+| tf_mixnet_s *tfp | 75.800 (24.200) | 92.788 (7.212) | 4.13 | bilinear | 224 | N/A |
+| tf_mobilenetv3_large_100 *tfp | 75.768 (24.232) | 92.710 (7.290) | 5.48 | bilinear | 224 | N/A |
+| tf_mixnet_s | 75.648 (24.352) | 92.636 (7.364) | 4.13 | bicubic | 224 | 0.875 |
+| tf_mobilenetv3_large_100 | 75.516 (24.484) | 92.600 (7.400) | 5.48 | bilinear | 224 | 0.875 |
+| tf_efficientnet_lite0 *tfp | 75.074 (24.926) | 92.314 (7.686) | 4.65 | bilinear | 224 | N/A |
+| tf_efficientnet_lite0 | 74.842 (25.158) | 92.170 (7.830) | 4.65 | bicubic | 224 | 0.875 |
+| tf_mobilenetv3_large_075 *tfp | 73.730 (26.270) | 91.616 (8.384) | 3.99 | bilinear | 224 |N/A |
+| tf_mobilenetv3_large_075 | 73.442 (26.558) | 91.352 (8.648) | 3.99 | bilinear | 224 | 0.875 |
+| tf_mobilenetv3_large_minimal_100 *tfp | 72.678 (27.322) | 90.860 (9.140) | 3.92 | bilinear | 224 | N/A |
+| tf_mobilenetv3_large_minimal_100 | 72.244 (27.756) | 90.636 (9.364) | 3.92 | bilinear | 224 | 0.875 |
+| tf_mobilenetv3_small_100 *tfp | 67.918 (32.082) | 87.958 (12.042 | 2.54 | bilinear | 224 | N/A |
+| tf_mobilenetv3_small_100 | 67.918 (32.082) | 87.662 (12.338) | 2.54 | bilinear | 224 | 0.875 |
+| tf_mobilenetv3_small_075 *tfp | 66.142 (33.858) | 86.498 (13.502) | 2.04 | bilinear | 224 | N/A |
+| tf_mobilenetv3_small_075 | 65.718 (34.282) | 86.136 (13.864) | 2.04 | bilinear | 224 | 0.875 |
+| tf_mobilenetv3_small_minimal_100 *tfp | 63.378 (36.622) | 84.802 (15.198) | 2.04 | bilinear | 224 | N/A |
+| tf_mobilenetv3_small_minimal_100 | 62.898 (37.102) | 84.230 (15.770) | 2.04 | bilinear | 224 | 0.875 |
+
+
+*tfp models validated with `tf-preprocessing` pipeline
+
+Google tf and tflite weights ported from official Tensorflow repositories
+* https://github.com/tensorflow/tpu/tree/master/models/official/mnasnet
+* https://github.com/tensorflow/tpu/tree/master/models/official/efficientnet
+* https://github.com/tensorflow/models/tree/master/research/slim/nets/mobilenet
+
+## Usage
+
+### Environment
+
+All development and testing has been done in Conda Python 3 environments on Linux x86-64 systems, specifically Python 3.6.x, 3.7.x, 3.8.x.
+
+Users have reported that a Python 3 Anaconda install in Windows works. I have not verified this myself.
+
+PyTorch versions 1.4, 1.5, 1.6 have been tested with this code.
+
+I've tried to keep the dependencies minimal, the setup is as per the PyTorch default install instructions for Conda:
+```
+conda create -n torch-env
+conda activate torch-env
+conda install -c pytorch pytorch torchvision cudatoolkit=10.2
+```
+
+### PyTorch Hub
+
+Models can be accessed via the PyTorch Hub API
+
+```
+>>> torch.hub.list('rwightman/gen-efficientnet-pytorch')
+['efficientnet_b0', ...]
+>>> model = torch.hub.load('rwightman/gen-efficientnet-pytorch', 'efficientnet_b0', pretrained=True)
+>>> model.eval()
+>>> output = model(torch.randn(1,3,224,224))
+```
+
+### Pip
+This package can be installed via pip.
+
+Install (after conda env/install):
+```
+pip install geffnet
+```
+
+Eval use:
+```
+>>> import geffnet
+>>> m = geffnet.create_model('mobilenetv3_large_100', pretrained=True)
+>>> m.eval()
+```
+
+Train use:
+```
+>>> import geffnet
+>>> # models can also be created by using the entrypoint directly
+>>> m = geffnet.efficientnet_b2(pretrained=True, drop_rate=0.25, drop_connect_rate=0.2)
+>>> m.train()
+```
+
+Create in a nn.Sequential container, for fast.ai, etc:
+```
+>>> import geffnet
+>>> m = geffnet.mixnet_l(pretrained=True, drop_rate=0.25, drop_connect_rate=0.2, as_sequential=True)
+```
+
+### Exporting
+
+Scripts are included to
+* export models to ONNX (`onnx_export.py`)
+* optimized ONNX graph (`onnx_optimize.py` or `onnx_validate.py` w/ `--onnx-output-opt` arg)
+* validate with ONNX runtime (`onnx_validate.py`)
+* convert ONNX model to Caffe2 (`onnx_to_caffe.py`)
+* validate in Caffe2 (`caffe2_validate.py`)
+* benchmark in Caffe2 w/ FLOPs, parameters output (`caffe2_benchmark.py`)
+
+As an example, to export the MobileNet-V3 pretrained model and then run an Imagenet validation:
+```
+python onnx_export.py --model mobilenetv3_large_100 ./mobilenetv3_100.onnx
+python onnx_validate.py /imagenet/validation/ --onnx-input ./mobilenetv3_100.onnx
+```
+
+These scripts were tested to be working as of PyTorch 1.6 and ONNX 1.7 w/ ONNX runtime 1.4. Caffe2 compatible
+export now requires additional args mentioned in the export script (not needed in earlier versions).
+
+#### Export Notes
+1. The TF ported weights with the 'SAME' conv padding activated cannot be exported to ONNX unless `_EXPORTABLE` flag in `config.py` is set to True. Use `config.set_exportable(True)` as in the `onnx_export.py` script.
+2. TF ported models with 'SAME' padding will have the padding fixed at export time to the resolution used for export. Even though dynamic padding is supported in opset >= 11, I can't get it working.
+3. ONNX optimize facility doesn't work reliably in PyTorch 1.6 / ONNX 1.7. Fortunately, the onnxruntime based inference is working very well now and includes on the fly optimization.
+3. ONNX / Caffe2 export/import frequently breaks with different PyTorch and ONNX version releases. Please check their respective issue trackers before filing issues here.
+
+
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/caffe2_benchmark.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/caffe2_benchmark.py
new file mode 100644
index 0000000000000000000000000000000000000000..93f28a1e63d9f7287ca02997c7991fe66dd0aeb9
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/caffe2_benchmark.py
@@ -0,0 +1,65 @@
+""" Caffe2 validation script
+
+This script runs Caffe2 benchmark on exported ONNX model.
+It is a useful tool for reporting model FLOPS.
+
+Copyright 2020 Ross Wightman
+"""
+import argparse
+from caffe2.python import core, workspace, model_helper
+from caffe2.proto import caffe2_pb2
+
+
+parser = argparse.ArgumentParser(description='Caffe2 Model Benchmark')
+parser.add_argument('--c2-prefix', default='', type=str, metavar='NAME',
+ help='caffe2 model pb name prefix')
+parser.add_argument('--c2-init', default='', type=str, metavar='PATH',
+ help='caffe2 model init .pb')
+parser.add_argument('--c2-predict', default='', type=str, metavar='PATH',
+ help='caffe2 model predict .pb')
+parser.add_argument('-b', '--batch-size', default=1, type=int,
+ metavar='N', help='mini-batch size (default: 1)')
+parser.add_argument('--img-size', default=224, type=int,
+ metavar='N', help='Input image dimension, uses model default if empty')
+
+
+def main():
+ args = parser.parse_args()
+ args.gpu_id = 0
+ if args.c2_prefix:
+ args.c2_init = args.c2_prefix + '.init.pb'
+ args.c2_predict = args.c2_prefix + '.predict.pb'
+
+ model = model_helper.ModelHelper(name="le_net", init_params=False)
+
+ # Bring in the init net from init_net.pb
+ init_net_proto = caffe2_pb2.NetDef()
+ with open(args.c2_init, "rb") as f:
+ init_net_proto.ParseFromString(f.read())
+ model.param_init_net = core.Net(init_net_proto)
+
+ # bring in the predict net from predict_net.pb
+ predict_net_proto = caffe2_pb2.NetDef()
+ with open(args.c2_predict, "rb") as f:
+ predict_net_proto.ParseFromString(f.read())
+ model.net = core.Net(predict_net_proto)
+
+ # CUDA performance not impressive
+ #device_opts = core.DeviceOption(caffe2_pb2.PROTO_CUDA, args.gpu_id)
+ #model.net.RunAllOnGPU(gpu_id=args.gpu_id, use_cudnn=True)
+ #model.param_init_net.RunAllOnGPU(gpu_id=args.gpu_id, use_cudnn=True)
+
+ input_blob = model.net.external_inputs[0]
+ model.param_init_net.GaussianFill(
+ [],
+ input_blob.GetUnscopedName(),
+ shape=(args.batch_size, 3, args.img_size, args.img_size),
+ mean=0.0,
+ std=1.0)
+ workspace.RunNetOnce(model.param_init_net)
+ workspace.CreateNet(model.net, overwrite=True)
+ workspace.BenchmarkNet(model.net.Proto().name, 5, 20, True)
+
+
+if __name__ == '__main__':
+ main()
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/caffe2_validate.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/caffe2_validate.py
new file mode 100644
index 0000000000000000000000000000000000000000..7cfaab38c095663fe32e4addbdf06b57bcb53614
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/caffe2_validate.py
@@ -0,0 +1,138 @@
+""" Caffe2 validation script
+
+This script is created to verify exported ONNX models running in Caffe2
+It utilizes the same PyTorch dataloader/processing pipeline for a
+fair comparison against the originals.
+
+Copyright 2020 Ross Wightman
+"""
+import argparse
+import numpy as np
+from caffe2.python import core, workspace, model_helper
+from caffe2.proto import caffe2_pb2
+from data import create_loader, resolve_data_config, Dataset
+from utils import AverageMeter
+import time
+
+parser = argparse.ArgumentParser(description='Caffe2 ImageNet Validation')
+parser.add_argument('data', metavar='DIR',
+ help='path to dataset')
+parser.add_argument('--c2-prefix', default='', type=str, metavar='NAME',
+ help='caffe2 model pb name prefix')
+parser.add_argument('--c2-init', default='', type=str, metavar='PATH',
+ help='caffe2 model init .pb')
+parser.add_argument('--c2-predict', default='', type=str, metavar='PATH',
+ help='caffe2 model predict .pb')
+parser.add_argument('-j', '--workers', default=2, type=int, metavar='N',
+ help='number of data loading workers (default: 2)')
+parser.add_argument('-b', '--batch-size', default=256, type=int,
+ metavar='N', help='mini-batch size (default: 256)')
+parser.add_argument('--img-size', default=None, type=int,
+ metavar='N', help='Input image dimension, uses model default if empty')
+parser.add_argument('--mean', type=float, nargs='+', default=None, metavar='MEAN',
+ help='Override mean pixel value of dataset')
+parser.add_argument('--std', type=float, nargs='+', default=None, metavar='STD',
+ help='Override std deviation of of dataset')
+parser.add_argument('--crop-pct', type=float, default=None, metavar='PCT',
+ help='Override default crop pct of 0.875')
+parser.add_argument('--interpolation', default='', type=str, metavar='NAME',
+ help='Image resize interpolation type (overrides model)')
+parser.add_argument('--tf-preprocessing', dest='tf_preprocessing', action='store_true',
+ help='use tensorflow mnasnet preporcessing')
+parser.add_argument('--print-freq', '-p', default=10, type=int,
+ metavar='N', help='print frequency (default: 10)')
+
+
+def main():
+ args = parser.parse_args()
+ args.gpu_id = 0
+ if args.c2_prefix:
+ args.c2_init = args.c2_prefix + '.init.pb'
+ args.c2_predict = args.c2_prefix + '.predict.pb'
+
+ model = model_helper.ModelHelper(name="validation_net", init_params=False)
+
+ # Bring in the init net from init_net.pb
+ init_net_proto = caffe2_pb2.NetDef()
+ with open(args.c2_init, "rb") as f:
+ init_net_proto.ParseFromString(f.read())
+ model.param_init_net = core.Net(init_net_proto)
+
+ # bring in the predict net from predict_net.pb
+ predict_net_proto = caffe2_pb2.NetDef()
+ with open(args.c2_predict, "rb") as f:
+ predict_net_proto.ParseFromString(f.read())
+ model.net = core.Net(predict_net_proto)
+
+ data_config = resolve_data_config(None, args)
+ loader = create_loader(
+ Dataset(args.data, load_bytes=args.tf_preprocessing),
+ input_size=data_config['input_size'],
+ batch_size=args.batch_size,
+ use_prefetcher=False,
+ interpolation=data_config['interpolation'],
+ mean=data_config['mean'],
+ std=data_config['std'],
+ num_workers=args.workers,
+ crop_pct=data_config['crop_pct'],
+ tensorflow_preprocessing=args.tf_preprocessing)
+
+ # this is so obvious, wonderful interface
+ input_blob = model.net.external_inputs[0]
+ output_blob = model.net.external_outputs[0]
+
+ if True:
+ device_opts = None
+ else:
+ # CUDA is crashing, no idea why, awesome error message, give it a try for kicks
+ device_opts = core.DeviceOption(caffe2_pb2.PROTO_CUDA, args.gpu_id)
+ model.net.RunAllOnGPU(gpu_id=args.gpu_id, use_cudnn=True)
+ model.param_init_net.RunAllOnGPU(gpu_id=args.gpu_id, use_cudnn=True)
+
+ model.param_init_net.GaussianFill(
+ [], input_blob.GetUnscopedName(),
+ shape=(1,) + data_config['input_size'], mean=0.0, std=1.0)
+ workspace.RunNetOnce(model.param_init_net)
+ workspace.CreateNet(model.net, overwrite=True)
+
+ batch_time = AverageMeter()
+ top1 = AverageMeter()
+ top5 = AverageMeter()
+ end = time.time()
+ for i, (input, target) in enumerate(loader):
+ # run the net and return prediction
+ caffe2_in = input.data.numpy()
+ workspace.FeedBlob(input_blob, caffe2_in, device_opts)
+ workspace.RunNet(model.net, num_iter=1)
+ output = workspace.FetchBlob(output_blob)
+
+ # measure accuracy and record loss
+ prec1, prec5 = accuracy_np(output.data, target.numpy())
+ top1.update(prec1.item(), input.size(0))
+ top5.update(prec5.item(), input.size(0))
+
+ # measure elapsed time
+ batch_time.update(time.time() - end)
+ end = time.time()
+
+ if i % args.print_freq == 0:
+ print('Test: [{0}/{1}]\t'
+ 'Time {batch_time.val:.3f} ({batch_time.avg:.3f}, {rate_avg:.3f}/s, {ms_avg:.3f} ms/sample) \t'
+ 'Prec@1 {top1.val:.3f} ({top1.avg:.3f})\t'
+ 'Prec@5 {top5.val:.3f} ({top5.avg:.3f})'.format(
+ i, len(loader), batch_time=batch_time, rate_avg=input.size(0) / batch_time.avg,
+ ms_avg=100 * batch_time.avg / input.size(0), top1=top1, top5=top5))
+
+ print(' * Prec@1 {top1.avg:.3f} ({top1a:.3f}) Prec@5 {top5.avg:.3f} ({top5a:.3f})'.format(
+ top1=top1, top1a=100-top1.avg, top5=top5, top5a=100.-top5.avg))
+
+
+def accuracy_np(output, target):
+ max_indices = np.argsort(output, axis=1)[:, ::-1]
+ top5 = 100 * np.equal(max_indices[:, :5], target[:, np.newaxis]).sum(axis=1).mean()
+ top1 = 100 * np.equal(max_indices[:, 0], target).mean()
+ return top1, top5
+
+
+if __name__ == '__main__':
+ main()
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/data/__init__.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/data/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..75824e6203af44be8b60c06dded81bf393f9134a
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/data/__init__.py
@@ -0,0 +1,3 @@
+from .dataset import Dataset
+from .transforms import *
+from .loader import create_loader
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/data/dataset.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/data/dataset.py
new file mode 100644
index 0000000000000000000000000000000000000000..8b4b07f0eb57dbe517714e170796647db8d7a6ed
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/data/dataset.py
@@ -0,0 +1,91 @@
+""" Quick n simple image folder dataset
+
+Copyright 2020 Ross Wightman
+"""
+import torch.utils.data as data
+
+import os
+import re
+import torch
+from PIL import Image
+
+
+IMG_EXTENSIONS = ['.png', '.jpg', '.jpeg']
+
+
+def natural_key(string_):
+ """See http://www.codinghorror.com/blog/archives/001018.html"""
+ return [int(s) if s.isdigit() else s for s in re.split(r'(\d+)', string_.lower())]
+
+
+def find_images_and_targets(folder, types=IMG_EXTENSIONS, class_to_idx=None, leaf_name_only=True, sort=True):
+ if class_to_idx is None:
+ class_to_idx = dict()
+ build_class_idx = True
+ else:
+ build_class_idx = False
+ labels = []
+ filenames = []
+ for root, subdirs, files in os.walk(folder, topdown=False):
+ rel_path = os.path.relpath(root, folder) if (root != folder) else ''
+ label = os.path.basename(rel_path) if leaf_name_only else rel_path.replace(os.path.sep, '_')
+ if build_class_idx and not subdirs:
+ class_to_idx[label] = None
+ for f in files:
+ base, ext = os.path.splitext(f)
+ if ext.lower() in types:
+ filenames.append(os.path.join(root, f))
+ labels.append(label)
+ if build_class_idx:
+ classes = sorted(class_to_idx.keys(), key=natural_key)
+ for idx, c in enumerate(classes):
+ class_to_idx[c] = idx
+ images_and_targets = zip(filenames, [class_to_idx[l] for l in labels])
+ if sort:
+ images_and_targets = sorted(images_and_targets, key=lambda k: natural_key(k[0]))
+ if build_class_idx:
+ return images_and_targets, classes, class_to_idx
+ else:
+ return images_and_targets
+
+
+class Dataset(data.Dataset):
+
+ def __init__(
+ self,
+ root,
+ transform=None,
+ load_bytes=False):
+
+ imgs, _, _ = find_images_and_targets(root)
+ if len(imgs) == 0:
+ raise(RuntimeError("Found 0 images in subfolders of: " + root + "\n"
+ "Supported image extensions are: " + ",".join(IMG_EXTENSIONS)))
+ self.root = root
+ self.imgs = imgs
+ self.transform = transform
+ self.load_bytes = load_bytes
+
+ def __getitem__(self, index):
+ path, target = self.imgs[index]
+ img = open(path, 'rb').read() if self.load_bytes else Image.open(path).convert('RGB')
+ if self.transform is not None:
+ img = self.transform(img)
+ if target is None:
+ target = torch.zeros(1).long()
+ return img, target
+
+ def __len__(self):
+ return len(self.imgs)
+
+ def filenames(self, indices=[], basename=False):
+ if indices:
+ if basename:
+ return [os.path.basename(self.imgs[i][0]) for i in indices]
+ else:
+ return [self.imgs[i][0] for i in indices]
+ else:
+ if basename:
+ return [os.path.basename(x[0]) for x in self.imgs]
+ else:
+ return [x[0] for x in self.imgs]
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/data/loader.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/data/loader.py
new file mode 100644
index 0000000000000000000000000000000000000000..23e0f69ea23ac0a9be7752986220ed434da72ef4
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/data/loader.py
@@ -0,0 +1,108 @@
+""" Fast Collate, CUDA Prefetcher
+
+Prefetcher and Fast Collate inspired by NVIDIA APEX example at
+https://github.com/NVIDIA/apex/commit/d5e2bb4bdeedd27b1dfaf5bb2b24d6c000dee9be#diff-cf86c282ff7fba81fad27a559379d5bf
+
+Hacked together by / Copyright 2020 Ross Wightman
+"""
+import torch
+import torch.utils.data
+from .transforms import *
+
+
+def fast_collate(batch):
+ targets = torch.tensor([b[1] for b in batch], dtype=torch.int64)
+ batch_size = len(targets)
+ tensor = torch.zeros((batch_size, *batch[0][0].shape), dtype=torch.uint8)
+ for i in range(batch_size):
+ tensor[i] += torch.from_numpy(batch[i][0])
+
+ return tensor, targets
+
+
+class PrefetchLoader:
+
+ def __init__(self,
+ loader,
+ mean=IMAGENET_DEFAULT_MEAN,
+ std=IMAGENET_DEFAULT_STD):
+ self.loader = loader
+ self.mean = torch.tensor([x * 255 for x in mean]).cuda().view(1, 3, 1, 1)
+ self.std = torch.tensor([x * 255 for x in std]).cuda().view(1, 3, 1, 1)
+
+ def __iter__(self):
+ stream = torch.cuda.Stream()
+ first = True
+
+ for next_input, next_target in self.loader:
+ with torch.cuda.stream(stream):
+ next_input = next_input.cuda(non_blocking=True)
+ next_target = next_target.cuda(non_blocking=True)
+ next_input = next_input.float().sub_(self.mean).div_(self.std)
+
+ if not first:
+ yield input, target
+ else:
+ first = False
+
+ torch.cuda.current_stream().wait_stream(stream)
+ input = next_input
+ target = next_target
+
+ yield input, target
+
+ def __len__(self):
+ return len(self.loader)
+
+ @property
+ def sampler(self):
+ return self.loader.sampler
+
+
+def create_loader(
+ dataset,
+ input_size,
+ batch_size,
+ is_training=False,
+ use_prefetcher=True,
+ interpolation='bilinear',
+ mean=IMAGENET_DEFAULT_MEAN,
+ std=IMAGENET_DEFAULT_STD,
+ num_workers=1,
+ crop_pct=None,
+ tensorflow_preprocessing=False
+):
+ if isinstance(input_size, tuple):
+ img_size = input_size[-2:]
+ else:
+ img_size = input_size
+
+ if tensorflow_preprocessing and use_prefetcher:
+ from data.tf_preprocessing import TfPreprocessTransform
+ transform = TfPreprocessTransform(
+ is_training=is_training, size=img_size, interpolation=interpolation)
+ else:
+ transform = transforms_imagenet_eval(
+ img_size,
+ interpolation=interpolation,
+ use_prefetcher=use_prefetcher,
+ mean=mean,
+ std=std,
+ crop_pct=crop_pct)
+
+ dataset.transform = transform
+
+ loader = torch.utils.data.DataLoader(
+ dataset,
+ batch_size=batch_size,
+ shuffle=False,
+ num_workers=num_workers,
+ collate_fn=fast_collate if use_prefetcher else torch.utils.data.dataloader.default_collate,
+ )
+ if use_prefetcher:
+ loader = PrefetchLoader(
+ loader,
+ mean=mean,
+ std=std)
+
+ return loader
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/data/tf_preprocessing.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/data/tf_preprocessing.py
new file mode 100644
index 0000000000000000000000000000000000000000..ee3adeaed6e06bcdec68e312e335b6e2fa31faec
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/data/tf_preprocessing.py
@@ -0,0 +1,234 @@
+""" Tensorflow Preprocessing Adapter
+
+Allows use of Tensorflow preprocessing pipeline in PyTorch Transform
+
+Copyright of original Tensorflow code below.
+
+Hacked together by / Copyright 2020 Ross Wightman
+"""
+# Copyright 2018 The TensorFlow Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ==============================================================================
+from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
+
+import tensorflow as tf
+import numpy as np
+
+IMAGE_SIZE = 224
+CROP_PADDING = 32
+
+
+def distorted_bounding_box_crop(image_bytes,
+ bbox,
+ min_object_covered=0.1,
+ aspect_ratio_range=(0.75, 1.33),
+ area_range=(0.05, 1.0),
+ max_attempts=100,
+ scope=None):
+ """Generates cropped_image using one of the bboxes randomly distorted.
+
+ See `tf.image.sample_distorted_bounding_box` for more documentation.
+
+ Args:
+ image_bytes: `Tensor` of binary image data.
+ bbox: `Tensor` of bounding boxes arranged `[1, num_boxes, coords]`
+ where each coordinate is [0, 1) and the coordinates are arranged
+ as `[ymin, xmin, ymax, xmax]`. If num_boxes is 0 then use the whole
+ image.
+ min_object_covered: An optional `float`. Defaults to `0.1`. The cropped
+ area of the image must contain at least this fraction of any bounding
+ box supplied.
+ aspect_ratio_range: An optional list of `float`s. The cropped area of the
+ image must have an aspect ratio = width / height within this range.
+ area_range: An optional list of `float`s. The cropped area of the image
+ must contain a fraction of the supplied image within in this range.
+ max_attempts: An optional `int`. Number of attempts at generating a cropped
+ region of the image of the specified constraints. After `max_attempts`
+ failures, return the entire image.
+ scope: Optional `str` for name scope.
+ Returns:
+ cropped image `Tensor`
+ """
+ with tf.name_scope(scope, 'distorted_bounding_box_crop', [image_bytes, bbox]):
+ shape = tf.image.extract_jpeg_shape(image_bytes)
+ sample_distorted_bounding_box = tf.image.sample_distorted_bounding_box(
+ shape,
+ bounding_boxes=bbox,
+ min_object_covered=min_object_covered,
+ aspect_ratio_range=aspect_ratio_range,
+ area_range=area_range,
+ max_attempts=max_attempts,
+ use_image_if_no_bounding_boxes=True)
+ bbox_begin, bbox_size, _ = sample_distorted_bounding_box
+
+ # Crop the image to the specified bounding box.
+ offset_y, offset_x, _ = tf.unstack(bbox_begin)
+ target_height, target_width, _ = tf.unstack(bbox_size)
+ crop_window = tf.stack([offset_y, offset_x, target_height, target_width])
+ image = tf.image.decode_and_crop_jpeg(image_bytes, crop_window, channels=3)
+
+ return image
+
+
+def _at_least_x_are_equal(a, b, x):
+ """At least `x` of `a` and `b` `Tensors` are equal."""
+ match = tf.equal(a, b)
+ match = tf.cast(match, tf.int32)
+ return tf.greater_equal(tf.reduce_sum(match), x)
+
+
+def _decode_and_random_crop(image_bytes, image_size, resize_method):
+ """Make a random crop of image_size."""
+ bbox = tf.constant([0.0, 0.0, 1.0, 1.0], dtype=tf.float32, shape=[1, 1, 4])
+ image = distorted_bounding_box_crop(
+ image_bytes,
+ bbox,
+ min_object_covered=0.1,
+ aspect_ratio_range=(3. / 4, 4. / 3.),
+ area_range=(0.08, 1.0),
+ max_attempts=10,
+ scope=None)
+ original_shape = tf.image.extract_jpeg_shape(image_bytes)
+ bad = _at_least_x_are_equal(original_shape, tf.shape(image), 3)
+
+ image = tf.cond(
+ bad,
+ lambda: _decode_and_center_crop(image_bytes, image_size),
+ lambda: tf.image.resize([image], [image_size, image_size], resize_method)[0])
+
+ return image
+
+
+def _decode_and_center_crop(image_bytes, image_size, resize_method):
+ """Crops to center of image with padding then scales image_size."""
+ shape = tf.image.extract_jpeg_shape(image_bytes)
+ image_height = shape[0]
+ image_width = shape[1]
+
+ padded_center_crop_size = tf.cast(
+ ((image_size / (image_size + CROP_PADDING)) *
+ tf.cast(tf.minimum(image_height, image_width), tf.float32)),
+ tf.int32)
+
+ offset_height = ((image_height - padded_center_crop_size) + 1) // 2
+ offset_width = ((image_width - padded_center_crop_size) + 1) // 2
+ crop_window = tf.stack([offset_height, offset_width,
+ padded_center_crop_size, padded_center_crop_size])
+ image = tf.image.decode_and_crop_jpeg(image_bytes, crop_window, channels=3)
+ image = tf.image.resize([image], [image_size, image_size], resize_method)[0]
+
+ return image
+
+
+def _flip(image):
+ """Random horizontal image flip."""
+ image = tf.image.random_flip_left_right(image)
+ return image
+
+
+def preprocess_for_train(image_bytes, use_bfloat16, image_size=IMAGE_SIZE, interpolation='bicubic'):
+ """Preprocesses the given image for evaluation.
+
+ Args:
+ image_bytes: `Tensor` representing an image binary of arbitrary size.
+ use_bfloat16: `bool` for whether to use bfloat16.
+ image_size: image size.
+ interpolation: image interpolation method
+
+ Returns:
+ A preprocessed image `Tensor`.
+ """
+ resize_method = tf.image.ResizeMethod.BICUBIC if interpolation == 'bicubic' else tf.image.ResizeMethod.BILINEAR
+ image = _decode_and_random_crop(image_bytes, image_size, resize_method)
+ image = _flip(image)
+ image = tf.reshape(image, [image_size, image_size, 3])
+ image = tf.image.convert_image_dtype(
+ image, dtype=tf.bfloat16 if use_bfloat16 else tf.float32)
+ return image
+
+
+def preprocess_for_eval(image_bytes, use_bfloat16, image_size=IMAGE_SIZE, interpolation='bicubic'):
+ """Preprocesses the given image for evaluation.
+
+ Args:
+ image_bytes: `Tensor` representing an image binary of arbitrary size.
+ use_bfloat16: `bool` for whether to use bfloat16.
+ image_size: image size.
+ interpolation: image interpolation method
+
+ Returns:
+ A preprocessed image `Tensor`.
+ """
+ resize_method = tf.image.ResizeMethod.BICUBIC if interpolation == 'bicubic' else tf.image.ResizeMethod.BILINEAR
+ image = _decode_and_center_crop(image_bytes, image_size, resize_method)
+ image = tf.reshape(image, [image_size, image_size, 3])
+ image = tf.image.convert_image_dtype(
+ image, dtype=tf.bfloat16 if use_bfloat16 else tf.float32)
+ return image
+
+
+def preprocess_image(image_bytes,
+ is_training=False,
+ use_bfloat16=False,
+ image_size=IMAGE_SIZE,
+ interpolation='bicubic'):
+ """Preprocesses the given image.
+
+ Args:
+ image_bytes: `Tensor` representing an image binary of arbitrary size.
+ is_training: `bool` for whether the preprocessing is for training.
+ use_bfloat16: `bool` for whether to use bfloat16.
+ image_size: image size.
+ interpolation: image interpolation method
+
+ Returns:
+ A preprocessed image `Tensor` with value range of [0, 255].
+ """
+ if is_training:
+ return preprocess_for_train(image_bytes, use_bfloat16, image_size, interpolation)
+ else:
+ return preprocess_for_eval(image_bytes, use_bfloat16, image_size, interpolation)
+
+
+class TfPreprocessTransform:
+
+ def __init__(self, is_training=False, size=224, interpolation='bicubic'):
+ self.is_training = is_training
+ self.size = size[0] if isinstance(size, tuple) else size
+ self.interpolation = interpolation
+ self._image_bytes = None
+ self.process_image = self._build_tf_graph()
+ self.sess = None
+
+ def _build_tf_graph(self):
+ with tf.device('/cpu:0'):
+ self._image_bytes = tf.placeholder(
+ shape=[],
+ dtype=tf.string,
+ )
+ img = preprocess_image(
+ self._image_bytes, self.is_training, False, self.size, self.interpolation)
+ return img
+
+ def __call__(self, image_bytes):
+ if self.sess is None:
+ self.sess = tf.Session()
+ img = self.sess.run(self.process_image, feed_dict={self._image_bytes: image_bytes})
+ img = img.round().clip(0, 255).astype(np.uint8)
+ if img.ndim < 3:
+ img = np.expand_dims(img, axis=-1)
+ img = np.rollaxis(img, 2) # HWC to CHW
+ return img
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/data/transforms.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/data/transforms.py
new file mode 100644
index 0000000000000000000000000000000000000000..a570d484683f8ad5612bc22fb9ecae7e75c36afc
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/data/transforms.py
@@ -0,0 +1,150 @@
+import torch
+from torchvision import transforms
+from PIL import Image
+import math
+import numpy as np
+
+DEFAULT_CROP_PCT = 0.875
+
+IMAGENET_DEFAULT_MEAN = (0.485, 0.456, 0.406)
+IMAGENET_DEFAULT_STD = (0.229, 0.224, 0.225)
+IMAGENET_INCEPTION_MEAN = (0.5, 0.5, 0.5)
+IMAGENET_INCEPTION_STD = (0.5, 0.5, 0.5)
+IMAGENET_DPN_MEAN = (124 / 255, 117 / 255, 104 / 255)
+IMAGENET_DPN_STD = tuple([1 / (.0167 * 255)] * 3)
+
+
+def resolve_data_config(model, args, default_cfg={}, verbose=True):
+ new_config = {}
+ default_cfg = default_cfg
+ if not default_cfg and model is not None and hasattr(model, 'default_cfg'):
+ default_cfg = model.default_cfg
+
+ # Resolve input/image size
+ # FIXME grayscale/chans arg to use different # channels?
+ in_chans = 3
+ input_size = (in_chans, 224, 224)
+ if args.img_size is not None:
+ # FIXME support passing img_size as tuple, non-square
+ assert isinstance(args.img_size, int)
+ input_size = (in_chans, args.img_size, args.img_size)
+ elif 'input_size' in default_cfg:
+ input_size = default_cfg['input_size']
+ new_config['input_size'] = input_size
+
+ # resolve interpolation method
+ new_config['interpolation'] = 'bicubic'
+ if args.interpolation:
+ new_config['interpolation'] = args.interpolation
+ elif 'interpolation' in default_cfg:
+ new_config['interpolation'] = default_cfg['interpolation']
+
+ # resolve dataset + model mean for normalization
+ new_config['mean'] = IMAGENET_DEFAULT_MEAN
+ if args.mean is not None:
+ mean = tuple(args.mean)
+ if len(mean) == 1:
+ mean = tuple(list(mean) * in_chans)
+ else:
+ assert len(mean) == in_chans
+ new_config['mean'] = mean
+ elif 'mean' in default_cfg:
+ new_config['mean'] = default_cfg['mean']
+
+ # resolve dataset + model std deviation for normalization
+ new_config['std'] = IMAGENET_DEFAULT_STD
+ if args.std is not None:
+ std = tuple(args.std)
+ if len(std) == 1:
+ std = tuple(list(std) * in_chans)
+ else:
+ assert len(std) == in_chans
+ new_config['std'] = std
+ elif 'std' in default_cfg:
+ new_config['std'] = default_cfg['std']
+
+ # resolve default crop percentage
+ new_config['crop_pct'] = DEFAULT_CROP_PCT
+ if args.crop_pct is not None:
+ new_config['crop_pct'] = args.crop_pct
+ elif 'crop_pct' in default_cfg:
+ new_config['crop_pct'] = default_cfg['crop_pct']
+
+ if verbose:
+ print('Data processing configuration for current model + dataset:')
+ for n, v in new_config.items():
+ print('\t%s: %s' % (n, str(v)))
+
+ return new_config
+
+
+class ToNumpy:
+
+ def __call__(self, pil_img):
+ np_img = np.array(pil_img, dtype=np.uint8)
+ if np_img.ndim < 3:
+ np_img = np.expand_dims(np_img, axis=-1)
+ np_img = np.rollaxis(np_img, 2) # HWC to CHW
+ return np_img
+
+
+class ToTensor:
+
+ def __init__(self, dtype=torch.float32):
+ self.dtype = dtype
+
+ def __call__(self, pil_img):
+ np_img = np.array(pil_img, dtype=np.uint8)
+ if np_img.ndim < 3:
+ np_img = np.expand_dims(np_img, axis=-1)
+ np_img = np.rollaxis(np_img, 2) # HWC to CHW
+ return torch.from_numpy(np_img).to(dtype=self.dtype)
+
+
+def _pil_interp(method):
+ if method == 'bicubic':
+ return Image.BICUBIC
+ elif method == 'lanczos':
+ return Image.LANCZOS
+ elif method == 'hamming':
+ return Image.HAMMING
+ else:
+ # default bilinear, do we want to allow nearest?
+ return Image.BILINEAR
+
+
+def transforms_imagenet_eval(
+ img_size=224,
+ crop_pct=None,
+ interpolation='bilinear',
+ use_prefetcher=False,
+ mean=IMAGENET_DEFAULT_MEAN,
+ std=IMAGENET_DEFAULT_STD):
+ crop_pct = crop_pct or DEFAULT_CROP_PCT
+
+ if isinstance(img_size, tuple):
+ assert len(img_size) == 2
+ if img_size[-1] == img_size[-2]:
+ # fall-back to older behaviour so Resize scales to shortest edge if target is square
+ scale_size = int(math.floor(img_size[0] / crop_pct))
+ else:
+ scale_size = tuple([int(x / crop_pct) for x in img_size])
+ else:
+ scale_size = int(math.floor(img_size / crop_pct))
+
+ tfl = [
+ transforms.Resize(scale_size, _pil_interp(interpolation)),
+ transforms.CenterCrop(img_size),
+ ]
+ if use_prefetcher:
+ # prefetcher and collate will handle tensor conversion and norm
+ tfl += [ToNumpy()]
+ else:
+ tfl += [
+ transforms.ToTensor(),
+ transforms.Normalize(
+ mean=torch.tensor(mean),
+ std=torch.tensor(std))
+ ]
+
+ return transforms.Compose(tfl)
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/__init__.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..2e441a5838d1e972823b9668ac8d459445f6f6ce
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/__init__.py
@@ -0,0 +1,5 @@
+from .gen_efficientnet import *
+from .mobilenetv3 import *
+from .model_factory import create_model
+from .config import is_exportable, is_scriptable, set_exportable, set_scriptable
+from .activations import *
\ No newline at end of file
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/activations/__init__.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/activations/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..813421a743ffc33b8eb53ebf62dd4a03d831b654
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/activations/__init__.py
@@ -0,0 +1,137 @@
+from geffnet import config
+from geffnet.activations.activations_me import *
+from geffnet.activations.activations_jit import *
+from geffnet.activations.activations import *
+import torch
+
+_has_silu = 'silu' in dir(torch.nn.functional)
+
+_ACT_FN_DEFAULT = dict(
+ silu=F.silu if _has_silu else swish,
+ swish=F.silu if _has_silu else swish,
+ mish=mish,
+ relu=F.relu,
+ relu6=F.relu6,
+ sigmoid=sigmoid,
+ tanh=tanh,
+ hard_sigmoid=hard_sigmoid,
+ hard_swish=hard_swish,
+)
+
+_ACT_FN_JIT = dict(
+ silu=F.silu if _has_silu else swish_jit,
+ swish=F.silu if _has_silu else swish_jit,
+ mish=mish_jit,
+)
+
+_ACT_FN_ME = dict(
+ silu=F.silu if _has_silu else swish_me,
+ swish=F.silu if _has_silu else swish_me,
+ mish=mish_me,
+ hard_swish=hard_swish_me,
+ hard_sigmoid_jit=hard_sigmoid_me,
+)
+
+_ACT_LAYER_DEFAULT = dict(
+ silu=nn.SiLU if _has_silu else Swish,
+ swish=nn.SiLU if _has_silu else Swish,
+ mish=Mish,
+ relu=nn.ReLU,
+ relu6=nn.ReLU6,
+ sigmoid=Sigmoid,
+ tanh=Tanh,
+ hard_sigmoid=HardSigmoid,
+ hard_swish=HardSwish,
+)
+
+_ACT_LAYER_JIT = dict(
+ silu=nn.SiLU if _has_silu else SwishJit,
+ swish=nn.SiLU if _has_silu else SwishJit,
+ mish=MishJit,
+)
+
+_ACT_LAYER_ME = dict(
+ silu=nn.SiLU if _has_silu else SwishMe,
+ swish=nn.SiLU if _has_silu else SwishMe,
+ mish=MishMe,
+ hard_swish=HardSwishMe,
+ hard_sigmoid=HardSigmoidMe
+)
+
+_OVERRIDE_FN = dict()
+_OVERRIDE_LAYER = dict()
+
+
+def add_override_act_fn(name, fn):
+ global _OVERRIDE_FN
+ _OVERRIDE_FN[name] = fn
+
+
+def update_override_act_fn(overrides):
+ assert isinstance(overrides, dict)
+ global _OVERRIDE_FN
+ _OVERRIDE_FN.update(overrides)
+
+
+def clear_override_act_fn():
+ global _OVERRIDE_FN
+ _OVERRIDE_FN = dict()
+
+
+def add_override_act_layer(name, fn):
+ _OVERRIDE_LAYER[name] = fn
+
+
+def update_override_act_layer(overrides):
+ assert isinstance(overrides, dict)
+ global _OVERRIDE_LAYER
+ _OVERRIDE_LAYER.update(overrides)
+
+
+def clear_override_act_layer():
+ global _OVERRIDE_LAYER
+ _OVERRIDE_LAYER = dict()
+
+
+def get_act_fn(name='relu'):
+ """ Activation Function Factory
+ Fetching activation fns by name with this function allows export or torch script friendly
+ functions to be returned dynamically based on current config.
+ """
+ if name in _OVERRIDE_FN:
+ return _OVERRIDE_FN[name]
+ use_me = not (config.is_exportable() or config.is_scriptable() or config.is_no_jit())
+ if use_me and name in _ACT_FN_ME:
+ # If not exporting or scripting the model, first look for a memory optimized version
+ # activation with custom autograd, then fallback to jit scripted, then a Python or Torch builtin
+ return _ACT_FN_ME[name]
+ if config.is_exportable() and name in ('silu', 'swish'):
+ # FIXME PyTorch SiLU doesn't ONNX export, this is a temp hack
+ return swish
+ use_jit = not (config.is_exportable() or config.is_no_jit())
+ # NOTE: export tracing should work with jit scripted components, but I keep running into issues
+ if use_jit and name in _ACT_FN_JIT: # jit scripted models should be okay for export/scripting
+ return _ACT_FN_JIT[name]
+ return _ACT_FN_DEFAULT[name]
+
+
+def get_act_layer(name='relu'):
+ """ Activation Layer Factory
+ Fetching activation layers by name with this function allows export or torch script friendly
+ functions to be returned dynamically based on current config.
+ """
+ if name in _OVERRIDE_LAYER:
+ return _OVERRIDE_LAYER[name]
+ use_me = not (config.is_exportable() or config.is_scriptable() or config.is_no_jit())
+ if use_me and name in _ACT_LAYER_ME:
+ return _ACT_LAYER_ME[name]
+ if config.is_exportable() and name in ('silu', 'swish'):
+ # FIXME PyTorch SiLU doesn't ONNX export, this is a temp hack
+ return Swish
+ use_jit = not (config.is_exportable() or config.is_no_jit())
+ # NOTE: export tracing should work with jit scripted components, but I keep running into issues
+ if use_jit and name in _ACT_FN_JIT: # jit scripted models should be okay for export/scripting
+ return _ACT_LAYER_JIT[name]
+ return _ACT_LAYER_DEFAULT[name]
+
+
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/activations/activations.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/activations/activations.py
new file mode 100644
index 0000000000000000000000000000000000000000..bdea692d1397673b2513d898c33edbcb37d94240
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/activations/activations.py
@@ -0,0 +1,102 @@
+""" Activations
+
+A collection of activations fn and modules with a common interface so that they can
+easily be swapped. All have an `inplace` arg even if not used.
+
+Copyright 2020 Ross Wightman
+"""
+from torch import nn as nn
+from torch.nn import functional as F
+
+
+def swish(x, inplace: bool = False):
+ """Swish - Described originally as SiLU (https://arxiv.org/abs/1702.03118v3)
+ and also as Swish (https://arxiv.org/abs/1710.05941).
+
+ TODO Rename to SiLU with addition to PyTorch
+ """
+ return x.mul_(x.sigmoid()) if inplace else x.mul(x.sigmoid())
+
+
+class Swish(nn.Module):
+ def __init__(self, inplace: bool = False):
+ super(Swish, self).__init__()
+ self.inplace = inplace
+
+ def forward(self, x):
+ return swish(x, self.inplace)
+
+
+def mish(x, inplace: bool = False):
+ """Mish: A Self Regularized Non-Monotonic Neural Activation Function - https://arxiv.org/abs/1908.08681
+ """
+ return x.mul(F.softplus(x).tanh())
+
+
+class Mish(nn.Module):
+ def __init__(self, inplace: bool = False):
+ super(Mish, self).__init__()
+ self.inplace = inplace
+
+ def forward(self, x):
+ return mish(x, self.inplace)
+
+
+def sigmoid(x, inplace: bool = False):
+ return x.sigmoid_() if inplace else x.sigmoid()
+
+
+# PyTorch has this, but not with a consistent inplace argmument interface
+class Sigmoid(nn.Module):
+ def __init__(self, inplace: bool = False):
+ super(Sigmoid, self).__init__()
+ self.inplace = inplace
+
+ def forward(self, x):
+ return x.sigmoid_() if self.inplace else x.sigmoid()
+
+
+def tanh(x, inplace: bool = False):
+ return x.tanh_() if inplace else x.tanh()
+
+
+# PyTorch has this, but not with a consistent inplace argmument interface
+class Tanh(nn.Module):
+ def __init__(self, inplace: bool = False):
+ super(Tanh, self).__init__()
+ self.inplace = inplace
+
+ def forward(self, x):
+ return x.tanh_() if self.inplace else x.tanh()
+
+
+def hard_swish(x, inplace: bool = False):
+ inner = F.relu6(x + 3.).div_(6.)
+ return x.mul_(inner) if inplace else x.mul(inner)
+
+
+class HardSwish(nn.Module):
+ def __init__(self, inplace: bool = False):
+ super(HardSwish, self).__init__()
+ self.inplace = inplace
+
+ def forward(self, x):
+ return hard_swish(x, self.inplace)
+
+
+def hard_sigmoid(x, inplace: bool = False):
+ if inplace:
+ return x.add_(3.).clamp_(0., 6.).div_(6.)
+ else:
+ return F.relu6(x + 3.) / 6.
+
+
+class HardSigmoid(nn.Module):
+ def __init__(self, inplace: bool = False):
+ super(HardSigmoid, self).__init__()
+ self.inplace = inplace
+
+ def forward(self, x):
+ return hard_sigmoid(x, self.inplace)
+
+
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/activations/activations_jit.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/activations/activations_jit.py
new file mode 100644
index 0000000000000000000000000000000000000000..7176b05e779787528a47f20d55d64d4a0f219360
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/activations/activations_jit.py
@@ -0,0 +1,79 @@
+""" Activations (jit)
+
+A collection of jit-scripted activations fn and modules with a common interface so that they can
+easily be swapped. All have an `inplace` arg even if not used.
+
+All jit scripted activations are lacking in-place variations on purpose, scripted kernel fusion does not
+currently work across in-place op boundaries, thus performance is equal to or less than the non-scripted
+versions if they contain in-place ops.
+
+Copyright 2020 Ross Wightman
+"""
+
+import torch
+from torch import nn as nn
+from torch.nn import functional as F
+
+__all__ = ['swish_jit', 'SwishJit', 'mish_jit', 'MishJit',
+ 'hard_sigmoid_jit', 'HardSigmoidJit', 'hard_swish_jit', 'HardSwishJit']
+
+
+@torch.jit.script
+def swish_jit(x, inplace: bool = False):
+ """Swish - Described originally as SiLU (https://arxiv.org/abs/1702.03118v3)
+ and also as Swish (https://arxiv.org/abs/1710.05941).
+
+ TODO Rename to SiLU with addition to PyTorch
+ """
+ return x.mul(x.sigmoid())
+
+
+@torch.jit.script
+def mish_jit(x, _inplace: bool = False):
+ """Mish: A Self Regularized Non-Monotonic Neural Activation Function - https://arxiv.org/abs/1908.08681
+ """
+ return x.mul(F.softplus(x).tanh())
+
+
+class SwishJit(nn.Module):
+ def __init__(self, inplace: bool = False):
+ super(SwishJit, self).__init__()
+
+ def forward(self, x):
+ return swish_jit(x)
+
+
+class MishJit(nn.Module):
+ def __init__(self, inplace: bool = False):
+ super(MishJit, self).__init__()
+
+ def forward(self, x):
+ return mish_jit(x)
+
+
+@torch.jit.script
+def hard_sigmoid_jit(x, inplace: bool = False):
+ # return F.relu6(x + 3.) / 6.
+ return (x + 3).clamp(min=0, max=6).div(6.) # clamp seems ever so slightly faster?
+
+
+class HardSigmoidJit(nn.Module):
+ def __init__(self, inplace: bool = False):
+ super(HardSigmoidJit, self).__init__()
+
+ def forward(self, x):
+ return hard_sigmoid_jit(x)
+
+
+@torch.jit.script
+def hard_swish_jit(x, inplace: bool = False):
+ # return x * (F.relu6(x + 3.) / 6)
+ return x * (x + 3).clamp(min=0, max=6).div(6.) # clamp seems ever so slightly faster?
+
+
+class HardSwishJit(nn.Module):
+ def __init__(self, inplace: bool = False):
+ super(HardSwishJit, self).__init__()
+
+ def forward(self, x):
+ return hard_swish_jit(x)
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/activations/activations_me.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/activations/activations_me.py
new file mode 100644
index 0000000000000000000000000000000000000000..e91df5a50fdbe40bc386e2541a4fda743ad95e9a
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/activations/activations_me.py
@@ -0,0 +1,174 @@
+""" Activations (memory-efficient w/ custom autograd)
+
+A collection of activations fn and modules with a common interface so that they can
+easily be swapped. All have an `inplace` arg even if not used.
+
+These activations are not compatible with jit scripting or ONNX export of the model, please use either
+the JIT or basic versions of the activations.
+
+Copyright 2020 Ross Wightman
+"""
+
+import torch
+from torch import nn as nn
+from torch.nn import functional as F
+
+
+__all__ = ['swish_me', 'SwishMe', 'mish_me', 'MishMe',
+ 'hard_sigmoid_me', 'HardSigmoidMe', 'hard_swish_me', 'HardSwishMe']
+
+
+@torch.jit.script
+def swish_jit_fwd(x):
+ return x.mul(torch.sigmoid(x))
+
+
+@torch.jit.script
+def swish_jit_bwd(x, grad_output):
+ x_sigmoid = torch.sigmoid(x)
+ return grad_output * (x_sigmoid * (1 + x * (1 - x_sigmoid)))
+
+
+class SwishJitAutoFn(torch.autograd.Function):
+ """ torch.jit.script optimised Swish w/ memory-efficient checkpoint
+ Inspired by conversation btw Jeremy Howard & Adam Pazske
+ https://twitter.com/jeremyphoward/status/1188251041835315200
+
+ Swish - Described originally as SiLU (https://arxiv.org/abs/1702.03118v3)
+ and also as Swish (https://arxiv.org/abs/1710.05941).
+
+ TODO Rename to SiLU with addition to PyTorch
+ """
+
+ @staticmethod
+ def forward(ctx, x):
+ ctx.save_for_backward(x)
+ return swish_jit_fwd(x)
+
+ @staticmethod
+ def backward(ctx, grad_output):
+ x = ctx.saved_tensors[0]
+ return swish_jit_bwd(x, grad_output)
+
+
+def swish_me(x, inplace=False):
+ return SwishJitAutoFn.apply(x)
+
+
+class SwishMe(nn.Module):
+ def __init__(self, inplace: bool = False):
+ super(SwishMe, self).__init__()
+
+ def forward(self, x):
+ return SwishJitAutoFn.apply(x)
+
+
+@torch.jit.script
+def mish_jit_fwd(x):
+ return x.mul(torch.tanh(F.softplus(x)))
+
+
+@torch.jit.script
+def mish_jit_bwd(x, grad_output):
+ x_sigmoid = torch.sigmoid(x)
+ x_tanh_sp = F.softplus(x).tanh()
+ return grad_output.mul(x_tanh_sp + x * x_sigmoid * (1 - x_tanh_sp * x_tanh_sp))
+
+
+class MishJitAutoFn(torch.autograd.Function):
+ """ Mish: A Self Regularized Non-Monotonic Neural Activation Function - https://arxiv.org/abs/1908.08681
+ A memory efficient, jit scripted variant of Mish
+ """
+ @staticmethod
+ def forward(ctx, x):
+ ctx.save_for_backward(x)
+ return mish_jit_fwd(x)
+
+ @staticmethod
+ def backward(ctx, grad_output):
+ x = ctx.saved_tensors[0]
+ return mish_jit_bwd(x, grad_output)
+
+
+def mish_me(x, inplace=False):
+ return MishJitAutoFn.apply(x)
+
+
+class MishMe(nn.Module):
+ def __init__(self, inplace: bool = False):
+ super(MishMe, self).__init__()
+
+ def forward(self, x):
+ return MishJitAutoFn.apply(x)
+
+
+@torch.jit.script
+def hard_sigmoid_jit_fwd(x, inplace: bool = False):
+ return (x + 3).clamp(min=0, max=6).div(6.)
+
+
+@torch.jit.script
+def hard_sigmoid_jit_bwd(x, grad_output):
+ m = torch.ones_like(x) * ((x >= -3.) & (x <= 3.)) / 6.
+ return grad_output * m
+
+
+class HardSigmoidJitAutoFn(torch.autograd.Function):
+ @staticmethod
+ def forward(ctx, x):
+ ctx.save_for_backward(x)
+ return hard_sigmoid_jit_fwd(x)
+
+ @staticmethod
+ def backward(ctx, grad_output):
+ x = ctx.saved_tensors[0]
+ return hard_sigmoid_jit_bwd(x, grad_output)
+
+
+def hard_sigmoid_me(x, inplace: bool = False):
+ return HardSigmoidJitAutoFn.apply(x)
+
+
+class HardSigmoidMe(nn.Module):
+ def __init__(self, inplace: bool = False):
+ super(HardSigmoidMe, self).__init__()
+
+ def forward(self, x):
+ return HardSigmoidJitAutoFn.apply(x)
+
+
+@torch.jit.script
+def hard_swish_jit_fwd(x):
+ return x * (x + 3).clamp(min=0, max=6).div(6.)
+
+
+@torch.jit.script
+def hard_swish_jit_bwd(x, grad_output):
+ m = torch.ones_like(x) * (x >= 3.)
+ m = torch.where((x >= -3.) & (x <= 3.), x / 3. + .5, m)
+ return grad_output * m
+
+
+class HardSwishJitAutoFn(torch.autograd.Function):
+ """A memory efficient, jit-scripted HardSwish activation"""
+ @staticmethod
+ def forward(ctx, x):
+ ctx.save_for_backward(x)
+ return hard_swish_jit_fwd(x)
+
+ @staticmethod
+ def backward(ctx, grad_output):
+ x = ctx.saved_tensors[0]
+ return hard_swish_jit_bwd(x, grad_output)
+
+
+def hard_swish_me(x, inplace=False):
+ return HardSwishJitAutoFn.apply(x)
+
+
+class HardSwishMe(nn.Module):
+ def __init__(self, inplace: bool = False):
+ super(HardSwishMe, self).__init__()
+
+ def forward(self, x):
+ return HardSwishJitAutoFn.apply(x)
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/config.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/config.py
new file mode 100644
index 0000000000000000000000000000000000000000..27d5307fd9ee0246f1e35f41520f17385d23f1dd
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/config.py
@@ -0,0 +1,123 @@
+""" Global layer config state
+"""
+from typing import Any, Optional
+
+__all__ = [
+ 'is_exportable', 'is_scriptable', 'is_no_jit', 'layer_config_kwargs',
+ 'set_exportable', 'set_scriptable', 'set_no_jit', 'set_layer_config'
+]
+
+# Set to True if prefer to have layers with no jit optimization (includes activations)
+_NO_JIT = False
+
+# Set to True if prefer to have activation layers with no jit optimization
+# NOTE not currently used as no difference between no_jit and no_activation jit as only layers obeying
+# the jit flags so far are activations. This will change as more layers are updated and/or added.
+_NO_ACTIVATION_JIT = False
+
+# Set to True if exporting a model with Same padding via ONNX
+_EXPORTABLE = False
+
+# Set to True if wanting to use torch.jit.script on a model
+_SCRIPTABLE = False
+
+
+def is_no_jit():
+ return _NO_JIT
+
+
+class set_no_jit:
+ def __init__(self, mode: bool) -> None:
+ global _NO_JIT
+ self.prev = _NO_JIT
+ _NO_JIT = mode
+
+ def __enter__(self) -> None:
+ pass
+
+ def __exit__(self, *args: Any) -> bool:
+ global _NO_JIT
+ _NO_JIT = self.prev
+ return False
+
+
+def is_exportable():
+ return _EXPORTABLE
+
+
+class set_exportable:
+ def __init__(self, mode: bool) -> None:
+ global _EXPORTABLE
+ self.prev = _EXPORTABLE
+ _EXPORTABLE = mode
+
+ def __enter__(self) -> None:
+ pass
+
+ def __exit__(self, *args: Any) -> bool:
+ global _EXPORTABLE
+ _EXPORTABLE = self.prev
+ return False
+
+
+def is_scriptable():
+ return _SCRIPTABLE
+
+
+class set_scriptable:
+ def __init__(self, mode: bool) -> None:
+ global _SCRIPTABLE
+ self.prev = _SCRIPTABLE
+ _SCRIPTABLE = mode
+
+ def __enter__(self) -> None:
+ pass
+
+ def __exit__(self, *args: Any) -> bool:
+ global _SCRIPTABLE
+ _SCRIPTABLE = self.prev
+ return False
+
+
+class set_layer_config:
+ """ Layer config context manager that allows setting all layer config flags at once.
+ If a flag arg is None, it will not change the current value.
+ """
+ def __init__(
+ self,
+ scriptable: Optional[bool] = None,
+ exportable: Optional[bool] = None,
+ no_jit: Optional[bool] = None,
+ no_activation_jit: Optional[bool] = None):
+ global _SCRIPTABLE
+ global _EXPORTABLE
+ global _NO_JIT
+ global _NO_ACTIVATION_JIT
+ self.prev = _SCRIPTABLE, _EXPORTABLE, _NO_JIT, _NO_ACTIVATION_JIT
+ if scriptable is not None:
+ _SCRIPTABLE = scriptable
+ if exportable is not None:
+ _EXPORTABLE = exportable
+ if no_jit is not None:
+ _NO_JIT = no_jit
+ if no_activation_jit is not None:
+ _NO_ACTIVATION_JIT = no_activation_jit
+
+ def __enter__(self) -> None:
+ pass
+
+ def __exit__(self, *args: Any) -> bool:
+ global _SCRIPTABLE
+ global _EXPORTABLE
+ global _NO_JIT
+ global _NO_ACTIVATION_JIT
+ _SCRIPTABLE, _EXPORTABLE, _NO_JIT, _NO_ACTIVATION_JIT = self.prev
+ return False
+
+
+def layer_config_kwargs(kwargs):
+ """ Consume config kwargs and return contextmgr obj """
+ return set_layer_config(
+ scriptable=kwargs.pop('scriptable', None),
+ exportable=kwargs.pop('exportable', None),
+ no_jit=kwargs.pop('no_jit', None))
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/conv2d_layers.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/conv2d_layers.py
new file mode 100644
index 0000000000000000000000000000000000000000..d8467460c4b36e54c83ce2dcd3ebe91d3432cad2
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/conv2d_layers.py
@@ -0,0 +1,304 @@
+""" Conv2D w/ SAME padding, CondConv, MixedConv
+
+A collection of conv layers and padding helpers needed by EfficientNet, MixNet, and
+MobileNetV3 models that maintain weight compatibility with original Tensorflow models.
+
+Copyright 2020 Ross Wightman
+"""
+import collections.abc
+import math
+from functools import partial
+from itertools import repeat
+from typing import Tuple, Optional
+
+import numpy as np
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+
+from .config import *
+
+
+# From PyTorch internals
+def _ntuple(n):
+ def parse(x):
+ if isinstance(x, collections.abc.Iterable):
+ return x
+ return tuple(repeat(x, n))
+ return parse
+
+
+_single = _ntuple(1)
+_pair = _ntuple(2)
+_triple = _ntuple(3)
+_quadruple = _ntuple(4)
+
+
+def _is_static_pad(kernel_size, stride=1, dilation=1, **_):
+ return stride == 1 and (dilation * (kernel_size - 1)) % 2 == 0
+
+
+def _get_padding(kernel_size, stride=1, dilation=1, **_):
+ padding = ((stride - 1) + dilation * (kernel_size - 1)) // 2
+ return padding
+
+
+def _calc_same_pad(i: int, k: int, s: int, d: int):
+ return max((-(i // -s) - 1) * s + (k - 1) * d + 1 - i, 0)
+
+
+def _same_pad_arg(input_size, kernel_size, stride, dilation):
+ ih, iw = input_size
+ kh, kw = kernel_size
+ pad_h = _calc_same_pad(ih, kh, stride[0], dilation[0])
+ pad_w = _calc_same_pad(iw, kw, stride[1], dilation[1])
+ return [pad_w // 2, pad_w - pad_w // 2, pad_h // 2, pad_h - pad_h // 2]
+
+
+def _split_channels(num_chan, num_groups):
+ split = [num_chan // num_groups for _ in range(num_groups)]
+ split[0] += num_chan - sum(split)
+ return split
+
+
+def conv2d_same(
+ x, weight: torch.Tensor, bias: Optional[torch.Tensor] = None, stride: Tuple[int, int] = (1, 1),
+ padding: Tuple[int, int] = (0, 0), dilation: Tuple[int, int] = (1, 1), groups: int = 1):
+ ih, iw = x.size()[-2:]
+ kh, kw = weight.size()[-2:]
+ pad_h = _calc_same_pad(ih, kh, stride[0], dilation[0])
+ pad_w = _calc_same_pad(iw, kw, stride[1], dilation[1])
+ x = F.pad(x, [pad_w // 2, pad_w - pad_w // 2, pad_h // 2, pad_h - pad_h // 2])
+ return F.conv2d(x, weight, bias, stride, (0, 0), dilation, groups)
+
+
+class Conv2dSame(nn.Conv2d):
+ """ Tensorflow like 'SAME' convolution wrapper for 2D convolutions
+ """
+
+ # pylint: disable=unused-argument
+ def __init__(self, in_channels, out_channels, kernel_size, stride=1,
+ padding=0, dilation=1, groups=1, bias=True):
+ super(Conv2dSame, self).__init__(
+ in_channels, out_channels, kernel_size, stride, 0, dilation, groups, bias)
+
+ def forward(self, x):
+ return conv2d_same(x, self.weight, self.bias, self.stride, self.padding, self.dilation, self.groups)
+
+
+class Conv2dSameExport(nn.Conv2d):
+ """ ONNX export friendly Tensorflow like 'SAME' convolution wrapper for 2D convolutions
+
+ NOTE: This does not currently work with torch.jit.script
+ """
+
+ # pylint: disable=unused-argument
+ def __init__(self, in_channels, out_channels, kernel_size, stride=1,
+ padding=0, dilation=1, groups=1, bias=True):
+ super(Conv2dSameExport, self).__init__(
+ in_channels, out_channels, kernel_size, stride, 0, dilation, groups, bias)
+ self.pad = None
+ self.pad_input_size = (0, 0)
+
+ def forward(self, x):
+ input_size = x.size()[-2:]
+ if self.pad is None:
+ pad_arg = _same_pad_arg(input_size, self.weight.size()[-2:], self.stride, self.dilation)
+ self.pad = nn.ZeroPad2d(pad_arg)
+ self.pad_input_size = input_size
+
+ if self.pad is not None:
+ x = self.pad(x)
+ return F.conv2d(
+ x, self.weight, self.bias, self.stride, self.padding, self.dilation, self.groups)
+
+
+def get_padding_value(padding, kernel_size, **kwargs):
+ dynamic = False
+ if isinstance(padding, str):
+ # for any string padding, the padding will be calculated for you, one of three ways
+ padding = padding.lower()
+ if padding == 'same':
+ # TF compatible 'SAME' padding, has a performance and GPU memory allocation impact
+ if _is_static_pad(kernel_size, **kwargs):
+ # static case, no extra overhead
+ padding = _get_padding(kernel_size, **kwargs)
+ else:
+ # dynamic padding
+ padding = 0
+ dynamic = True
+ elif padding == 'valid':
+ # 'VALID' padding, same as padding=0
+ padding = 0
+ else:
+ # Default to PyTorch style 'same'-ish symmetric padding
+ padding = _get_padding(kernel_size, **kwargs)
+ return padding, dynamic
+
+
+def create_conv2d_pad(in_chs, out_chs, kernel_size, **kwargs):
+ padding = kwargs.pop('padding', '')
+ kwargs.setdefault('bias', False)
+ padding, is_dynamic = get_padding_value(padding, kernel_size, **kwargs)
+ if is_dynamic:
+ if is_exportable():
+ assert not is_scriptable()
+ return Conv2dSameExport(in_chs, out_chs, kernel_size, **kwargs)
+ else:
+ return Conv2dSame(in_chs, out_chs, kernel_size, **kwargs)
+ else:
+ return nn.Conv2d(in_chs, out_chs, kernel_size, padding=padding, **kwargs)
+
+
+class MixedConv2d(nn.ModuleDict):
+ """ Mixed Grouped Convolution
+ Based on MDConv and GroupedConv in MixNet impl:
+ https://github.com/tensorflow/tpu/blob/master/models/official/mnasnet/mixnet/custom_layers.py
+ """
+
+ def __init__(self, in_channels, out_channels, kernel_size=3,
+ stride=1, padding='', dilation=1, depthwise=False, **kwargs):
+ super(MixedConv2d, self).__init__()
+
+ kernel_size = kernel_size if isinstance(kernel_size, list) else [kernel_size]
+ num_groups = len(kernel_size)
+ in_splits = _split_channels(in_channels, num_groups)
+ out_splits = _split_channels(out_channels, num_groups)
+ self.in_channels = sum(in_splits)
+ self.out_channels = sum(out_splits)
+ for idx, (k, in_ch, out_ch) in enumerate(zip(kernel_size, in_splits, out_splits)):
+ conv_groups = out_ch if depthwise else 1
+ self.add_module(
+ str(idx),
+ create_conv2d_pad(
+ in_ch, out_ch, k, stride=stride,
+ padding=padding, dilation=dilation, groups=conv_groups, **kwargs)
+ )
+ self.splits = in_splits
+
+ def forward(self, x):
+ x_split = torch.split(x, self.splits, 1)
+ x_out = [conv(x_split[i]) for i, conv in enumerate(self.values())]
+ x = torch.cat(x_out, 1)
+ return x
+
+
+def get_condconv_initializer(initializer, num_experts, expert_shape):
+ def condconv_initializer(weight):
+ """CondConv initializer function."""
+ num_params = np.prod(expert_shape)
+ if (len(weight.shape) != 2 or weight.shape[0] != num_experts or
+ weight.shape[1] != num_params):
+ raise (ValueError(
+ 'CondConv variables must have shape [num_experts, num_params]'))
+ for i in range(num_experts):
+ initializer(weight[i].view(expert_shape))
+ return condconv_initializer
+
+
+class CondConv2d(nn.Module):
+ """ Conditional Convolution
+ Inspired by: https://github.com/tensorflow/tpu/blob/master/models/official/efficientnet/condconv/condconv_layers.py
+
+ Grouped convolution hackery for parallel execution of the per-sample kernel filters inspired by this discussion:
+ https://github.com/pytorch/pytorch/issues/17983
+ """
+ __constants__ = ['bias', 'in_channels', 'out_channels', 'dynamic_padding']
+
+ def __init__(self, in_channels, out_channels, kernel_size=3,
+ stride=1, padding='', dilation=1, groups=1, bias=False, num_experts=4):
+ super(CondConv2d, self).__init__()
+
+ self.in_channels = in_channels
+ self.out_channels = out_channels
+ self.kernel_size = _pair(kernel_size)
+ self.stride = _pair(stride)
+ padding_val, is_padding_dynamic = get_padding_value(
+ padding, kernel_size, stride=stride, dilation=dilation)
+ self.dynamic_padding = is_padding_dynamic # if in forward to work with torchscript
+ self.padding = _pair(padding_val)
+ self.dilation = _pair(dilation)
+ self.groups = groups
+ self.num_experts = num_experts
+
+ self.weight_shape = (self.out_channels, self.in_channels // self.groups) + self.kernel_size
+ weight_num_param = 1
+ for wd in self.weight_shape:
+ weight_num_param *= wd
+ self.weight = torch.nn.Parameter(torch.Tensor(self.num_experts, weight_num_param))
+
+ if bias:
+ self.bias_shape = (self.out_channels,)
+ self.bias = torch.nn.Parameter(torch.Tensor(self.num_experts, self.out_channels))
+ else:
+ self.register_parameter('bias', None)
+
+ self.reset_parameters()
+
+ def reset_parameters(self):
+ init_weight = get_condconv_initializer(
+ partial(nn.init.kaiming_uniform_, a=math.sqrt(5)), self.num_experts, self.weight_shape)
+ init_weight(self.weight)
+ if self.bias is not None:
+ fan_in = np.prod(self.weight_shape[1:])
+ bound = 1 / math.sqrt(fan_in)
+ init_bias = get_condconv_initializer(
+ partial(nn.init.uniform_, a=-bound, b=bound), self.num_experts, self.bias_shape)
+ init_bias(self.bias)
+
+ def forward(self, x, routing_weights):
+ B, C, H, W = x.shape
+ weight = torch.matmul(routing_weights, self.weight)
+ new_weight_shape = (B * self.out_channels, self.in_channels // self.groups) + self.kernel_size
+ weight = weight.view(new_weight_shape)
+ bias = None
+ if self.bias is not None:
+ bias = torch.matmul(routing_weights, self.bias)
+ bias = bias.view(B * self.out_channels)
+ # move batch elements with channels so each batch element can be efficiently convolved with separate kernel
+ x = x.view(1, B * C, H, W)
+ if self.dynamic_padding:
+ out = conv2d_same(
+ x, weight, bias, stride=self.stride, padding=self.padding,
+ dilation=self.dilation, groups=self.groups * B)
+ else:
+ out = F.conv2d(
+ x, weight, bias, stride=self.stride, padding=self.padding,
+ dilation=self.dilation, groups=self.groups * B)
+ out = out.permute([1, 0, 2, 3]).view(B, self.out_channels, out.shape[-2], out.shape[-1])
+
+ # Literal port (from TF definition)
+ # x = torch.split(x, 1, 0)
+ # weight = torch.split(weight, 1, 0)
+ # if self.bias is not None:
+ # bias = torch.matmul(routing_weights, self.bias)
+ # bias = torch.split(bias, 1, 0)
+ # else:
+ # bias = [None] * B
+ # out = []
+ # for xi, wi, bi in zip(x, weight, bias):
+ # wi = wi.view(*self.weight_shape)
+ # if bi is not None:
+ # bi = bi.view(*self.bias_shape)
+ # out.append(self.conv_fn(
+ # xi, wi, bi, stride=self.stride, padding=self.padding,
+ # dilation=self.dilation, groups=self.groups))
+ # out = torch.cat(out, 0)
+ return out
+
+
+def select_conv2d(in_chs, out_chs, kernel_size, **kwargs):
+ assert 'groups' not in kwargs # only use 'depthwise' bool arg
+ if isinstance(kernel_size, list):
+ assert 'num_experts' not in kwargs # MixNet + CondConv combo not supported currently
+ # We're going to use only lists for defining the MixedConv2d kernel groups,
+ # ints, tuples, other iterables will continue to pass to normal conv and specify h, w.
+ m = MixedConv2d(in_chs, out_chs, kernel_size, **kwargs)
+ else:
+ depthwise = kwargs.pop('depthwise', False)
+ groups = out_chs if depthwise else 1
+ if 'num_experts' in kwargs and kwargs['num_experts'] > 0:
+ m = CondConv2d(in_chs, out_chs, kernel_size, groups=groups, **kwargs)
+ else:
+ m = create_conv2d_pad(in_chs, out_chs, kernel_size, groups=groups, **kwargs)
+ return m
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/efficientnet_builder.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/efficientnet_builder.py
new file mode 100644
index 0000000000000000000000000000000000000000..95dd63d400e70d70664c5a433a2772363f865e61
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/efficientnet_builder.py
@@ -0,0 +1,683 @@
+""" EfficientNet / MobileNetV3 Blocks and Builder
+
+Copyright 2020 Ross Wightman
+"""
+import re
+from copy import deepcopy
+
+from .conv2d_layers import *
+from geffnet.activations import *
+
+__all__ = ['get_bn_args_tf', 'resolve_bn_args', 'resolve_se_args', 'resolve_act_layer', 'make_divisible',
+ 'round_channels', 'drop_connect', 'SqueezeExcite', 'ConvBnAct', 'DepthwiseSeparableConv',
+ 'InvertedResidual', 'CondConvResidual', 'EdgeResidual', 'EfficientNetBuilder', 'decode_arch_def',
+ 'initialize_weight_default', 'initialize_weight_goog', 'BN_MOMENTUM_TF_DEFAULT', 'BN_EPS_TF_DEFAULT'
+]
+
+# Defaults used for Google/Tensorflow training of mobile networks /w RMSprop as per
+# papers and TF reference implementations. PT momentum equiv for TF decay is (1 - TF decay)
+# NOTE: momentum varies btw .99 and .9997 depending on source
+# .99 in official TF TPU impl
+# .9997 (/w .999 in search space) for paper
+#
+# PyTorch defaults are momentum = .1, eps = 1e-5
+#
+BN_MOMENTUM_TF_DEFAULT = 1 - 0.99
+BN_EPS_TF_DEFAULT = 1e-3
+_BN_ARGS_TF = dict(momentum=BN_MOMENTUM_TF_DEFAULT, eps=BN_EPS_TF_DEFAULT)
+
+
+def get_bn_args_tf():
+ return _BN_ARGS_TF.copy()
+
+
+def resolve_bn_args(kwargs):
+ bn_args = get_bn_args_tf() if kwargs.pop('bn_tf', False) else {}
+ bn_momentum = kwargs.pop('bn_momentum', None)
+ if bn_momentum is not None:
+ bn_args['momentum'] = bn_momentum
+ bn_eps = kwargs.pop('bn_eps', None)
+ if bn_eps is not None:
+ bn_args['eps'] = bn_eps
+ return bn_args
+
+
+_SE_ARGS_DEFAULT = dict(
+ gate_fn=sigmoid,
+ act_layer=None, # None == use containing block's activation layer
+ reduce_mid=False,
+ divisor=1)
+
+
+def resolve_se_args(kwargs, in_chs, act_layer=None):
+ se_kwargs = kwargs.copy() if kwargs is not None else {}
+ # fill in args that aren't specified with the defaults
+ for k, v in _SE_ARGS_DEFAULT.items():
+ se_kwargs.setdefault(k, v)
+ # some models, like MobilNetV3, calculate SE reduction chs from the containing block's mid_ch instead of in_ch
+ if not se_kwargs.pop('reduce_mid'):
+ se_kwargs['reduced_base_chs'] = in_chs
+ # act_layer override, if it remains None, the containing block's act_layer will be used
+ if se_kwargs['act_layer'] is None:
+ assert act_layer is not None
+ se_kwargs['act_layer'] = act_layer
+ return se_kwargs
+
+
+def resolve_act_layer(kwargs, default='relu'):
+ act_layer = kwargs.pop('act_layer', default)
+ if isinstance(act_layer, str):
+ act_layer = get_act_layer(act_layer)
+ return act_layer
+
+
+def make_divisible(v: int, divisor: int = 8, min_value: int = None):
+ min_value = min_value or divisor
+ new_v = max(min_value, int(v + divisor / 2) // divisor * divisor)
+ if new_v < 0.9 * v: # ensure round down does not go down by more than 10%.
+ new_v += divisor
+ return new_v
+
+
+def round_channels(channels, multiplier=1.0, divisor=8, channel_min=None):
+ """Round number of filters based on depth multiplier."""
+ if not multiplier:
+ return channels
+ channels *= multiplier
+ return make_divisible(channels, divisor, channel_min)
+
+
+def drop_connect(inputs, training: bool = False, drop_connect_rate: float = 0.):
+ """Apply drop connect."""
+ if not training:
+ return inputs
+
+ keep_prob = 1 - drop_connect_rate
+ random_tensor = keep_prob + torch.rand(
+ (inputs.size()[0], 1, 1, 1), dtype=inputs.dtype, device=inputs.device)
+ random_tensor.floor_() # binarize
+ output = inputs.div(keep_prob) * random_tensor
+ return output
+
+
+class SqueezeExcite(nn.Module):
+
+ def __init__(self, in_chs, se_ratio=0.25, reduced_base_chs=None, act_layer=nn.ReLU, gate_fn=sigmoid, divisor=1):
+ super(SqueezeExcite, self).__init__()
+ reduced_chs = make_divisible((reduced_base_chs or in_chs) * se_ratio, divisor)
+ self.conv_reduce = nn.Conv2d(in_chs, reduced_chs, 1, bias=True)
+ self.act1 = act_layer(inplace=True)
+ self.conv_expand = nn.Conv2d(reduced_chs, in_chs, 1, bias=True)
+ self.gate_fn = gate_fn
+
+ def forward(self, x):
+ x_se = x.mean((2, 3), keepdim=True)
+ x_se = self.conv_reduce(x_se)
+ x_se = self.act1(x_se)
+ x_se = self.conv_expand(x_se)
+ x = x * self.gate_fn(x_se)
+ return x
+
+
+class ConvBnAct(nn.Module):
+ def __init__(self, in_chs, out_chs, kernel_size,
+ stride=1, pad_type='', act_layer=nn.ReLU, norm_layer=nn.BatchNorm2d, norm_kwargs=None):
+ super(ConvBnAct, self).__init__()
+ assert stride in [1, 2]
+ norm_kwargs = norm_kwargs or {}
+ self.conv = select_conv2d(in_chs, out_chs, kernel_size, stride=stride, padding=pad_type)
+ self.bn1 = norm_layer(out_chs, **norm_kwargs)
+ self.act1 = act_layer(inplace=True)
+
+ def forward(self, x):
+ x = self.conv(x)
+ x = self.bn1(x)
+ x = self.act1(x)
+ return x
+
+
+class DepthwiseSeparableConv(nn.Module):
+ """ DepthwiseSeparable block
+ Used for DS convs in MobileNet-V1 and in the place of IR blocks with an expansion
+ factor of 1.0. This is an alternative to having a IR with optional first pw conv.
+ """
+ def __init__(self, in_chs, out_chs, dw_kernel_size=3,
+ stride=1, pad_type='', act_layer=nn.ReLU, noskip=False,
+ pw_kernel_size=1, pw_act=False, se_ratio=0., se_kwargs=None,
+ norm_layer=nn.BatchNorm2d, norm_kwargs=None, drop_connect_rate=0.):
+ super(DepthwiseSeparableConv, self).__init__()
+ assert stride in [1, 2]
+ norm_kwargs = norm_kwargs or {}
+ self.has_residual = (stride == 1 and in_chs == out_chs) and not noskip
+ self.drop_connect_rate = drop_connect_rate
+
+ self.conv_dw = select_conv2d(
+ in_chs, in_chs, dw_kernel_size, stride=stride, padding=pad_type, depthwise=True)
+ self.bn1 = norm_layer(in_chs, **norm_kwargs)
+ self.act1 = act_layer(inplace=True)
+
+ # Squeeze-and-excitation
+ if se_ratio is not None and se_ratio > 0.:
+ se_kwargs = resolve_se_args(se_kwargs, in_chs, act_layer)
+ self.se = SqueezeExcite(in_chs, se_ratio=se_ratio, **se_kwargs)
+ else:
+ self.se = nn.Identity()
+
+ self.conv_pw = select_conv2d(in_chs, out_chs, pw_kernel_size, padding=pad_type)
+ self.bn2 = norm_layer(out_chs, **norm_kwargs)
+ self.act2 = act_layer(inplace=True) if pw_act else nn.Identity()
+
+ def forward(self, x):
+ residual = x
+
+ x = self.conv_dw(x)
+ x = self.bn1(x)
+ x = self.act1(x)
+
+ x = self.se(x)
+
+ x = self.conv_pw(x)
+ x = self.bn2(x)
+ x = self.act2(x)
+
+ if self.has_residual:
+ if self.drop_connect_rate > 0.:
+ x = drop_connect(x, self.training, self.drop_connect_rate)
+ x += residual
+ return x
+
+
+class InvertedResidual(nn.Module):
+ """ Inverted residual block w/ optional SE"""
+
+ def __init__(self, in_chs, out_chs, dw_kernel_size=3,
+ stride=1, pad_type='', act_layer=nn.ReLU, noskip=False,
+ exp_ratio=1.0, exp_kernel_size=1, pw_kernel_size=1,
+ se_ratio=0., se_kwargs=None, norm_layer=nn.BatchNorm2d, norm_kwargs=None,
+ conv_kwargs=None, drop_connect_rate=0.):
+ super(InvertedResidual, self).__init__()
+ norm_kwargs = norm_kwargs or {}
+ conv_kwargs = conv_kwargs or {}
+ mid_chs: int = make_divisible(in_chs * exp_ratio)
+ self.has_residual = (in_chs == out_chs and stride == 1) and not noskip
+ self.drop_connect_rate = drop_connect_rate
+
+ # Point-wise expansion
+ self.conv_pw = select_conv2d(in_chs, mid_chs, exp_kernel_size, padding=pad_type, **conv_kwargs)
+ self.bn1 = norm_layer(mid_chs, **norm_kwargs)
+ self.act1 = act_layer(inplace=True)
+
+ # Depth-wise convolution
+ self.conv_dw = select_conv2d(
+ mid_chs, mid_chs, dw_kernel_size, stride=stride, padding=pad_type, depthwise=True, **conv_kwargs)
+ self.bn2 = norm_layer(mid_chs, **norm_kwargs)
+ self.act2 = act_layer(inplace=True)
+
+ # Squeeze-and-excitation
+ if se_ratio is not None and se_ratio > 0.:
+ se_kwargs = resolve_se_args(se_kwargs, in_chs, act_layer)
+ self.se = SqueezeExcite(mid_chs, se_ratio=se_ratio, **se_kwargs)
+ else:
+ self.se = nn.Identity() # for jit.script compat
+
+ # Point-wise linear projection
+ self.conv_pwl = select_conv2d(mid_chs, out_chs, pw_kernel_size, padding=pad_type, **conv_kwargs)
+ self.bn3 = norm_layer(out_chs, **norm_kwargs)
+
+ def forward(self, x):
+ residual = x
+
+ # Point-wise expansion
+ x = self.conv_pw(x)
+ x = self.bn1(x)
+ x = self.act1(x)
+
+ # Depth-wise convolution
+ x = self.conv_dw(x)
+ x = self.bn2(x)
+ x = self.act2(x)
+
+ # Squeeze-and-excitation
+ x = self.se(x)
+
+ # Point-wise linear projection
+ x = self.conv_pwl(x)
+ x = self.bn3(x)
+
+ if self.has_residual:
+ if self.drop_connect_rate > 0.:
+ x = drop_connect(x, self.training, self.drop_connect_rate)
+ x += residual
+ return x
+
+
+class CondConvResidual(InvertedResidual):
+ """ Inverted residual block w/ CondConv routing"""
+
+ def __init__(self, in_chs, out_chs, dw_kernel_size=3,
+ stride=1, pad_type='', act_layer=nn.ReLU, noskip=False,
+ exp_ratio=1.0, exp_kernel_size=1, pw_kernel_size=1,
+ se_ratio=0., se_kwargs=None, norm_layer=nn.BatchNorm2d, norm_kwargs=None,
+ num_experts=0, drop_connect_rate=0.):
+
+ self.num_experts = num_experts
+ conv_kwargs = dict(num_experts=self.num_experts)
+
+ super(CondConvResidual, self).__init__(
+ in_chs, out_chs, dw_kernel_size=dw_kernel_size, stride=stride, pad_type=pad_type,
+ act_layer=act_layer, noskip=noskip, exp_ratio=exp_ratio, exp_kernel_size=exp_kernel_size,
+ pw_kernel_size=pw_kernel_size, se_ratio=se_ratio, se_kwargs=se_kwargs,
+ norm_layer=norm_layer, norm_kwargs=norm_kwargs, conv_kwargs=conv_kwargs,
+ drop_connect_rate=drop_connect_rate)
+
+ self.routing_fn = nn.Linear(in_chs, self.num_experts)
+
+ def forward(self, x):
+ residual = x
+
+ # CondConv routing
+ pooled_inputs = F.adaptive_avg_pool2d(x, 1).flatten(1)
+ routing_weights = torch.sigmoid(self.routing_fn(pooled_inputs))
+
+ # Point-wise expansion
+ x = self.conv_pw(x, routing_weights)
+ x = self.bn1(x)
+ x = self.act1(x)
+
+ # Depth-wise convolution
+ x = self.conv_dw(x, routing_weights)
+ x = self.bn2(x)
+ x = self.act2(x)
+
+ # Squeeze-and-excitation
+ x = self.se(x)
+
+ # Point-wise linear projection
+ x = self.conv_pwl(x, routing_weights)
+ x = self.bn3(x)
+
+ if self.has_residual:
+ if self.drop_connect_rate > 0.:
+ x = drop_connect(x, self.training, self.drop_connect_rate)
+ x += residual
+ return x
+
+
+class EdgeResidual(nn.Module):
+ """ EdgeTPU Residual block with expansion convolution followed by pointwise-linear w/ stride"""
+
+ def __init__(self, in_chs, out_chs, exp_kernel_size=3, exp_ratio=1.0, fake_in_chs=0,
+ stride=1, pad_type='', act_layer=nn.ReLU, noskip=False, pw_kernel_size=1,
+ se_ratio=0., se_kwargs=None, norm_layer=nn.BatchNorm2d, norm_kwargs=None, drop_connect_rate=0.):
+ super(EdgeResidual, self).__init__()
+ norm_kwargs = norm_kwargs or {}
+ mid_chs = make_divisible(fake_in_chs * exp_ratio) if fake_in_chs > 0 else make_divisible(in_chs * exp_ratio)
+ self.has_residual = (in_chs == out_chs and stride == 1) and not noskip
+ self.drop_connect_rate = drop_connect_rate
+
+ # Expansion convolution
+ self.conv_exp = select_conv2d(in_chs, mid_chs, exp_kernel_size, padding=pad_type)
+ self.bn1 = norm_layer(mid_chs, **norm_kwargs)
+ self.act1 = act_layer(inplace=True)
+
+ # Squeeze-and-excitation
+ if se_ratio is not None and se_ratio > 0.:
+ se_kwargs = resolve_se_args(se_kwargs, in_chs, act_layer)
+ self.se = SqueezeExcite(mid_chs, se_ratio=se_ratio, **se_kwargs)
+ else:
+ self.se = nn.Identity()
+
+ # Point-wise linear projection
+ self.conv_pwl = select_conv2d(mid_chs, out_chs, pw_kernel_size, stride=stride, padding=pad_type)
+ self.bn2 = nn.BatchNorm2d(out_chs, **norm_kwargs)
+
+ def forward(self, x):
+ residual = x
+
+ # Expansion convolution
+ x = self.conv_exp(x)
+ x = self.bn1(x)
+ x = self.act1(x)
+
+ # Squeeze-and-excitation
+ x = self.se(x)
+
+ # Point-wise linear projection
+ x = self.conv_pwl(x)
+ x = self.bn2(x)
+
+ if self.has_residual:
+ if self.drop_connect_rate > 0.:
+ x = drop_connect(x, self.training, self.drop_connect_rate)
+ x += residual
+
+ return x
+
+
+class EfficientNetBuilder:
+ """ Build Trunk Blocks for Efficient/Mobile Networks
+
+ This ended up being somewhat of a cross between
+ https://github.com/tensorflow/tpu/blob/master/models/official/mnasnet/mnasnet_models.py
+ and
+ https://github.com/facebookresearch/maskrcnn-benchmark/blob/master/maskrcnn_benchmark/modeling/backbone/fbnet_builder.py
+
+ """
+
+ def __init__(self, channel_multiplier=1.0, channel_divisor=8, channel_min=None,
+ pad_type='', act_layer=None, se_kwargs=None,
+ norm_layer=nn.BatchNorm2d, norm_kwargs=None, drop_connect_rate=0.):
+ self.channel_multiplier = channel_multiplier
+ self.channel_divisor = channel_divisor
+ self.channel_min = channel_min
+ self.pad_type = pad_type
+ self.act_layer = act_layer
+ self.se_kwargs = se_kwargs
+ self.norm_layer = norm_layer
+ self.norm_kwargs = norm_kwargs
+ self.drop_connect_rate = drop_connect_rate
+
+ # updated during build
+ self.in_chs = None
+ self.block_idx = 0
+ self.block_count = 0
+
+ def _round_channels(self, chs):
+ return round_channels(chs, self.channel_multiplier, self.channel_divisor, self.channel_min)
+
+ def _make_block(self, ba):
+ bt = ba.pop('block_type')
+ ba['in_chs'] = self.in_chs
+ ba['out_chs'] = self._round_channels(ba['out_chs'])
+ if 'fake_in_chs' in ba and ba['fake_in_chs']:
+ # FIXME this is a hack to work around mismatch in origin impl input filters for EdgeTPU
+ ba['fake_in_chs'] = self._round_channels(ba['fake_in_chs'])
+ ba['norm_layer'] = self.norm_layer
+ ba['norm_kwargs'] = self.norm_kwargs
+ ba['pad_type'] = self.pad_type
+ # block act fn overrides the model default
+ ba['act_layer'] = ba['act_layer'] if ba['act_layer'] is not None else self.act_layer
+ assert ba['act_layer'] is not None
+ if bt == 'ir':
+ ba['drop_connect_rate'] = self.drop_connect_rate * self.block_idx / self.block_count
+ ba['se_kwargs'] = self.se_kwargs
+ if ba.get('num_experts', 0) > 0:
+ block = CondConvResidual(**ba)
+ else:
+ block = InvertedResidual(**ba)
+ elif bt == 'ds' or bt == 'dsa':
+ ba['drop_connect_rate'] = self.drop_connect_rate * self.block_idx / self.block_count
+ ba['se_kwargs'] = self.se_kwargs
+ block = DepthwiseSeparableConv(**ba)
+ elif bt == 'er':
+ ba['drop_connect_rate'] = self.drop_connect_rate * self.block_idx / self.block_count
+ ba['se_kwargs'] = self.se_kwargs
+ block = EdgeResidual(**ba)
+ elif bt == 'cn':
+ block = ConvBnAct(**ba)
+ else:
+ assert False, 'Uknkown block type (%s) while building model.' % bt
+ self.in_chs = ba['out_chs'] # update in_chs for arg of next block
+ return block
+
+ def _make_stack(self, stack_args):
+ blocks = []
+ # each stack (stage) contains a list of block arguments
+ for i, ba in enumerate(stack_args):
+ if i >= 1:
+ # only the first block in any stack can have a stride > 1
+ ba['stride'] = 1
+ block = self._make_block(ba)
+ blocks.append(block)
+ self.block_idx += 1 # incr global idx (across all stacks)
+ return nn.Sequential(*blocks)
+
+ def __call__(self, in_chs, block_args):
+ """ Build the blocks
+ Args:
+ in_chs: Number of input-channels passed to first block
+ block_args: A list of lists, outer list defines stages, inner
+ list contains strings defining block configuration(s)
+ Return:
+ List of block stacks (each stack wrapped in nn.Sequential)
+ """
+ self.in_chs = in_chs
+ self.block_count = sum([len(x) for x in block_args])
+ self.block_idx = 0
+ blocks = []
+ # outer list of block_args defines the stacks ('stages' by some conventions)
+ for stack_idx, stack in enumerate(block_args):
+ assert isinstance(stack, list)
+ stack = self._make_stack(stack)
+ blocks.append(stack)
+ return blocks
+
+
+def _parse_ksize(ss):
+ if ss.isdigit():
+ return int(ss)
+ else:
+ return [int(k) for k in ss.split('.')]
+
+
+def _decode_block_str(block_str):
+ """ Decode block definition string
+
+ Gets a list of block arg (dicts) through a string notation of arguments.
+ E.g. ir_r2_k3_s2_e1_i32_o16_se0.25_noskip
+
+ All args can exist in any order with the exception of the leading string which
+ is assumed to indicate the block type.
+
+ leading string - block type (
+ ir = InvertedResidual, ds = DepthwiseSep, dsa = DeptwhiseSep with pw act, cn = ConvBnAct)
+ r - number of repeat blocks,
+ k - kernel size,
+ s - strides (1-9),
+ e - expansion ratio,
+ c - output channels,
+ se - squeeze/excitation ratio
+ n - activation fn ('re', 'r6', 'hs', or 'sw')
+ Args:
+ block_str: a string representation of block arguments.
+ Returns:
+ A list of block args (dicts)
+ Raises:
+ ValueError: if the string def not properly specified (TODO)
+ """
+ assert isinstance(block_str, str)
+ ops = block_str.split('_')
+ block_type = ops[0] # take the block type off the front
+ ops = ops[1:]
+ options = {}
+ noskip = False
+ for op in ops:
+ # string options being checked on individual basis, combine if they grow
+ if op == 'noskip':
+ noskip = True
+ elif op.startswith('n'):
+ # activation fn
+ key = op[0]
+ v = op[1:]
+ if v == 're':
+ value = get_act_layer('relu')
+ elif v == 'r6':
+ value = get_act_layer('relu6')
+ elif v == 'hs':
+ value = get_act_layer('hard_swish')
+ elif v == 'sw':
+ value = get_act_layer('swish')
+ else:
+ continue
+ options[key] = value
+ else:
+ # all numeric options
+ splits = re.split(r'(\d.*)', op)
+ if len(splits) >= 2:
+ key, value = splits[:2]
+ options[key] = value
+
+ # if act_layer is None, the model default (passed to model init) will be used
+ act_layer = options['n'] if 'n' in options else None
+ exp_kernel_size = _parse_ksize(options['a']) if 'a' in options else 1
+ pw_kernel_size = _parse_ksize(options['p']) if 'p' in options else 1
+ fake_in_chs = int(options['fc']) if 'fc' in options else 0 # FIXME hack to deal with in_chs issue in TPU def
+
+ num_repeat = int(options['r'])
+ # each type of block has different valid arguments, fill accordingly
+ if block_type == 'ir':
+ block_args = dict(
+ block_type=block_type,
+ dw_kernel_size=_parse_ksize(options['k']),
+ exp_kernel_size=exp_kernel_size,
+ pw_kernel_size=pw_kernel_size,
+ out_chs=int(options['c']),
+ exp_ratio=float(options['e']),
+ se_ratio=float(options['se']) if 'se' in options else None,
+ stride=int(options['s']),
+ act_layer=act_layer,
+ noskip=noskip,
+ )
+ if 'cc' in options:
+ block_args['num_experts'] = int(options['cc'])
+ elif block_type == 'ds' or block_type == 'dsa':
+ block_args = dict(
+ block_type=block_type,
+ dw_kernel_size=_parse_ksize(options['k']),
+ pw_kernel_size=pw_kernel_size,
+ out_chs=int(options['c']),
+ se_ratio=float(options['se']) if 'se' in options else None,
+ stride=int(options['s']),
+ act_layer=act_layer,
+ pw_act=block_type == 'dsa',
+ noskip=block_type == 'dsa' or noskip,
+ )
+ elif block_type == 'er':
+ block_args = dict(
+ block_type=block_type,
+ exp_kernel_size=_parse_ksize(options['k']),
+ pw_kernel_size=pw_kernel_size,
+ out_chs=int(options['c']),
+ exp_ratio=float(options['e']),
+ fake_in_chs=fake_in_chs,
+ se_ratio=float(options['se']) if 'se' in options else None,
+ stride=int(options['s']),
+ act_layer=act_layer,
+ noskip=noskip,
+ )
+ elif block_type == 'cn':
+ block_args = dict(
+ block_type=block_type,
+ kernel_size=int(options['k']),
+ out_chs=int(options['c']),
+ stride=int(options['s']),
+ act_layer=act_layer,
+ )
+ else:
+ assert False, 'Unknown block type (%s)' % block_type
+
+ return block_args, num_repeat
+
+
+def _scale_stage_depth(stack_args, repeats, depth_multiplier=1.0, depth_trunc='ceil'):
+ """ Per-stage depth scaling
+ Scales the block repeats in each stage. This depth scaling impl maintains
+ compatibility with the EfficientNet scaling method, while allowing sensible
+ scaling for other models that may have multiple block arg definitions in each stage.
+ """
+
+ # We scale the total repeat count for each stage, there may be multiple
+ # block arg defs per stage so we need to sum.
+ num_repeat = sum(repeats)
+ if depth_trunc == 'round':
+ # Truncating to int by rounding allows stages with few repeats to remain
+ # proportionally smaller for longer. This is a good choice when stage definitions
+ # include single repeat stages that we'd prefer to keep that way as long as possible
+ num_repeat_scaled = max(1, round(num_repeat * depth_multiplier))
+ else:
+ # The default for EfficientNet truncates repeats to int via 'ceil'.
+ # Any multiplier > 1.0 will result in an increased depth for every stage.
+ num_repeat_scaled = int(math.ceil(num_repeat * depth_multiplier))
+
+ # Proportionally distribute repeat count scaling to each block definition in the stage.
+ # Allocation is done in reverse as it results in the first block being less likely to be scaled.
+ # The first block makes less sense to repeat in most of the arch definitions.
+ repeats_scaled = []
+ for r in repeats[::-1]:
+ rs = max(1, round((r / num_repeat * num_repeat_scaled)))
+ repeats_scaled.append(rs)
+ num_repeat -= r
+ num_repeat_scaled -= rs
+ repeats_scaled = repeats_scaled[::-1]
+
+ # Apply the calculated scaling to each block arg in the stage
+ sa_scaled = []
+ for ba, rep in zip(stack_args, repeats_scaled):
+ sa_scaled.extend([deepcopy(ba) for _ in range(rep)])
+ return sa_scaled
+
+
+def decode_arch_def(arch_def, depth_multiplier=1.0, depth_trunc='ceil', experts_multiplier=1, fix_first_last=False):
+ arch_args = []
+ for stack_idx, block_strings in enumerate(arch_def):
+ assert isinstance(block_strings, list)
+ stack_args = []
+ repeats = []
+ for block_str in block_strings:
+ assert isinstance(block_str, str)
+ ba, rep = _decode_block_str(block_str)
+ if ba.get('num_experts', 0) > 0 and experts_multiplier > 1:
+ ba['num_experts'] *= experts_multiplier
+ stack_args.append(ba)
+ repeats.append(rep)
+ if fix_first_last and (stack_idx == 0 or stack_idx == len(arch_def) - 1):
+ arch_args.append(_scale_stage_depth(stack_args, repeats, 1.0, depth_trunc))
+ else:
+ arch_args.append(_scale_stage_depth(stack_args, repeats, depth_multiplier, depth_trunc))
+ return arch_args
+
+
+def initialize_weight_goog(m, n='', fix_group_fanout=True):
+ # weight init as per Tensorflow Official impl
+ # https://github.com/tensorflow/tpu/blob/master/models/official/mnasnet/mnasnet_model.py
+ if isinstance(m, CondConv2d):
+ fan_out = m.kernel_size[0] * m.kernel_size[1] * m.out_channels
+ if fix_group_fanout:
+ fan_out //= m.groups
+ init_weight_fn = get_condconv_initializer(
+ lambda w: w.data.normal_(0, math.sqrt(2.0 / fan_out)), m.num_experts, m.weight_shape)
+ init_weight_fn(m.weight)
+ if m.bias is not None:
+ m.bias.data.zero_()
+ elif isinstance(m, nn.Conv2d):
+ fan_out = m.kernel_size[0] * m.kernel_size[1] * m.out_channels
+ if fix_group_fanout:
+ fan_out //= m.groups
+ m.weight.data.normal_(0, math.sqrt(2.0 / fan_out))
+ if m.bias is not None:
+ m.bias.data.zero_()
+ elif isinstance(m, nn.BatchNorm2d):
+ m.weight.data.fill_(1.0)
+ m.bias.data.zero_()
+ elif isinstance(m, nn.Linear):
+ fan_out = m.weight.size(0) # fan-out
+ fan_in = 0
+ if 'routing_fn' in n:
+ fan_in = m.weight.size(1)
+ init_range = 1.0 / math.sqrt(fan_in + fan_out)
+ m.weight.data.uniform_(-init_range, init_range)
+ m.bias.data.zero_()
+
+
+def initialize_weight_default(m, n=''):
+ if isinstance(m, CondConv2d):
+ init_fn = get_condconv_initializer(partial(
+ nn.init.kaiming_normal_, mode='fan_out', nonlinearity='relu'), m.num_experts, m.weight_shape)
+ init_fn(m.weight)
+ elif isinstance(m, nn.Conv2d):
+ nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu')
+ elif isinstance(m, nn.BatchNorm2d):
+ m.weight.data.fill_(1.0)
+ m.bias.data.zero_()
+ elif isinstance(m, nn.Linear):
+ nn.init.kaiming_uniform_(m.weight, mode='fan_in', nonlinearity='linear')
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/gen_efficientnet.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/gen_efficientnet.py
new file mode 100644
index 0000000000000000000000000000000000000000..cd170d4cc5bed6ca82b61539902b470d3320c691
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/gen_efficientnet.py
@@ -0,0 +1,1450 @@
+""" Generic Efficient Networks
+
+A generic MobileNet class with building blocks to support a variety of models:
+
+* EfficientNet (B0-B8, L2 + Tensorflow pretrained AutoAug/RandAug/AdvProp/NoisyStudent ports)
+ - EfficientNet: Rethinking Model Scaling for CNNs - https://arxiv.org/abs/1905.11946
+ - CondConv: Conditionally Parameterized Convolutions for Efficient Inference - https://arxiv.org/abs/1904.04971
+ - Adversarial Examples Improve Image Recognition - https://arxiv.org/abs/1911.09665
+ - Self-training with Noisy Student improves ImageNet classification - https://arxiv.org/abs/1911.04252
+
+* EfficientNet-Lite
+
+* MixNet (Small, Medium, and Large)
+ - MixConv: Mixed Depthwise Convolutional Kernels - https://arxiv.org/abs/1907.09595
+
+* MNasNet B1, A1 (SE), Small
+ - MnasNet: Platform-Aware Neural Architecture Search for Mobile - https://arxiv.org/abs/1807.11626
+
+* FBNet-C
+ - FBNet: Hardware-Aware Efficient ConvNet Design via Differentiable NAS - https://arxiv.org/abs/1812.03443
+
+* Single-Path NAS Pixel1
+ - Single-Path NAS: Designing Hardware-Efficient ConvNets - https://arxiv.org/abs/1904.02877
+
+* And likely more...
+
+Hacked together by / Copyright 2020 Ross Wightman
+"""
+import torch.nn as nn
+import torch.nn.functional as F
+
+from .config import layer_config_kwargs, is_scriptable
+from .conv2d_layers import select_conv2d
+from .helpers import load_pretrained
+from .efficientnet_builder import *
+
+__all__ = ['GenEfficientNet', 'mnasnet_050', 'mnasnet_075', 'mnasnet_100', 'mnasnet_b1', 'mnasnet_140',
+ 'semnasnet_050', 'semnasnet_075', 'semnasnet_100', 'mnasnet_a1', 'semnasnet_140', 'mnasnet_small',
+ 'mobilenetv2_100', 'mobilenetv2_140', 'mobilenetv2_110d', 'mobilenetv2_120d',
+ 'fbnetc_100', 'spnasnet_100', 'efficientnet_b0', 'efficientnet_b1', 'efficientnet_b2', 'efficientnet_b3',
+ 'efficientnet_b4', 'efficientnet_b5', 'efficientnet_b6', 'efficientnet_b7', 'efficientnet_b8',
+ 'efficientnet_l2', 'efficientnet_es', 'efficientnet_em', 'efficientnet_el',
+ 'efficientnet_cc_b0_4e', 'efficientnet_cc_b0_8e', 'efficientnet_cc_b1_8e',
+ 'efficientnet_lite0', 'efficientnet_lite1', 'efficientnet_lite2', 'efficientnet_lite3', 'efficientnet_lite4',
+ 'tf_efficientnet_b0', 'tf_efficientnet_b1', 'tf_efficientnet_b2', 'tf_efficientnet_b3',
+ 'tf_efficientnet_b4', 'tf_efficientnet_b5', 'tf_efficientnet_b6', 'tf_efficientnet_b7', 'tf_efficientnet_b8',
+ 'tf_efficientnet_b0_ap', 'tf_efficientnet_b1_ap', 'tf_efficientnet_b2_ap', 'tf_efficientnet_b3_ap',
+ 'tf_efficientnet_b4_ap', 'tf_efficientnet_b5_ap', 'tf_efficientnet_b6_ap', 'tf_efficientnet_b7_ap',
+ 'tf_efficientnet_b8_ap', 'tf_efficientnet_b0_ns', 'tf_efficientnet_b1_ns', 'tf_efficientnet_b2_ns',
+ 'tf_efficientnet_b3_ns', 'tf_efficientnet_b4_ns', 'tf_efficientnet_b5_ns', 'tf_efficientnet_b6_ns',
+ 'tf_efficientnet_b7_ns', 'tf_efficientnet_l2_ns', 'tf_efficientnet_l2_ns_475',
+ 'tf_efficientnet_es', 'tf_efficientnet_em', 'tf_efficientnet_el',
+ 'tf_efficientnet_cc_b0_4e', 'tf_efficientnet_cc_b0_8e', 'tf_efficientnet_cc_b1_8e',
+ 'tf_efficientnet_lite0', 'tf_efficientnet_lite1', 'tf_efficientnet_lite2', 'tf_efficientnet_lite3',
+ 'tf_efficientnet_lite4',
+ 'mixnet_s', 'mixnet_m', 'mixnet_l', 'mixnet_xl', 'tf_mixnet_s', 'tf_mixnet_m', 'tf_mixnet_l']
+
+
+model_urls = {
+ 'mnasnet_050': None,
+ 'mnasnet_075': None,
+ 'mnasnet_100':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/mnasnet_b1-74cb7081.pth',
+ 'mnasnet_140': None,
+ 'mnasnet_small': None,
+
+ 'semnasnet_050': None,
+ 'semnasnet_075': None,
+ 'semnasnet_100':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/mnasnet_a1-d9418771.pth',
+ 'semnasnet_140': None,
+
+ 'mobilenetv2_100':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/mobilenetv2_100_ra-b33bc2c4.pth',
+ 'mobilenetv2_110d':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/mobilenetv2_110d_ra-77090ade.pth',
+ 'mobilenetv2_120d':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/mobilenetv2_120d_ra-5987e2ed.pth',
+ 'mobilenetv2_140':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/mobilenetv2_140_ra-21a4e913.pth',
+
+ 'fbnetc_100':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/fbnetc_100-c345b898.pth',
+ 'spnasnet_100':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/spnasnet_100-048bc3f4.pth',
+
+ 'efficientnet_b0':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/efficientnet_b0_ra-3dd342df.pth',
+ 'efficientnet_b1':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/efficientnet_b1-533bc792.pth',
+ 'efficientnet_b2':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/efficientnet_b2_ra-bcdf34b7.pth',
+ 'efficientnet_b3':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/efficientnet_b3_ra2-cf984f9c.pth',
+ 'efficientnet_b4': None,
+ 'efficientnet_b5': None,
+ 'efficientnet_b6': None,
+ 'efficientnet_b7': None,
+ 'efficientnet_b8': None,
+ 'efficientnet_l2': None,
+
+ 'efficientnet_es':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/efficientnet_es_ra-f111e99c.pth',
+ 'efficientnet_em': None,
+ 'efficientnet_el': None,
+
+ 'efficientnet_cc_b0_4e': None,
+ 'efficientnet_cc_b0_8e': None,
+ 'efficientnet_cc_b1_8e': None,
+
+ 'efficientnet_lite0': 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/efficientnet_lite0_ra-37913777.pth',
+ 'efficientnet_lite1': None,
+ 'efficientnet_lite2': None,
+ 'efficientnet_lite3': None,
+ 'efficientnet_lite4': None,
+
+ 'tf_efficientnet_b0':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_b0_aa-827b6e33.pth',
+ 'tf_efficientnet_b1':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_b1_aa-ea7a6ee0.pth',
+ 'tf_efficientnet_b2':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_b2_aa-60c94f97.pth',
+ 'tf_efficientnet_b3':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_b3_aa-84b4657e.pth',
+ 'tf_efficientnet_b4':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_b4_aa-818f208c.pth',
+ 'tf_efficientnet_b5':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_b5_ra-9a3e5369.pth',
+ 'tf_efficientnet_b6':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_b6_aa-80ba17e4.pth',
+ 'tf_efficientnet_b7':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_b7_ra-6c08e654.pth',
+ 'tf_efficientnet_b8':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_b8_ra-572d5dd9.pth',
+
+ 'tf_efficientnet_b0_ap':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_b0_ap-f262efe1.pth',
+ 'tf_efficientnet_b1_ap':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_b1_ap-44ef0a3d.pth',
+ 'tf_efficientnet_b2_ap':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_b2_ap-2f8e7636.pth',
+ 'tf_efficientnet_b3_ap':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_b3_ap-aad25bdd.pth',
+ 'tf_efficientnet_b4_ap':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_b4_ap-dedb23e6.pth',
+ 'tf_efficientnet_b5_ap':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_b5_ap-9e82fae8.pth',
+ 'tf_efficientnet_b6_ap':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_b6_ap-4ffb161f.pth',
+ 'tf_efficientnet_b7_ap':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_b7_ap-ddb28fec.pth',
+ 'tf_efficientnet_b8_ap':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_b8_ap-00e169fa.pth',
+
+ 'tf_efficientnet_b0_ns':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_b0_ns-c0e6a31c.pth',
+ 'tf_efficientnet_b1_ns':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_b1_ns-99dd0c41.pth',
+ 'tf_efficientnet_b2_ns':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_b2_ns-00306e48.pth',
+ 'tf_efficientnet_b3_ns':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_b3_ns-9d44bf68.pth',
+ 'tf_efficientnet_b4_ns':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_b4_ns-d6313a46.pth',
+ 'tf_efficientnet_b5_ns':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_b5_ns-6f26d0cf.pth',
+ 'tf_efficientnet_b6_ns':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_b6_ns-51548356.pth',
+ 'tf_efficientnet_b7_ns':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_b7_ns-1dbc32de.pth',
+ 'tf_efficientnet_l2_ns_475':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_l2_ns_475-bebbd00a.pth',
+ 'tf_efficientnet_l2_ns':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_l2_ns-df73bb44.pth',
+
+ 'tf_efficientnet_es':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_es-ca1afbfe.pth',
+ 'tf_efficientnet_em':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_em-e78cfe58.pth',
+ 'tf_efficientnet_el':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_el-5143854e.pth',
+
+ 'tf_efficientnet_cc_b0_4e':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_cc_b0_4e-4362b6b2.pth',
+ 'tf_efficientnet_cc_b0_8e':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_cc_b0_8e-66184a25.pth',
+ 'tf_efficientnet_cc_b1_8e':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_cc_b1_8e-f7c79ae1.pth',
+
+ 'tf_efficientnet_lite0':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_lite0-0aa007d2.pth',
+ 'tf_efficientnet_lite1':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_lite1-bde8b488.pth',
+ 'tf_efficientnet_lite2':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_lite2-dcccb7df.pth',
+ 'tf_efficientnet_lite3':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_lite3-b733e338.pth',
+ 'tf_efficientnet_lite4':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_efficientnet_lite4-741542c3.pth',
+
+ 'mixnet_s': 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/mixnet_s-a907afbc.pth',
+ 'mixnet_m': 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/mixnet_m-4647fc68.pth',
+ 'mixnet_l': 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/mixnet_l-5a9a2ed8.pth',
+ 'mixnet_xl': 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/mixnet_xl_ra-aac3c00c.pth',
+
+ 'tf_mixnet_s':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_mixnet_s-89d3354b.pth',
+ 'tf_mixnet_m':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_mixnet_m-0f4d8805.pth',
+ 'tf_mixnet_l':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_mixnet_l-6c92e0c8.pth',
+}
+
+
+class GenEfficientNet(nn.Module):
+ """ Generic EfficientNets
+
+ An implementation of mobile optimized networks that covers:
+ * EfficientNet (B0-B8, L2, CondConv, EdgeTPU)
+ * MixNet (Small, Medium, and Large, XL)
+ * MNASNet A1, B1, and small
+ * FBNet C
+ * Single-Path NAS Pixel1
+ """
+
+ def __init__(self, block_args, num_classes=1000, in_chans=3, num_features=1280, stem_size=32, fix_stem=False,
+ channel_multiplier=1.0, channel_divisor=8, channel_min=None,
+ pad_type='', act_layer=nn.ReLU, drop_rate=0., drop_connect_rate=0.,
+ se_kwargs=None, norm_layer=nn.BatchNorm2d, norm_kwargs=None,
+ weight_init='goog'):
+ super(GenEfficientNet, self).__init__()
+ self.drop_rate = drop_rate
+
+ if not fix_stem:
+ stem_size = round_channels(stem_size, channel_multiplier, channel_divisor, channel_min)
+ self.conv_stem = select_conv2d(in_chans, stem_size, 3, stride=2, padding=pad_type)
+ self.bn1 = norm_layer(stem_size, **norm_kwargs)
+ self.act1 = act_layer(inplace=True)
+ in_chs = stem_size
+
+ builder = EfficientNetBuilder(
+ channel_multiplier, channel_divisor, channel_min,
+ pad_type, act_layer, se_kwargs, norm_layer, norm_kwargs, drop_connect_rate)
+ self.blocks = nn.Sequential(*builder(in_chs, block_args))
+ in_chs = builder.in_chs
+
+ self.conv_head = select_conv2d(in_chs, num_features, 1, padding=pad_type)
+ self.bn2 = norm_layer(num_features, **norm_kwargs)
+ self.act2 = act_layer(inplace=True)
+ self.global_pool = nn.AdaptiveAvgPool2d(1)
+ self.classifier = nn.Linear(num_features, num_classes)
+
+ for n, m in self.named_modules():
+ if weight_init == 'goog':
+ initialize_weight_goog(m, n)
+ else:
+ initialize_weight_default(m, n)
+
+ def features(self, x):
+ x = self.conv_stem(x)
+ x = self.bn1(x)
+ x = self.act1(x)
+ x = self.blocks(x)
+ x = self.conv_head(x)
+ x = self.bn2(x)
+ x = self.act2(x)
+ return x
+
+ def as_sequential(self):
+ layers = [self.conv_stem, self.bn1, self.act1]
+ layers.extend(self.blocks)
+ layers.extend([
+ self.conv_head, self.bn2, self.act2,
+ self.global_pool, nn.Flatten(), nn.Dropout(self.drop_rate), self.classifier])
+ return nn.Sequential(*layers)
+
+ def forward(self, x):
+ x = self.features(x)
+ x = self.global_pool(x)
+ x = x.flatten(1)
+ if self.drop_rate > 0.:
+ x = F.dropout(x, p=self.drop_rate, training=self.training)
+ return self.classifier(x)
+
+
+def _create_model(model_kwargs, variant, pretrained=False):
+ as_sequential = model_kwargs.pop('as_sequential', False)
+ model = GenEfficientNet(**model_kwargs)
+ if pretrained:
+ load_pretrained(model, model_urls[variant])
+ if as_sequential:
+ model = model.as_sequential()
+ return model
+
+
+def _gen_mnasnet_a1(variant, channel_multiplier=1.0, pretrained=False, **kwargs):
+ """Creates a mnasnet-a1 model.
+
+ Ref impl: https://github.com/tensorflow/tpu/tree/master/models/official/mnasnet
+ Paper: https://arxiv.org/pdf/1807.11626.pdf.
+
+ Args:
+ channel_multiplier: multiplier to number of channels per layer.
+ """
+ arch_def = [
+ # stage 0, 112x112 in
+ ['ds_r1_k3_s1_e1_c16_noskip'],
+ # stage 1, 112x112 in
+ ['ir_r2_k3_s2_e6_c24'],
+ # stage 2, 56x56 in
+ ['ir_r3_k5_s2_e3_c40_se0.25'],
+ # stage 3, 28x28 in
+ ['ir_r4_k3_s2_e6_c80'],
+ # stage 4, 14x14in
+ ['ir_r2_k3_s1_e6_c112_se0.25'],
+ # stage 5, 14x14in
+ ['ir_r3_k5_s2_e6_c160_se0.25'],
+ # stage 6, 7x7 in
+ ['ir_r1_k3_s1_e6_c320'],
+ ]
+ with layer_config_kwargs(kwargs):
+ model_kwargs = dict(
+ block_args=decode_arch_def(arch_def),
+ stem_size=32,
+ channel_multiplier=channel_multiplier,
+ act_layer=resolve_act_layer(kwargs, 'relu'),
+ norm_kwargs=resolve_bn_args(kwargs),
+ **kwargs
+ )
+ model = _create_model(model_kwargs, variant, pretrained)
+ return model
+
+
+def _gen_mnasnet_b1(variant, channel_multiplier=1.0, pretrained=False, **kwargs):
+ """Creates a mnasnet-b1 model.
+
+ Ref impl: https://github.com/tensorflow/tpu/tree/master/models/official/mnasnet
+ Paper: https://arxiv.org/pdf/1807.11626.pdf.
+
+ Args:
+ channel_multiplier: multiplier to number of channels per layer.
+ """
+ arch_def = [
+ # stage 0, 112x112 in
+ ['ds_r1_k3_s1_c16_noskip'],
+ # stage 1, 112x112 in
+ ['ir_r3_k3_s2_e3_c24'],
+ # stage 2, 56x56 in
+ ['ir_r3_k5_s2_e3_c40'],
+ # stage 3, 28x28 in
+ ['ir_r3_k5_s2_e6_c80'],
+ # stage 4, 14x14in
+ ['ir_r2_k3_s1_e6_c96'],
+ # stage 5, 14x14in
+ ['ir_r4_k5_s2_e6_c192'],
+ # stage 6, 7x7 in
+ ['ir_r1_k3_s1_e6_c320_noskip']
+ ]
+ with layer_config_kwargs(kwargs):
+ model_kwargs = dict(
+ block_args=decode_arch_def(arch_def),
+ stem_size=32,
+ channel_multiplier=channel_multiplier,
+ act_layer=resolve_act_layer(kwargs, 'relu'),
+ norm_kwargs=resolve_bn_args(kwargs),
+ **kwargs
+ )
+ model = _create_model(model_kwargs, variant, pretrained)
+ return model
+
+
+def _gen_mnasnet_small(variant, channel_multiplier=1.0, pretrained=False, **kwargs):
+ """Creates a mnasnet-b1 model.
+
+ Ref impl: https://github.com/tensorflow/tpu/tree/master/models/official/mnasnet
+ Paper: https://arxiv.org/pdf/1807.11626.pdf.
+
+ Args:
+ channel_multiplier: multiplier to number of channels per layer.
+ """
+ arch_def = [
+ ['ds_r1_k3_s1_c8'],
+ ['ir_r1_k3_s2_e3_c16'],
+ ['ir_r2_k3_s2_e6_c16'],
+ ['ir_r4_k5_s2_e6_c32_se0.25'],
+ ['ir_r3_k3_s1_e6_c32_se0.25'],
+ ['ir_r3_k5_s2_e6_c88_se0.25'],
+ ['ir_r1_k3_s1_e6_c144']
+ ]
+ with layer_config_kwargs(kwargs):
+ model_kwargs = dict(
+ block_args=decode_arch_def(arch_def),
+ stem_size=8,
+ channel_multiplier=channel_multiplier,
+ act_layer=resolve_act_layer(kwargs, 'relu'),
+ norm_kwargs=resolve_bn_args(kwargs),
+ **kwargs
+ )
+ model = _create_model(model_kwargs, variant, pretrained)
+ return model
+
+
+def _gen_mobilenet_v2(
+ variant, channel_multiplier=1.0, depth_multiplier=1.0, fix_stem_head=False, pretrained=False, **kwargs):
+ """ Generate MobileNet-V2 network
+ Ref impl: https://github.com/tensorflow/models/blob/master/research/slim/nets/mobilenet/mobilenet_v2.py
+ Paper: https://arxiv.org/abs/1801.04381
+ """
+ arch_def = [
+ ['ds_r1_k3_s1_c16'],
+ ['ir_r2_k3_s2_e6_c24'],
+ ['ir_r3_k3_s2_e6_c32'],
+ ['ir_r4_k3_s2_e6_c64'],
+ ['ir_r3_k3_s1_e6_c96'],
+ ['ir_r3_k3_s2_e6_c160'],
+ ['ir_r1_k3_s1_e6_c320'],
+ ]
+ with layer_config_kwargs(kwargs):
+ model_kwargs = dict(
+ block_args=decode_arch_def(arch_def, depth_multiplier=depth_multiplier, fix_first_last=fix_stem_head),
+ num_features=1280 if fix_stem_head else round_channels(1280, channel_multiplier, 8, None),
+ stem_size=32,
+ fix_stem=fix_stem_head,
+ channel_multiplier=channel_multiplier,
+ norm_kwargs=resolve_bn_args(kwargs),
+ act_layer=nn.ReLU6,
+ **kwargs
+ )
+ model = _create_model(model_kwargs, variant, pretrained)
+ return model
+
+
+def _gen_fbnetc(variant, channel_multiplier=1.0, pretrained=False, **kwargs):
+ """ FBNet-C
+
+ Paper: https://arxiv.org/abs/1812.03443
+ Ref Impl: https://github.com/facebookresearch/maskrcnn-benchmark/blob/master/maskrcnn_benchmark/modeling/backbone/fbnet_modeldef.py
+
+ NOTE: the impl above does not relate to the 'C' variant here, that was derived from paper,
+ it was used to confirm some building block details
+ """
+ arch_def = [
+ ['ir_r1_k3_s1_e1_c16'],
+ ['ir_r1_k3_s2_e6_c24', 'ir_r2_k3_s1_e1_c24'],
+ ['ir_r1_k5_s2_e6_c32', 'ir_r1_k5_s1_e3_c32', 'ir_r1_k5_s1_e6_c32', 'ir_r1_k3_s1_e6_c32'],
+ ['ir_r1_k5_s2_e6_c64', 'ir_r1_k5_s1_e3_c64', 'ir_r2_k5_s1_e6_c64'],
+ ['ir_r3_k5_s1_e6_c112', 'ir_r1_k5_s1_e3_c112'],
+ ['ir_r4_k5_s2_e6_c184'],
+ ['ir_r1_k3_s1_e6_c352'],
+ ]
+ with layer_config_kwargs(kwargs):
+ model_kwargs = dict(
+ block_args=decode_arch_def(arch_def),
+ stem_size=16,
+ num_features=1984, # paper suggests this, but is not 100% clear
+ channel_multiplier=channel_multiplier,
+ act_layer=resolve_act_layer(kwargs, 'relu'),
+ norm_kwargs=resolve_bn_args(kwargs),
+ **kwargs
+ )
+ model = _create_model(model_kwargs, variant, pretrained)
+ return model
+
+
+def _gen_spnasnet(variant, channel_multiplier=1.0, pretrained=False, **kwargs):
+ """Creates the Single-Path NAS model from search targeted for Pixel1 phone.
+
+ Paper: https://arxiv.org/abs/1904.02877
+
+ Args:
+ channel_multiplier: multiplier to number of channels per layer.
+ """
+ arch_def = [
+ # stage 0, 112x112 in
+ ['ds_r1_k3_s1_c16_noskip'],
+ # stage 1, 112x112 in
+ ['ir_r3_k3_s2_e3_c24'],
+ # stage 2, 56x56 in
+ ['ir_r1_k5_s2_e6_c40', 'ir_r3_k3_s1_e3_c40'],
+ # stage 3, 28x28 in
+ ['ir_r1_k5_s2_e6_c80', 'ir_r3_k3_s1_e3_c80'],
+ # stage 4, 14x14in
+ ['ir_r1_k5_s1_e6_c96', 'ir_r3_k5_s1_e3_c96'],
+ # stage 5, 14x14in
+ ['ir_r4_k5_s2_e6_c192'],
+ # stage 6, 7x7 in
+ ['ir_r1_k3_s1_e6_c320_noskip']
+ ]
+ with layer_config_kwargs(kwargs):
+ model_kwargs = dict(
+ block_args=decode_arch_def(arch_def),
+ stem_size=32,
+ channel_multiplier=channel_multiplier,
+ act_layer=resolve_act_layer(kwargs, 'relu'),
+ norm_kwargs=resolve_bn_args(kwargs),
+ **kwargs
+ )
+ model = _create_model(model_kwargs, variant, pretrained)
+ return model
+
+
+def _gen_efficientnet(variant, channel_multiplier=1.0, depth_multiplier=1.0, pretrained=False, **kwargs):
+ """Creates an EfficientNet model.
+
+ Ref impl: https://github.com/tensorflow/tpu/blob/master/models/official/efficientnet/efficientnet_model.py
+ Paper: https://arxiv.org/abs/1905.11946
+
+ EfficientNet params
+ name: (channel_multiplier, depth_multiplier, resolution, dropout_rate)
+ 'efficientnet-b0': (1.0, 1.0, 224, 0.2),
+ 'efficientnet-b1': (1.0, 1.1, 240, 0.2),
+ 'efficientnet-b2': (1.1, 1.2, 260, 0.3),
+ 'efficientnet-b3': (1.2, 1.4, 300, 0.3),
+ 'efficientnet-b4': (1.4, 1.8, 380, 0.4),
+ 'efficientnet-b5': (1.6, 2.2, 456, 0.4),
+ 'efficientnet-b6': (1.8, 2.6, 528, 0.5),
+ 'efficientnet-b7': (2.0, 3.1, 600, 0.5),
+ 'efficientnet-b8': (2.2, 3.6, 672, 0.5),
+
+ Args:
+ channel_multiplier: multiplier to number of channels per layer
+ depth_multiplier: multiplier to number of repeats per stage
+
+ """
+ arch_def = [
+ ['ds_r1_k3_s1_e1_c16_se0.25'],
+ ['ir_r2_k3_s2_e6_c24_se0.25'],
+ ['ir_r2_k5_s2_e6_c40_se0.25'],
+ ['ir_r3_k3_s2_e6_c80_se0.25'],
+ ['ir_r3_k5_s1_e6_c112_se0.25'],
+ ['ir_r4_k5_s2_e6_c192_se0.25'],
+ ['ir_r1_k3_s1_e6_c320_se0.25'],
+ ]
+ with layer_config_kwargs(kwargs):
+ model_kwargs = dict(
+ block_args=decode_arch_def(arch_def, depth_multiplier),
+ num_features=round_channels(1280, channel_multiplier, 8, None),
+ stem_size=32,
+ channel_multiplier=channel_multiplier,
+ act_layer=resolve_act_layer(kwargs, 'swish'),
+ norm_kwargs=resolve_bn_args(kwargs),
+ **kwargs,
+ )
+ model = _create_model(model_kwargs, variant, pretrained)
+ return model
+
+
+def _gen_efficientnet_edge(variant, channel_multiplier=1.0, depth_multiplier=1.0, pretrained=False, **kwargs):
+ arch_def = [
+ # NOTE `fc` is present to override a mismatch between stem channels and in chs not
+ # present in other models
+ ['er_r1_k3_s1_e4_c24_fc24_noskip'],
+ ['er_r2_k3_s2_e8_c32'],
+ ['er_r4_k3_s2_e8_c48'],
+ ['ir_r5_k5_s2_e8_c96'],
+ ['ir_r4_k5_s1_e8_c144'],
+ ['ir_r2_k5_s2_e8_c192'],
+ ]
+ with layer_config_kwargs(kwargs):
+ model_kwargs = dict(
+ block_args=decode_arch_def(arch_def, depth_multiplier),
+ num_features=round_channels(1280, channel_multiplier, 8, None),
+ stem_size=32,
+ channel_multiplier=channel_multiplier,
+ act_layer=resolve_act_layer(kwargs, 'relu'),
+ norm_kwargs=resolve_bn_args(kwargs),
+ **kwargs,
+ )
+ model = _create_model(model_kwargs, variant, pretrained)
+ return model
+
+
+def _gen_efficientnet_condconv(
+ variant, channel_multiplier=1.0, depth_multiplier=1.0, experts_multiplier=1, pretrained=False, **kwargs):
+ """Creates an efficientnet-condconv model."""
+ arch_def = [
+ ['ds_r1_k3_s1_e1_c16_se0.25'],
+ ['ir_r2_k3_s2_e6_c24_se0.25'],
+ ['ir_r2_k5_s2_e6_c40_se0.25'],
+ ['ir_r3_k3_s2_e6_c80_se0.25'],
+ ['ir_r3_k5_s1_e6_c112_se0.25_cc4'],
+ ['ir_r4_k5_s2_e6_c192_se0.25_cc4'],
+ ['ir_r1_k3_s1_e6_c320_se0.25_cc4'],
+ ]
+ with layer_config_kwargs(kwargs):
+ model_kwargs = dict(
+ block_args=decode_arch_def(arch_def, depth_multiplier, experts_multiplier=experts_multiplier),
+ num_features=round_channels(1280, channel_multiplier, 8, None),
+ stem_size=32,
+ channel_multiplier=channel_multiplier,
+ act_layer=resolve_act_layer(kwargs, 'swish'),
+ norm_kwargs=resolve_bn_args(kwargs),
+ **kwargs,
+ )
+ model = _create_model(model_kwargs, variant, pretrained)
+ return model
+
+
+def _gen_efficientnet_lite(variant, channel_multiplier=1.0, depth_multiplier=1.0, pretrained=False, **kwargs):
+ """Creates an EfficientNet-Lite model.
+
+ Ref impl: https://github.com/tensorflow/tpu/tree/master/models/official/efficientnet/lite
+ Paper: https://arxiv.org/abs/1905.11946
+
+ EfficientNet params
+ name: (channel_multiplier, depth_multiplier, resolution, dropout_rate)
+ 'efficientnet-lite0': (1.0, 1.0, 224, 0.2),
+ 'efficientnet-lite1': (1.0, 1.1, 240, 0.2),
+ 'efficientnet-lite2': (1.1, 1.2, 260, 0.3),
+ 'efficientnet-lite3': (1.2, 1.4, 280, 0.3),
+ 'efficientnet-lite4': (1.4, 1.8, 300, 0.3),
+
+ Args:
+ channel_multiplier: multiplier to number of channels per layer
+ depth_multiplier: multiplier to number of repeats per stage
+ """
+ arch_def = [
+ ['ds_r1_k3_s1_e1_c16'],
+ ['ir_r2_k3_s2_e6_c24'],
+ ['ir_r2_k5_s2_e6_c40'],
+ ['ir_r3_k3_s2_e6_c80'],
+ ['ir_r3_k5_s1_e6_c112'],
+ ['ir_r4_k5_s2_e6_c192'],
+ ['ir_r1_k3_s1_e6_c320'],
+ ]
+ with layer_config_kwargs(kwargs):
+ model_kwargs = dict(
+ block_args=decode_arch_def(arch_def, depth_multiplier, fix_first_last=True),
+ num_features=1280,
+ stem_size=32,
+ fix_stem=True,
+ channel_multiplier=channel_multiplier,
+ act_layer=nn.ReLU6,
+ norm_kwargs=resolve_bn_args(kwargs),
+ **kwargs,
+ )
+ model = _create_model(model_kwargs, variant, pretrained)
+ return model
+
+
+def _gen_mixnet_s(variant, channel_multiplier=1.0, pretrained=False, **kwargs):
+ """Creates a MixNet Small model.
+
+ Ref impl: https://github.com/tensorflow/tpu/tree/master/models/official/mnasnet/mixnet
+ Paper: https://arxiv.org/abs/1907.09595
+ """
+ arch_def = [
+ # stage 0, 112x112 in
+ ['ds_r1_k3_s1_e1_c16'], # relu
+ # stage 1, 112x112 in
+ ['ir_r1_k3_a1.1_p1.1_s2_e6_c24', 'ir_r1_k3_a1.1_p1.1_s1_e3_c24'], # relu
+ # stage 2, 56x56 in
+ ['ir_r1_k3.5.7_s2_e6_c40_se0.5_nsw', 'ir_r3_k3.5_a1.1_p1.1_s1_e6_c40_se0.5_nsw'], # swish
+ # stage 3, 28x28 in
+ ['ir_r1_k3.5.7_p1.1_s2_e6_c80_se0.25_nsw', 'ir_r2_k3.5_p1.1_s1_e6_c80_se0.25_nsw'], # swish
+ # stage 4, 14x14in
+ ['ir_r1_k3.5.7_a1.1_p1.1_s1_e6_c120_se0.5_nsw', 'ir_r2_k3.5.7.9_a1.1_p1.1_s1_e3_c120_se0.5_nsw'], # swish
+ # stage 5, 14x14in
+ ['ir_r1_k3.5.7.9.11_s2_e6_c200_se0.5_nsw', 'ir_r2_k3.5.7.9_p1.1_s1_e6_c200_se0.5_nsw'], # swish
+ # 7x7
+ ]
+ with layer_config_kwargs(kwargs):
+ model_kwargs = dict(
+ block_args=decode_arch_def(arch_def),
+ num_features=1536,
+ stem_size=16,
+ channel_multiplier=channel_multiplier,
+ act_layer=resolve_act_layer(kwargs, 'relu'),
+ norm_kwargs=resolve_bn_args(kwargs),
+ **kwargs
+ )
+ model = _create_model(model_kwargs, variant, pretrained)
+ return model
+
+
+def _gen_mixnet_m(variant, channel_multiplier=1.0, depth_multiplier=1.0, pretrained=False, **kwargs):
+ """Creates a MixNet Medium-Large model.
+
+ Ref impl: https://github.com/tensorflow/tpu/tree/master/models/official/mnasnet/mixnet
+ Paper: https://arxiv.org/abs/1907.09595
+ """
+ arch_def = [
+ # stage 0, 112x112 in
+ ['ds_r1_k3_s1_e1_c24'], # relu
+ # stage 1, 112x112 in
+ ['ir_r1_k3.5.7_a1.1_p1.1_s2_e6_c32', 'ir_r1_k3_a1.1_p1.1_s1_e3_c32'], # relu
+ # stage 2, 56x56 in
+ ['ir_r1_k3.5.7.9_s2_e6_c40_se0.5_nsw', 'ir_r3_k3.5_a1.1_p1.1_s1_e6_c40_se0.5_nsw'], # swish
+ # stage 3, 28x28 in
+ ['ir_r1_k3.5.7_s2_e6_c80_se0.25_nsw', 'ir_r3_k3.5.7.9_a1.1_p1.1_s1_e6_c80_se0.25_nsw'], # swish
+ # stage 4, 14x14in
+ ['ir_r1_k3_s1_e6_c120_se0.5_nsw', 'ir_r3_k3.5.7.9_a1.1_p1.1_s1_e3_c120_se0.5_nsw'], # swish
+ # stage 5, 14x14in
+ ['ir_r1_k3.5.7.9_s2_e6_c200_se0.5_nsw', 'ir_r3_k3.5.7.9_p1.1_s1_e6_c200_se0.5_nsw'], # swish
+ # 7x7
+ ]
+ with layer_config_kwargs(kwargs):
+ model_kwargs = dict(
+ block_args=decode_arch_def(arch_def, depth_multiplier, depth_trunc='round'),
+ num_features=1536,
+ stem_size=24,
+ channel_multiplier=channel_multiplier,
+ act_layer=resolve_act_layer(kwargs, 'relu'),
+ norm_kwargs=resolve_bn_args(kwargs),
+ **kwargs
+ )
+ model = _create_model(model_kwargs, variant, pretrained)
+ return model
+
+
+def mnasnet_050(pretrained=False, **kwargs):
+ """ MNASNet B1, depth multiplier of 0.5. """
+ model = _gen_mnasnet_b1('mnasnet_050', 0.5, pretrained=pretrained, **kwargs)
+ return model
+
+
+def mnasnet_075(pretrained=False, **kwargs):
+ """ MNASNet B1, depth multiplier of 0.75. """
+ model = _gen_mnasnet_b1('mnasnet_075', 0.75, pretrained=pretrained, **kwargs)
+ return model
+
+
+def mnasnet_100(pretrained=False, **kwargs):
+ """ MNASNet B1, depth multiplier of 1.0. """
+ model = _gen_mnasnet_b1('mnasnet_100', 1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def mnasnet_b1(pretrained=False, **kwargs):
+ """ MNASNet B1, depth multiplier of 1.0. """
+ return mnasnet_100(pretrained, **kwargs)
+
+
+def mnasnet_140(pretrained=False, **kwargs):
+ """ MNASNet B1, depth multiplier of 1.4 """
+ model = _gen_mnasnet_b1('mnasnet_140', 1.4, pretrained=pretrained, **kwargs)
+ return model
+
+
+def semnasnet_050(pretrained=False, **kwargs):
+ """ MNASNet A1 (w/ SE), depth multiplier of 0.5 """
+ model = _gen_mnasnet_a1('semnasnet_050', 0.5, pretrained=pretrained, **kwargs)
+ return model
+
+
+def semnasnet_075(pretrained=False, **kwargs):
+ """ MNASNet A1 (w/ SE), depth multiplier of 0.75. """
+ model = _gen_mnasnet_a1('semnasnet_075', 0.75, pretrained=pretrained, **kwargs)
+ return model
+
+
+def semnasnet_100(pretrained=False, **kwargs):
+ """ MNASNet A1 (w/ SE), depth multiplier of 1.0. """
+ model = _gen_mnasnet_a1('semnasnet_100', 1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def mnasnet_a1(pretrained=False, **kwargs):
+ """ MNASNet A1 (w/ SE), depth multiplier of 1.0. """
+ return semnasnet_100(pretrained, **kwargs)
+
+
+def semnasnet_140(pretrained=False, **kwargs):
+ """ MNASNet A1 (w/ SE), depth multiplier of 1.4. """
+ model = _gen_mnasnet_a1('semnasnet_140', 1.4, pretrained=pretrained, **kwargs)
+ return model
+
+
+def mnasnet_small(pretrained=False, **kwargs):
+ """ MNASNet Small, depth multiplier of 1.0. """
+ model = _gen_mnasnet_small('mnasnet_small', 1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def mobilenetv2_100(pretrained=False, **kwargs):
+ """ MobileNet V2 w/ 1.0 channel multiplier """
+ model = _gen_mobilenet_v2('mobilenetv2_100', 1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def mobilenetv2_140(pretrained=False, **kwargs):
+ """ MobileNet V2 w/ 1.4 channel multiplier """
+ model = _gen_mobilenet_v2('mobilenetv2_140', 1.4, pretrained=pretrained, **kwargs)
+ return model
+
+
+def mobilenetv2_110d(pretrained=False, **kwargs):
+ """ MobileNet V2 w/ 1.1 channel, 1.2 depth multipliers"""
+ model = _gen_mobilenet_v2(
+ 'mobilenetv2_110d', 1.1, depth_multiplier=1.2, fix_stem_head=True, pretrained=pretrained, **kwargs)
+ return model
+
+
+def mobilenetv2_120d(pretrained=False, **kwargs):
+ """ MobileNet V2 w/ 1.2 channel, 1.4 depth multipliers """
+ model = _gen_mobilenet_v2(
+ 'mobilenetv2_120d', 1.2, depth_multiplier=1.4, fix_stem_head=True, pretrained=pretrained, **kwargs)
+ return model
+
+
+def fbnetc_100(pretrained=False, **kwargs):
+ """ FBNet-C """
+ if pretrained:
+ # pretrained model trained with non-default BN epsilon
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ model = _gen_fbnetc('fbnetc_100', 1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def spnasnet_100(pretrained=False, **kwargs):
+ """ Single-Path NAS Pixel1"""
+ model = _gen_spnasnet('spnasnet_100', 1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def efficientnet_b0(pretrained=False, **kwargs):
+ """ EfficientNet-B0 """
+ # NOTE for train set drop_rate=0.2, drop_connect_rate=0.2
+ model = _gen_efficientnet(
+ 'efficientnet_b0', channel_multiplier=1.0, depth_multiplier=1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def efficientnet_b1(pretrained=False, **kwargs):
+ """ EfficientNet-B1 """
+ # NOTE for train set drop_rate=0.2, drop_connect_rate=0.2
+ model = _gen_efficientnet(
+ 'efficientnet_b1', channel_multiplier=1.0, depth_multiplier=1.1, pretrained=pretrained, **kwargs)
+ return model
+
+
+def efficientnet_b2(pretrained=False, **kwargs):
+ """ EfficientNet-B2 """
+ # NOTE for train set drop_rate=0.3, drop_connect_rate=0.2
+ model = _gen_efficientnet(
+ 'efficientnet_b2', channel_multiplier=1.1, depth_multiplier=1.2, pretrained=pretrained, **kwargs)
+ return model
+
+
+def efficientnet_b3(pretrained=False, **kwargs):
+ """ EfficientNet-B3 """
+ # NOTE for train set drop_rate=0.3, drop_connect_rate=0.2
+ model = _gen_efficientnet(
+ 'efficientnet_b3', channel_multiplier=1.2, depth_multiplier=1.4, pretrained=pretrained, **kwargs)
+ return model
+
+
+def efficientnet_b4(pretrained=False, **kwargs):
+ """ EfficientNet-B4 """
+ # NOTE for train set drop_rate=0.4, drop_connect_rate=0.2
+ model = _gen_efficientnet(
+ 'efficientnet_b4', channel_multiplier=1.4, depth_multiplier=1.8, pretrained=pretrained, **kwargs)
+ return model
+
+
+def efficientnet_b5(pretrained=False, **kwargs):
+ """ EfficientNet-B5 """
+ # NOTE for train set drop_rate=0.4, drop_connect_rate=0.2
+ model = _gen_efficientnet(
+ 'efficientnet_b5', channel_multiplier=1.6, depth_multiplier=2.2, pretrained=pretrained, **kwargs)
+ return model
+
+
+def efficientnet_b6(pretrained=False, **kwargs):
+ """ EfficientNet-B6 """
+ # NOTE for train set drop_rate=0.5, drop_connect_rate=0.2
+ model = _gen_efficientnet(
+ 'efficientnet_b6', channel_multiplier=1.8, depth_multiplier=2.6, pretrained=pretrained, **kwargs)
+ return model
+
+
+def efficientnet_b7(pretrained=False, **kwargs):
+ """ EfficientNet-B7 """
+ # NOTE for train set drop_rate=0.5, drop_connect_rate=0.2
+ model = _gen_efficientnet(
+ 'efficientnet_b7', channel_multiplier=2.0, depth_multiplier=3.1, pretrained=pretrained, **kwargs)
+ return model
+
+
+def efficientnet_b8(pretrained=False, **kwargs):
+ """ EfficientNet-B8 """
+ # NOTE for train set drop_rate=0.5, drop_connect_rate=0.2
+ model = _gen_efficientnet(
+ 'efficientnet_b8', channel_multiplier=2.2, depth_multiplier=3.6, pretrained=pretrained, **kwargs)
+ return model
+
+
+def efficientnet_l2(pretrained=False, **kwargs):
+ """ EfficientNet-L2. """
+ # NOTE for train, drop_rate should be 0.5
+ model = _gen_efficientnet(
+ 'efficientnet_l2', channel_multiplier=4.3, depth_multiplier=5.3, pretrained=pretrained, **kwargs)
+ return model
+
+
+def efficientnet_es(pretrained=False, **kwargs):
+ """ EfficientNet-Edge Small. """
+ model = _gen_efficientnet_edge(
+ 'efficientnet_es', channel_multiplier=1.0, depth_multiplier=1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def efficientnet_em(pretrained=False, **kwargs):
+ """ EfficientNet-Edge-Medium. """
+ model = _gen_efficientnet_edge(
+ 'efficientnet_em', channel_multiplier=1.0, depth_multiplier=1.1, pretrained=pretrained, **kwargs)
+ return model
+
+
+def efficientnet_el(pretrained=False, **kwargs):
+ """ EfficientNet-Edge-Large. """
+ model = _gen_efficientnet_edge(
+ 'efficientnet_el', channel_multiplier=1.2, depth_multiplier=1.4, pretrained=pretrained, **kwargs)
+ return model
+
+
+def efficientnet_cc_b0_4e(pretrained=False, **kwargs):
+ """ EfficientNet-CondConv-B0 w/ 8 Experts """
+ # NOTE for train set drop_rate=0.25, drop_connect_rate=0.2
+ model = _gen_efficientnet_condconv(
+ 'efficientnet_cc_b0_4e', channel_multiplier=1.0, depth_multiplier=1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def efficientnet_cc_b0_8e(pretrained=False, **kwargs):
+ """ EfficientNet-CondConv-B0 w/ 8 Experts """
+ # NOTE for train set drop_rate=0.25, drop_connect_rate=0.2
+ model = _gen_efficientnet_condconv(
+ 'efficientnet_cc_b0_8e', channel_multiplier=1.0, depth_multiplier=1.0, experts_multiplier=2,
+ pretrained=pretrained, **kwargs)
+ return model
+
+
+def efficientnet_cc_b1_8e(pretrained=False, **kwargs):
+ """ EfficientNet-CondConv-B1 w/ 8 Experts """
+ # NOTE for train set drop_rate=0.25, drop_connect_rate=0.2
+ model = _gen_efficientnet_condconv(
+ 'efficientnet_cc_b1_8e', channel_multiplier=1.0, depth_multiplier=1.1, experts_multiplier=2,
+ pretrained=pretrained, **kwargs)
+ return model
+
+
+def efficientnet_lite0(pretrained=False, **kwargs):
+ """ EfficientNet-Lite0 """
+ model = _gen_efficientnet_lite(
+ 'efficientnet_lite0', channel_multiplier=1.0, depth_multiplier=1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def efficientnet_lite1(pretrained=False, **kwargs):
+ """ EfficientNet-Lite1 """
+ model = _gen_efficientnet_lite(
+ 'efficientnet_lite1', channel_multiplier=1.0, depth_multiplier=1.1, pretrained=pretrained, **kwargs)
+ return model
+
+
+def efficientnet_lite2(pretrained=False, **kwargs):
+ """ EfficientNet-Lite2 """
+ model = _gen_efficientnet_lite(
+ 'efficientnet_lite2', channel_multiplier=1.1, depth_multiplier=1.2, pretrained=pretrained, **kwargs)
+ return model
+
+
+def efficientnet_lite3(pretrained=False, **kwargs):
+ """ EfficientNet-Lite3 """
+ model = _gen_efficientnet_lite(
+ 'efficientnet_lite3', channel_multiplier=1.2, depth_multiplier=1.4, pretrained=pretrained, **kwargs)
+ return model
+
+
+def efficientnet_lite4(pretrained=False, **kwargs):
+ """ EfficientNet-Lite4 """
+ model = _gen_efficientnet_lite(
+ 'efficientnet_lite4', channel_multiplier=1.4, depth_multiplier=1.8, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_b0(pretrained=False, **kwargs):
+ """ EfficientNet-B0 AutoAug. Tensorflow compatible variant """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_b0', channel_multiplier=1.0, depth_multiplier=1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_b1(pretrained=False, **kwargs):
+ """ EfficientNet-B1 AutoAug. Tensorflow compatible variant """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_b1', channel_multiplier=1.0, depth_multiplier=1.1, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_b2(pretrained=False, **kwargs):
+ """ EfficientNet-B2 AutoAug. Tensorflow compatible variant """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_b2', channel_multiplier=1.1, depth_multiplier=1.2, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_b3(pretrained=False, **kwargs):
+ """ EfficientNet-B3 AutoAug. Tensorflow compatible variant """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_b3', channel_multiplier=1.2, depth_multiplier=1.4, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_b4(pretrained=False, **kwargs):
+ """ EfficientNet-B4 AutoAug. Tensorflow compatible variant """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_b4', channel_multiplier=1.4, depth_multiplier=1.8, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_b5(pretrained=False, **kwargs):
+ """ EfficientNet-B5 RandAug. Tensorflow compatible variant """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_b5', channel_multiplier=1.6, depth_multiplier=2.2, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_b6(pretrained=False, **kwargs):
+ """ EfficientNet-B6 AutoAug. Tensorflow compatible variant """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_b6', channel_multiplier=1.8, depth_multiplier=2.6, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_b7(pretrained=False, **kwargs):
+ """ EfficientNet-B7 RandAug. Tensorflow compatible variant """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_b7', channel_multiplier=2.0, depth_multiplier=3.1, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_b8(pretrained=False, **kwargs):
+ """ EfficientNet-B8 RandAug. Tensorflow compatible variant """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_b8', channel_multiplier=2.2, depth_multiplier=3.6, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_b0_ap(pretrained=False, **kwargs):
+ """ EfficientNet-B0 AdvProp. Tensorflow compatible variant
+ Paper: Adversarial Examples Improve Image Recognition (https://arxiv.org/abs/1911.09665)
+ """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_b0_ap', channel_multiplier=1.0, depth_multiplier=1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_b1_ap(pretrained=False, **kwargs):
+ """ EfficientNet-B1 AdvProp. Tensorflow compatible variant
+ Paper: Adversarial Examples Improve Image Recognition (https://arxiv.org/abs/1911.09665)
+ """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_b1_ap', channel_multiplier=1.0, depth_multiplier=1.1, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_b2_ap(pretrained=False, **kwargs):
+ """ EfficientNet-B2 AdvProp. Tensorflow compatible variant
+ Paper: Adversarial Examples Improve Image Recognition (https://arxiv.org/abs/1911.09665)
+ """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_b2_ap', channel_multiplier=1.1, depth_multiplier=1.2, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_b3_ap(pretrained=False, **kwargs):
+ """ EfficientNet-B3 AdvProp. Tensorflow compatible variant
+ Paper: Adversarial Examples Improve Image Recognition (https://arxiv.org/abs/1911.09665)
+ """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_b3_ap', channel_multiplier=1.2, depth_multiplier=1.4, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_b4_ap(pretrained=False, **kwargs):
+ """ EfficientNet-B4 AdvProp. Tensorflow compatible variant
+ Paper: Adversarial Examples Improve Image Recognition (https://arxiv.org/abs/1911.09665)
+ """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_b4_ap', channel_multiplier=1.4, depth_multiplier=1.8, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_b5_ap(pretrained=False, **kwargs):
+ """ EfficientNet-B5 AdvProp. Tensorflow compatible variant
+ Paper: Adversarial Examples Improve Image Recognition (https://arxiv.org/abs/1911.09665)
+ """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_b5_ap', channel_multiplier=1.6, depth_multiplier=2.2, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_b6_ap(pretrained=False, **kwargs):
+ """ EfficientNet-B6 AdvProp. Tensorflow compatible variant
+ Paper: Adversarial Examples Improve Image Recognition (https://arxiv.org/abs/1911.09665)
+ """
+ # NOTE for train, drop_rate should be 0.5
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_b6_ap', channel_multiplier=1.8, depth_multiplier=2.6, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_b7_ap(pretrained=False, **kwargs):
+ """ EfficientNet-B7 AdvProp. Tensorflow compatible variant
+ Paper: Adversarial Examples Improve Image Recognition (https://arxiv.org/abs/1911.09665)
+ """
+ # NOTE for train, drop_rate should be 0.5
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_b7_ap', channel_multiplier=2.0, depth_multiplier=3.1, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_b8_ap(pretrained=False, **kwargs):
+ """ EfficientNet-B8 AdvProp. Tensorflow compatible variant
+ Paper: Adversarial Examples Improve Image Recognition (https://arxiv.org/abs/1911.09665)
+ """
+ # NOTE for train, drop_rate should be 0.5
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_b8_ap', channel_multiplier=2.2, depth_multiplier=3.6, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_b0_ns(pretrained=False, **kwargs):
+ """ EfficientNet-B0 NoisyStudent. Tensorflow compatible variant
+ Paper: Self-training with Noisy Student improves ImageNet classification (https://arxiv.org/abs/1911.04252)
+ """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_b0_ns', channel_multiplier=1.0, depth_multiplier=1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_b1_ns(pretrained=False, **kwargs):
+ """ EfficientNet-B1 NoisyStudent. Tensorflow compatible variant
+ Paper: Self-training with Noisy Student improves ImageNet classification (https://arxiv.org/abs/1911.04252)
+ """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_b1_ns', channel_multiplier=1.0, depth_multiplier=1.1, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_b2_ns(pretrained=False, **kwargs):
+ """ EfficientNet-B2 NoisyStudent. Tensorflow compatible variant
+ Paper: Self-training with Noisy Student improves ImageNet classification (https://arxiv.org/abs/1911.04252)
+ """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_b2_ns', channel_multiplier=1.1, depth_multiplier=1.2, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_b3_ns(pretrained=False, **kwargs):
+ """ EfficientNet-B3 NoisyStudent. Tensorflow compatible variant
+ Paper: Self-training with Noisy Student improves ImageNet classification (https://arxiv.org/abs/1911.04252)
+ """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_b3_ns', channel_multiplier=1.2, depth_multiplier=1.4, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_b4_ns(pretrained=False, **kwargs):
+ """ EfficientNet-B4 NoisyStudent. Tensorflow compatible variant
+ Paper: Self-training with Noisy Student improves ImageNet classification (https://arxiv.org/abs/1911.04252)
+ """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_b4_ns', channel_multiplier=1.4, depth_multiplier=1.8, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_b5_ns(pretrained=False, **kwargs):
+ """ EfficientNet-B5 NoisyStudent. Tensorflow compatible variant
+ Paper: Self-training with Noisy Student improves ImageNet classification (https://arxiv.org/abs/1911.04252)
+ """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_b5_ns', channel_multiplier=1.6, depth_multiplier=2.2, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_b6_ns(pretrained=False, **kwargs):
+ """ EfficientNet-B6 NoisyStudent. Tensorflow compatible variant
+ Paper: Self-training with Noisy Student improves ImageNet classification (https://arxiv.org/abs/1911.04252)
+ """
+ # NOTE for train, drop_rate should be 0.5
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_b6_ns', channel_multiplier=1.8, depth_multiplier=2.6, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_b7_ns(pretrained=False, **kwargs):
+ """ EfficientNet-B7 NoisyStudent. Tensorflow compatible variant
+ Paper: Self-training with Noisy Student improves ImageNet classification (https://arxiv.org/abs/1911.04252)
+ """
+ # NOTE for train, drop_rate should be 0.5
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_b7_ns', channel_multiplier=2.0, depth_multiplier=3.1, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_l2_ns_475(pretrained=False, **kwargs):
+ """ EfficientNet-L2 NoisyStudent @ 475x475. Tensorflow compatible variant
+ Paper: Self-training with Noisy Student improves ImageNet classification (https://arxiv.org/abs/1911.04252)
+ """
+ # NOTE for train, drop_rate should be 0.5
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_l2_ns_475', channel_multiplier=4.3, depth_multiplier=5.3, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_l2_ns(pretrained=False, **kwargs):
+ """ EfficientNet-L2 NoisyStudent. Tensorflow compatible variant
+ Paper: Self-training with Noisy Student improves ImageNet classification (https://arxiv.org/abs/1911.04252)
+ """
+ # NOTE for train, drop_rate should be 0.5
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet(
+ 'tf_efficientnet_l2_ns', channel_multiplier=4.3, depth_multiplier=5.3, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_es(pretrained=False, **kwargs):
+ """ EfficientNet-Edge Small. Tensorflow compatible variant """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet_edge(
+ 'tf_efficientnet_es', channel_multiplier=1.0, depth_multiplier=1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_em(pretrained=False, **kwargs):
+ """ EfficientNet-Edge-Medium. Tensorflow compatible variant """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet_edge(
+ 'tf_efficientnet_em', channel_multiplier=1.0, depth_multiplier=1.1, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_el(pretrained=False, **kwargs):
+ """ EfficientNet-Edge-Large. Tensorflow compatible variant """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet_edge(
+ 'tf_efficientnet_el', channel_multiplier=1.2, depth_multiplier=1.4, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_cc_b0_4e(pretrained=False, **kwargs):
+ """ EfficientNet-CondConv-B0 w/ 4 Experts """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet_condconv(
+ 'tf_efficientnet_cc_b0_4e', channel_multiplier=1.0, depth_multiplier=1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_cc_b0_8e(pretrained=False, **kwargs):
+ """ EfficientNet-CondConv-B0 w/ 8 Experts """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet_condconv(
+ 'tf_efficientnet_cc_b0_8e', channel_multiplier=1.0, depth_multiplier=1.0, experts_multiplier=2,
+ pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_cc_b1_8e(pretrained=False, **kwargs):
+ """ EfficientNet-CondConv-B1 w/ 8 Experts """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet_condconv(
+ 'tf_efficientnet_cc_b1_8e', channel_multiplier=1.0, depth_multiplier=1.1, experts_multiplier=2,
+ pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_lite0(pretrained=False, **kwargs):
+ """ EfficientNet-Lite0. Tensorflow compatible variant """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet_lite(
+ 'tf_efficientnet_lite0', channel_multiplier=1.0, depth_multiplier=1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_lite1(pretrained=False, **kwargs):
+ """ EfficientNet-Lite1. Tensorflow compatible variant """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet_lite(
+ 'tf_efficientnet_lite1', channel_multiplier=1.0, depth_multiplier=1.1, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_lite2(pretrained=False, **kwargs):
+ """ EfficientNet-Lite2. Tensorflow compatible variant """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet_lite(
+ 'tf_efficientnet_lite2', channel_multiplier=1.1, depth_multiplier=1.2, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_lite3(pretrained=False, **kwargs):
+ """ EfficientNet-Lite3. Tensorflow compatible variant """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet_lite(
+ 'tf_efficientnet_lite3', channel_multiplier=1.2, depth_multiplier=1.4, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_efficientnet_lite4(pretrained=False, **kwargs):
+ """ EfficientNet-Lite4. Tensorflow compatible variant """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_efficientnet_lite(
+ 'tf_efficientnet_lite4', channel_multiplier=1.4, depth_multiplier=1.8, pretrained=pretrained, **kwargs)
+ return model
+
+
+def mixnet_s(pretrained=False, **kwargs):
+ """Creates a MixNet Small model.
+ """
+ # NOTE for train set drop_rate=0.2
+ model = _gen_mixnet_s(
+ 'mixnet_s', channel_multiplier=1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def mixnet_m(pretrained=False, **kwargs):
+ """Creates a MixNet Medium model.
+ """
+ # NOTE for train set drop_rate=0.25
+ model = _gen_mixnet_m(
+ 'mixnet_m', channel_multiplier=1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def mixnet_l(pretrained=False, **kwargs):
+ """Creates a MixNet Large model.
+ """
+ # NOTE for train set drop_rate=0.25
+ model = _gen_mixnet_m(
+ 'mixnet_l', channel_multiplier=1.3, pretrained=pretrained, **kwargs)
+ return model
+
+
+def mixnet_xl(pretrained=False, **kwargs):
+ """Creates a MixNet Extra-Large model.
+ Not a paper spec, experimental def by RW w/ depth scaling.
+ """
+ # NOTE for train set drop_rate=0.25, drop_connect_rate=0.2
+ model = _gen_mixnet_m(
+ 'mixnet_xl', channel_multiplier=1.6, depth_multiplier=1.2, pretrained=pretrained, **kwargs)
+ return model
+
+
+def mixnet_xxl(pretrained=False, **kwargs):
+ """Creates a MixNet Double Extra Large model.
+ Not a paper spec, experimental def by RW w/ depth scaling.
+ """
+ # NOTE for train set drop_rate=0.3, drop_connect_rate=0.2
+ model = _gen_mixnet_m(
+ 'mixnet_xxl', channel_multiplier=2.4, depth_multiplier=1.3, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_mixnet_s(pretrained=False, **kwargs):
+ """Creates a MixNet Small model. Tensorflow compatible variant
+ """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_mixnet_s(
+ 'tf_mixnet_s', channel_multiplier=1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_mixnet_m(pretrained=False, **kwargs):
+ """Creates a MixNet Medium model. Tensorflow compatible variant
+ """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_mixnet_m(
+ 'tf_mixnet_m', channel_multiplier=1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_mixnet_l(pretrained=False, **kwargs):
+ """Creates a MixNet Large model. Tensorflow compatible variant
+ """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_mixnet_m(
+ 'tf_mixnet_l', channel_multiplier=1.3, pretrained=pretrained, **kwargs)
+ return model
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/helpers.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/helpers.py
new file mode 100644
index 0000000000000000000000000000000000000000..3f83a07d690c7ad681c777c19b1e7a5bb95da007
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/helpers.py
@@ -0,0 +1,71 @@
+""" Checkpoint loading / state_dict helpers
+Copyright 2020 Ross Wightman
+"""
+import torch
+import os
+from collections import OrderedDict
+try:
+ from torch.hub import load_state_dict_from_url
+except ImportError:
+ from torch.utils.model_zoo import load_url as load_state_dict_from_url
+
+
+def load_checkpoint(model, checkpoint_path):
+ if checkpoint_path and os.path.isfile(checkpoint_path):
+ print("=> Loading checkpoint '{}'".format(checkpoint_path))
+ checkpoint = torch.load(checkpoint_path)
+ if isinstance(checkpoint, dict) and 'state_dict' in checkpoint:
+ new_state_dict = OrderedDict()
+ for k, v in checkpoint['state_dict'].items():
+ if k.startswith('module'):
+ name = k[7:] # remove `module.`
+ else:
+ name = k
+ new_state_dict[name] = v
+ model.load_state_dict(new_state_dict)
+ else:
+ model.load_state_dict(checkpoint)
+ print("=> Loaded checkpoint '{}'".format(checkpoint_path))
+ else:
+ print("=> Error: No checkpoint found at '{}'".format(checkpoint_path))
+ raise FileNotFoundError()
+
+
+def load_pretrained(model, url, filter_fn=None, strict=True):
+ if not url:
+ print("=> Warning: Pretrained model URL is empty, using random initialization.")
+ return
+
+ state_dict = load_state_dict_from_url(url, progress=False, map_location='cpu')
+
+ input_conv = 'conv_stem'
+ classifier = 'classifier'
+ in_chans = getattr(model, input_conv).weight.shape[1]
+ num_classes = getattr(model, classifier).weight.shape[0]
+
+ input_conv_weight = input_conv + '.weight'
+ pretrained_in_chans = state_dict[input_conv_weight].shape[1]
+ if in_chans != pretrained_in_chans:
+ if in_chans == 1:
+ print('=> Converting pretrained input conv {} from {} to 1 channel'.format(
+ input_conv_weight, pretrained_in_chans))
+ conv1_weight = state_dict[input_conv_weight]
+ state_dict[input_conv_weight] = conv1_weight.sum(dim=1, keepdim=True)
+ else:
+ print('=> Discarding pretrained input conv {} since input channel count != {}'.format(
+ input_conv_weight, pretrained_in_chans))
+ del state_dict[input_conv_weight]
+ strict = False
+
+ classifier_weight = classifier + '.weight'
+ pretrained_num_classes = state_dict[classifier_weight].shape[0]
+ if num_classes != pretrained_num_classes:
+ print('=> Discarding pretrained classifier since num_classes != {}'.format(pretrained_num_classes))
+ del state_dict[classifier_weight]
+ del state_dict[classifier + '.bias']
+ strict = False
+
+ if filter_fn is not None:
+ state_dict = filter_fn(state_dict)
+
+ model.load_state_dict(state_dict, strict=strict)
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/mobilenetv3.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/mobilenetv3.py
new file mode 100644
index 0000000000000000000000000000000000000000..b5966c28f7207e98ee50745b1bc8f3663c650f9d
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/mobilenetv3.py
@@ -0,0 +1,364 @@
+""" MobileNet-V3
+
+A PyTorch impl of MobileNet-V3, compatible with TF weights from official impl.
+
+Paper: Searching for MobileNetV3 - https://arxiv.org/abs/1905.02244
+
+Hacked together by / Copyright 2020 Ross Wightman
+"""
+import torch.nn as nn
+import torch.nn.functional as F
+
+from .activations import get_act_fn, get_act_layer, HardSwish
+from .config import layer_config_kwargs
+from .conv2d_layers import select_conv2d
+from .helpers import load_pretrained
+from .efficientnet_builder import *
+
+__all__ = ['mobilenetv3_rw', 'mobilenetv3_large_075', 'mobilenetv3_large_100', 'mobilenetv3_large_minimal_100',
+ 'mobilenetv3_small_075', 'mobilenetv3_small_100', 'mobilenetv3_small_minimal_100',
+ 'tf_mobilenetv3_large_075', 'tf_mobilenetv3_large_100', 'tf_mobilenetv3_large_minimal_100',
+ 'tf_mobilenetv3_small_075', 'tf_mobilenetv3_small_100', 'tf_mobilenetv3_small_minimal_100']
+
+model_urls = {
+ 'mobilenetv3_rw':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/mobilenetv3_100-35495452.pth',
+ 'mobilenetv3_large_075': None,
+ 'mobilenetv3_large_100':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/mobilenetv3_large_100_ra-f55367f5.pth',
+ 'mobilenetv3_large_minimal_100': None,
+ 'mobilenetv3_small_075': None,
+ 'mobilenetv3_small_100': None,
+ 'mobilenetv3_small_minimal_100': None,
+ 'tf_mobilenetv3_large_075':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_mobilenetv3_large_075-150ee8b0.pth',
+ 'tf_mobilenetv3_large_100':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_mobilenetv3_large_100-427764d5.pth',
+ 'tf_mobilenetv3_large_minimal_100':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_mobilenetv3_large_minimal_100-8596ae28.pth',
+ 'tf_mobilenetv3_small_075':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_mobilenetv3_small_075-da427f52.pth',
+ 'tf_mobilenetv3_small_100':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_mobilenetv3_small_100-37f49e2b.pth',
+ 'tf_mobilenetv3_small_minimal_100':
+ 'https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/tf_mobilenetv3_small_minimal_100-922a7843.pth',
+}
+
+
+class MobileNetV3(nn.Module):
+ """ MobileNet-V3
+
+ A this model utilizes the MobileNet-v3 specific 'efficient head', where global pooling is done before the
+ head convolution without a final batch-norm layer before the classifier.
+
+ Paper: https://arxiv.org/abs/1905.02244
+ """
+
+ def __init__(self, block_args, num_classes=1000, in_chans=3, stem_size=16, num_features=1280, head_bias=True,
+ channel_multiplier=1.0, pad_type='', act_layer=HardSwish, drop_rate=0., drop_connect_rate=0.,
+ se_kwargs=None, norm_layer=nn.BatchNorm2d, norm_kwargs=None, weight_init='goog'):
+ super(MobileNetV3, self).__init__()
+ self.drop_rate = drop_rate
+
+ stem_size = round_channels(stem_size, channel_multiplier)
+ self.conv_stem = select_conv2d(in_chans, stem_size, 3, stride=2, padding=pad_type)
+ self.bn1 = nn.BatchNorm2d(stem_size, **norm_kwargs)
+ self.act1 = act_layer(inplace=True)
+ in_chs = stem_size
+
+ builder = EfficientNetBuilder(
+ channel_multiplier, pad_type=pad_type, act_layer=act_layer, se_kwargs=se_kwargs,
+ norm_layer=norm_layer, norm_kwargs=norm_kwargs, drop_connect_rate=drop_connect_rate)
+ self.blocks = nn.Sequential(*builder(in_chs, block_args))
+ in_chs = builder.in_chs
+
+ self.global_pool = nn.AdaptiveAvgPool2d(1)
+ self.conv_head = select_conv2d(in_chs, num_features, 1, padding=pad_type, bias=head_bias)
+ self.act2 = act_layer(inplace=True)
+ self.classifier = nn.Linear(num_features, num_classes)
+
+ for m in self.modules():
+ if weight_init == 'goog':
+ initialize_weight_goog(m)
+ else:
+ initialize_weight_default(m)
+
+ def as_sequential(self):
+ layers = [self.conv_stem, self.bn1, self.act1]
+ layers.extend(self.blocks)
+ layers.extend([
+ self.global_pool, self.conv_head, self.act2,
+ nn.Flatten(), nn.Dropout(self.drop_rate), self.classifier])
+ return nn.Sequential(*layers)
+
+ def features(self, x):
+ x = self.conv_stem(x)
+ x = self.bn1(x)
+ x = self.act1(x)
+ x = self.blocks(x)
+ x = self.global_pool(x)
+ x = self.conv_head(x)
+ x = self.act2(x)
+ return x
+
+ def forward(self, x):
+ x = self.features(x)
+ x = x.flatten(1)
+ if self.drop_rate > 0.:
+ x = F.dropout(x, p=self.drop_rate, training=self.training)
+ return self.classifier(x)
+
+
+def _create_model(model_kwargs, variant, pretrained=False):
+ as_sequential = model_kwargs.pop('as_sequential', False)
+ model = MobileNetV3(**model_kwargs)
+ if pretrained and model_urls[variant]:
+ load_pretrained(model, model_urls[variant])
+ if as_sequential:
+ model = model.as_sequential()
+ return model
+
+
+def _gen_mobilenet_v3_rw(variant, channel_multiplier=1.0, pretrained=False, **kwargs):
+ """Creates a MobileNet-V3 model (RW variant).
+
+ Paper: https://arxiv.org/abs/1905.02244
+
+ This was my first attempt at reproducing the MobileNet-V3 from paper alone. It came close to the
+ eventual Tensorflow reference impl but has a few differences:
+ 1. This model has no bias on the head convolution
+ 2. This model forces no residual (noskip) on the first DWS block, this is different than MnasNet
+ 3. This model always uses ReLU for the SE activation layer, other models in the family inherit their act layer
+ from their parent block
+ 4. This model does not enforce divisible by 8 limitation on the SE reduction channel count
+
+ Overall the changes are fairly minor and result in a very small parameter count difference and no
+ top-1/5
+
+ Args:
+ channel_multiplier: multiplier to number of channels per layer.
+ """
+ arch_def = [
+ # stage 0, 112x112 in
+ ['ds_r1_k3_s1_e1_c16_nre_noskip'], # relu
+ # stage 1, 112x112 in
+ ['ir_r1_k3_s2_e4_c24_nre', 'ir_r1_k3_s1_e3_c24_nre'], # relu
+ # stage 2, 56x56 in
+ ['ir_r3_k5_s2_e3_c40_se0.25_nre'], # relu
+ # stage 3, 28x28 in
+ ['ir_r1_k3_s2_e6_c80', 'ir_r1_k3_s1_e2.5_c80', 'ir_r2_k3_s1_e2.3_c80'], # hard-swish
+ # stage 4, 14x14in
+ ['ir_r2_k3_s1_e6_c112_se0.25'], # hard-swish
+ # stage 5, 14x14in
+ ['ir_r3_k5_s2_e6_c160_se0.25'], # hard-swish
+ # stage 6, 7x7 in
+ ['cn_r1_k1_s1_c960'], # hard-swish
+ ]
+ with layer_config_kwargs(kwargs):
+ model_kwargs = dict(
+ block_args=decode_arch_def(arch_def),
+ head_bias=False, # one of my mistakes
+ channel_multiplier=channel_multiplier,
+ act_layer=resolve_act_layer(kwargs, 'hard_swish'),
+ se_kwargs=dict(gate_fn=get_act_fn('hard_sigmoid'), reduce_mid=True),
+ norm_kwargs=resolve_bn_args(kwargs),
+ **kwargs,
+ )
+ model = _create_model(model_kwargs, variant, pretrained)
+ return model
+
+
+def _gen_mobilenet_v3(variant, channel_multiplier=1.0, pretrained=False, **kwargs):
+ """Creates a MobileNet-V3 large/small/minimal models.
+
+ Ref impl: https://github.com/tensorflow/models/blob/master/research/slim/nets/mobilenet/mobilenet_v3.py
+ Paper: https://arxiv.org/abs/1905.02244
+
+ Args:
+ channel_multiplier: multiplier to number of channels per layer.
+ """
+ if 'small' in variant:
+ num_features = 1024
+ if 'minimal' in variant:
+ act_layer = 'relu'
+ arch_def = [
+ # stage 0, 112x112 in
+ ['ds_r1_k3_s2_e1_c16'],
+ # stage 1, 56x56 in
+ ['ir_r1_k3_s2_e4.5_c24', 'ir_r1_k3_s1_e3.67_c24'],
+ # stage 2, 28x28 in
+ ['ir_r1_k3_s2_e4_c40', 'ir_r2_k3_s1_e6_c40'],
+ # stage 3, 14x14 in
+ ['ir_r2_k3_s1_e3_c48'],
+ # stage 4, 14x14in
+ ['ir_r3_k3_s2_e6_c96'],
+ # stage 6, 7x7 in
+ ['cn_r1_k1_s1_c576'],
+ ]
+ else:
+ act_layer = 'hard_swish'
+ arch_def = [
+ # stage 0, 112x112 in
+ ['ds_r1_k3_s2_e1_c16_se0.25_nre'], # relu
+ # stage 1, 56x56 in
+ ['ir_r1_k3_s2_e4.5_c24_nre', 'ir_r1_k3_s1_e3.67_c24_nre'], # relu
+ # stage 2, 28x28 in
+ ['ir_r1_k5_s2_e4_c40_se0.25', 'ir_r2_k5_s1_e6_c40_se0.25'], # hard-swish
+ # stage 3, 14x14 in
+ ['ir_r2_k5_s1_e3_c48_se0.25'], # hard-swish
+ # stage 4, 14x14in
+ ['ir_r3_k5_s2_e6_c96_se0.25'], # hard-swish
+ # stage 6, 7x7 in
+ ['cn_r1_k1_s1_c576'], # hard-swish
+ ]
+ else:
+ num_features = 1280
+ if 'minimal' in variant:
+ act_layer = 'relu'
+ arch_def = [
+ # stage 0, 112x112 in
+ ['ds_r1_k3_s1_e1_c16'],
+ # stage 1, 112x112 in
+ ['ir_r1_k3_s2_e4_c24', 'ir_r1_k3_s1_e3_c24'],
+ # stage 2, 56x56 in
+ ['ir_r3_k3_s2_e3_c40'],
+ # stage 3, 28x28 in
+ ['ir_r1_k3_s2_e6_c80', 'ir_r1_k3_s1_e2.5_c80', 'ir_r2_k3_s1_e2.3_c80'],
+ # stage 4, 14x14in
+ ['ir_r2_k3_s1_e6_c112'],
+ # stage 5, 14x14in
+ ['ir_r3_k3_s2_e6_c160'],
+ # stage 6, 7x7 in
+ ['cn_r1_k1_s1_c960'],
+ ]
+ else:
+ act_layer = 'hard_swish'
+ arch_def = [
+ # stage 0, 112x112 in
+ ['ds_r1_k3_s1_e1_c16_nre'], # relu
+ # stage 1, 112x112 in
+ ['ir_r1_k3_s2_e4_c24_nre', 'ir_r1_k3_s1_e3_c24_nre'], # relu
+ # stage 2, 56x56 in
+ ['ir_r3_k5_s2_e3_c40_se0.25_nre'], # relu
+ # stage 3, 28x28 in
+ ['ir_r1_k3_s2_e6_c80', 'ir_r1_k3_s1_e2.5_c80', 'ir_r2_k3_s1_e2.3_c80'], # hard-swish
+ # stage 4, 14x14in
+ ['ir_r2_k3_s1_e6_c112_se0.25'], # hard-swish
+ # stage 5, 14x14in
+ ['ir_r3_k5_s2_e6_c160_se0.25'], # hard-swish
+ # stage 6, 7x7 in
+ ['cn_r1_k1_s1_c960'], # hard-swish
+ ]
+ with layer_config_kwargs(kwargs):
+ model_kwargs = dict(
+ block_args=decode_arch_def(arch_def),
+ num_features=num_features,
+ stem_size=16,
+ channel_multiplier=channel_multiplier,
+ act_layer=resolve_act_layer(kwargs, act_layer),
+ se_kwargs=dict(
+ act_layer=get_act_layer('relu'), gate_fn=get_act_fn('hard_sigmoid'), reduce_mid=True, divisor=8),
+ norm_kwargs=resolve_bn_args(kwargs),
+ **kwargs,
+ )
+ model = _create_model(model_kwargs, variant, pretrained)
+ return model
+
+
+def mobilenetv3_rw(pretrained=False, **kwargs):
+ """ MobileNet-V3 RW
+ Attn: See note in gen function for this variant.
+ """
+ # NOTE for train set drop_rate=0.2
+ if pretrained:
+ # pretrained model trained with non-default BN epsilon
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ model = _gen_mobilenet_v3_rw('mobilenetv3_rw', 1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def mobilenetv3_large_075(pretrained=False, **kwargs):
+ """ MobileNet V3 Large 0.75"""
+ # NOTE for train set drop_rate=0.2
+ model = _gen_mobilenet_v3('mobilenetv3_large_075', 0.75, pretrained=pretrained, **kwargs)
+ return model
+
+
+def mobilenetv3_large_100(pretrained=False, **kwargs):
+ """ MobileNet V3 Large 1.0 """
+ # NOTE for train set drop_rate=0.2
+ model = _gen_mobilenet_v3('mobilenetv3_large_100', 1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def mobilenetv3_large_minimal_100(pretrained=False, **kwargs):
+ """ MobileNet V3 Large (Minimalistic) 1.0 """
+ # NOTE for train set drop_rate=0.2
+ model = _gen_mobilenet_v3('mobilenetv3_large_minimal_100', 1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def mobilenetv3_small_075(pretrained=False, **kwargs):
+ """ MobileNet V3 Small 0.75 """
+ model = _gen_mobilenet_v3('mobilenetv3_small_075', 0.75, pretrained=pretrained, **kwargs)
+ return model
+
+
+def mobilenetv3_small_100(pretrained=False, **kwargs):
+ """ MobileNet V3 Small 1.0 """
+ model = _gen_mobilenet_v3('mobilenetv3_small_100', 1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def mobilenetv3_small_minimal_100(pretrained=False, **kwargs):
+ """ MobileNet V3 Small (Minimalistic) 1.0 """
+ model = _gen_mobilenet_v3('mobilenetv3_small_minimal_100', 1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_mobilenetv3_large_075(pretrained=False, **kwargs):
+ """ MobileNet V3 Large 0.75. Tensorflow compat variant. """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_mobilenet_v3('tf_mobilenetv3_large_075', 0.75, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_mobilenetv3_large_100(pretrained=False, **kwargs):
+ """ MobileNet V3 Large 1.0. Tensorflow compat variant. """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_mobilenet_v3('tf_mobilenetv3_large_100', 1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_mobilenetv3_large_minimal_100(pretrained=False, **kwargs):
+ """ MobileNet V3 Large Minimalistic 1.0. Tensorflow compat variant. """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_mobilenet_v3('tf_mobilenetv3_large_minimal_100', 1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_mobilenetv3_small_075(pretrained=False, **kwargs):
+ """ MobileNet V3 Small 0.75. Tensorflow compat variant. """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_mobilenet_v3('tf_mobilenetv3_small_075', 0.75, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_mobilenetv3_small_100(pretrained=False, **kwargs):
+ """ MobileNet V3 Small 1.0. Tensorflow compat variant."""
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_mobilenet_v3('tf_mobilenetv3_small_100', 1.0, pretrained=pretrained, **kwargs)
+ return model
+
+
+def tf_mobilenetv3_small_minimal_100(pretrained=False, **kwargs):
+ """ MobileNet V3 Small Minimalistic 1.0. Tensorflow compat variant. """
+ kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
+ kwargs['pad_type'] = 'same'
+ model = _gen_mobilenet_v3('tf_mobilenetv3_small_minimal_100', 1.0, pretrained=pretrained, **kwargs)
+ return model
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/model_factory.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/model_factory.py
new file mode 100644
index 0000000000000000000000000000000000000000..4d46ea8baedaf3d787826eb3bb314b4230514647
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/model_factory.py
@@ -0,0 +1,27 @@
+from .config import set_layer_config
+from .helpers import load_checkpoint
+
+from .gen_efficientnet import *
+from .mobilenetv3 import *
+
+
+def create_model(
+ model_name='mnasnet_100',
+ pretrained=None,
+ num_classes=1000,
+ in_chans=3,
+ checkpoint_path='',
+ **kwargs):
+
+ model_kwargs = dict(num_classes=num_classes, in_chans=in_chans, pretrained=pretrained, **kwargs)
+
+ if model_name in globals():
+ create_fn = globals()[model_name]
+ model = create_fn(**model_kwargs)
+ else:
+ raise RuntimeError('Unknown model (%s)' % model_name)
+
+ if checkpoint_path and not pretrained:
+ load_checkpoint(model, checkpoint_path)
+
+ return model
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/version.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/version.py
new file mode 100644
index 0000000000000000000000000000000000000000..a6221b3de7b1490c5e712e8b5fcc94c3d9d04295
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/geffnet/version.py
@@ -0,0 +1 @@
+__version__ = '1.0.2'
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/hubconf.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/hubconf.py
new file mode 100644
index 0000000000000000000000000000000000000000..45b17b99bbeba34596569e6e50f6e8a2ebc45c54
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/hubconf.py
@@ -0,0 +1,84 @@
+dependencies = ['torch', 'math']
+
+from geffnet import efficientnet_b0
+from geffnet import efficientnet_b1
+from geffnet import efficientnet_b2
+from geffnet import efficientnet_b3
+
+from geffnet import efficientnet_es
+
+from geffnet import efficientnet_lite0
+
+from geffnet import mixnet_s
+from geffnet import mixnet_m
+from geffnet import mixnet_l
+from geffnet import mixnet_xl
+
+from geffnet import mobilenetv2_100
+from geffnet import mobilenetv2_110d
+from geffnet import mobilenetv2_120d
+from geffnet import mobilenetv2_140
+
+from geffnet import mobilenetv3_large_100
+from geffnet import mobilenetv3_rw
+from geffnet import mnasnet_a1
+from geffnet import mnasnet_b1
+from geffnet import fbnetc_100
+from geffnet import spnasnet_100
+
+from geffnet import tf_efficientnet_b0
+from geffnet import tf_efficientnet_b1
+from geffnet import tf_efficientnet_b2
+from geffnet import tf_efficientnet_b3
+from geffnet import tf_efficientnet_b4
+from geffnet import tf_efficientnet_b5
+from geffnet import tf_efficientnet_b6
+from geffnet import tf_efficientnet_b7
+from geffnet import tf_efficientnet_b8
+
+from geffnet import tf_efficientnet_b0_ap
+from geffnet import tf_efficientnet_b1_ap
+from geffnet import tf_efficientnet_b2_ap
+from geffnet import tf_efficientnet_b3_ap
+from geffnet import tf_efficientnet_b4_ap
+from geffnet import tf_efficientnet_b5_ap
+from geffnet import tf_efficientnet_b6_ap
+from geffnet import tf_efficientnet_b7_ap
+from geffnet import tf_efficientnet_b8_ap
+
+from geffnet import tf_efficientnet_b0_ns
+from geffnet import tf_efficientnet_b1_ns
+from geffnet import tf_efficientnet_b2_ns
+from geffnet import tf_efficientnet_b3_ns
+from geffnet import tf_efficientnet_b4_ns
+from geffnet import tf_efficientnet_b5_ns
+from geffnet import tf_efficientnet_b6_ns
+from geffnet import tf_efficientnet_b7_ns
+from geffnet import tf_efficientnet_l2_ns_475
+from geffnet import tf_efficientnet_l2_ns
+
+from geffnet import tf_efficientnet_es
+from geffnet import tf_efficientnet_em
+from geffnet import tf_efficientnet_el
+
+from geffnet import tf_efficientnet_cc_b0_4e
+from geffnet import tf_efficientnet_cc_b0_8e
+from geffnet import tf_efficientnet_cc_b1_8e
+
+from geffnet import tf_efficientnet_lite0
+from geffnet import tf_efficientnet_lite1
+from geffnet import tf_efficientnet_lite2
+from geffnet import tf_efficientnet_lite3
+from geffnet import tf_efficientnet_lite4
+
+from geffnet import tf_mixnet_s
+from geffnet import tf_mixnet_m
+from geffnet import tf_mixnet_l
+
+from geffnet import tf_mobilenetv3_large_075
+from geffnet import tf_mobilenetv3_large_100
+from geffnet import tf_mobilenetv3_large_minimal_100
+from geffnet import tf_mobilenetv3_small_075
+from geffnet import tf_mobilenetv3_small_100
+from geffnet import tf_mobilenetv3_small_minimal_100
+
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/onnx_export.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/onnx_export.py
new file mode 100644
index 0000000000000000000000000000000000000000..7a5162ce214830df501bdb81edb66c095122f69d
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/onnx_export.py
@@ -0,0 +1,120 @@
+""" ONNX export script
+
+Export PyTorch models as ONNX graphs.
+
+This export script originally started as an adaptation of code snippets found at
+https://pytorch.org/tutorials/advanced/super_resolution_with_onnxruntime.html
+
+The default parameters work with PyTorch 1.6 and ONNX 1.7 and produce an optimal ONNX graph
+for hosting in the ONNX runtime (see onnx_validate.py). To export an ONNX model compatible
+with caffe2 (see caffe2_benchmark.py and caffe2_validate.py), the --keep-init and --aten-fallback
+flags are currently required.
+
+Older versions of PyTorch/ONNX (tested PyTorch 1.4, ONNX 1.5) do not need extra flags for
+caffe2 compatibility, but they produce a model that isn't as fast running on ONNX runtime.
+
+Most new release of PyTorch and ONNX cause some sort of breakage in the export / usage of ONNX models.
+Please do your research and search ONNX and PyTorch issue tracker before asking me. Thanks.
+
+Copyright 2020 Ross Wightman
+"""
+import argparse
+import torch
+import numpy as np
+
+import onnx
+import geffnet
+
+parser = argparse.ArgumentParser(description='PyTorch ImageNet Validation')
+parser.add_argument('output', metavar='ONNX_FILE',
+ help='output model filename')
+parser.add_argument('--model', '-m', metavar='MODEL', default='mobilenetv3_large_100',
+ help='model architecture (default: mobilenetv3_large_100)')
+parser.add_argument('--opset', type=int, default=10,
+ help='ONNX opset to use (default: 10)')
+parser.add_argument('--keep-init', action='store_true', default=False,
+ help='Keep initializers as input. Needed for Caffe2 compatible export in newer PyTorch/ONNX.')
+parser.add_argument('--aten-fallback', action='store_true', default=False,
+ help='Fallback to ATEN ops. Helps fix AdaptiveAvgPool issue with Caffe2 in newer PyTorch/ONNX.')
+parser.add_argument('--dynamic-size', action='store_true', default=False,
+ help='Export model width dynamic width/height. Not recommended for "tf" models with SAME padding.')
+parser.add_argument('-b', '--batch-size', default=1, type=int,
+ metavar='N', help='mini-batch size (default: 1)')
+parser.add_argument('--img-size', default=None, type=int,
+ metavar='N', help='Input image dimension, uses model default if empty')
+parser.add_argument('--mean', type=float, nargs='+', default=None, metavar='MEAN',
+ help='Override mean pixel value of dataset')
+parser.add_argument('--std', type=float, nargs='+', default=None, metavar='STD',
+ help='Override std deviation of of dataset')
+parser.add_argument('--num-classes', type=int, default=1000,
+ help='Number classes in dataset')
+parser.add_argument('--checkpoint', default='', type=str, metavar='PATH',
+ help='path to checkpoint (default: none)')
+
+
+def main():
+ args = parser.parse_args()
+
+ args.pretrained = True
+ if args.checkpoint:
+ args.pretrained = False
+
+ print("==> Creating PyTorch {} model".format(args.model))
+ # NOTE exportable=True flag disables autofn/jit scripted activations and uses Conv2dSameExport layers
+ # for models using SAME padding
+ model = geffnet.create_model(
+ args.model,
+ num_classes=args.num_classes,
+ in_chans=3,
+ pretrained=args.pretrained,
+ checkpoint_path=args.checkpoint,
+ exportable=True)
+
+ model.eval()
+
+ example_input = torch.randn((args.batch_size, 3, args.img_size or 224, args.img_size or 224), requires_grad=True)
+
+ # Run model once before export trace, sets padding for models with Conv2dSameExport. This means
+ # that the padding for models with Conv2dSameExport (most models with tf_ prefix) is fixed for
+ # the input img_size specified in this script.
+ # Opset >= 11 should allow for dynamic padding, however I cannot get it to work due to
+ # issues in the tracing of the dynamic padding or errors attempting to export the model after jit
+ # scripting it (an approach that should work). Perhaps in a future PyTorch or ONNX versions...
+ model(example_input)
+
+ print("==> Exporting model to ONNX format at '{}'".format(args.output))
+ input_names = ["input0"]
+ output_names = ["output0"]
+ dynamic_axes = {'input0': {0: 'batch'}, 'output0': {0: 'batch'}}
+ if args.dynamic_size:
+ dynamic_axes['input0'][2] = 'height'
+ dynamic_axes['input0'][3] = 'width'
+ if args.aten_fallback:
+ export_type = torch.onnx.OperatorExportTypes.ONNX_ATEN_FALLBACK
+ else:
+ export_type = torch.onnx.OperatorExportTypes.ONNX
+
+ torch_out = torch.onnx._export(
+ model, example_input, args.output, export_params=True, verbose=True, input_names=input_names,
+ output_names=output_names, keep_initializers_as_inputs=args.keep_init, dynamic_axes=dynamic_axes,
+ opset_version=args.opset, operator_export_type=export_type)
+
+ print("==> Loading and checking exported model from '{}'".format(args.output))
+ onnx_model = onnx.load(args.output)
+ onnx.checker.check_model(onnx_model) # assuming throw on error
+ print("==> Passed")
+
+ if args.keep_init and args.aten_fallback:
+ import caffe2.python.onnx.backend as onnx_caffe2
+ # Caffe2 loading only works properly in newer PyTorch/ONNX combos when
+ # keep_initializers_as_inputs and aten_fallback are set to True.
+ print("==> Loading model into Caffe2 backend and comparing forward pass.".format(args.output))
+ caffe2_backend = onnx_caffe2.prepare(onnx_model)
+ B = {onnx_model.graph.input[0].name: x.data.numpy()}
+ c2_out = caffe2_backend.run(B)[0]
+ np.testing.assert_almost_equal(torch_out.data.numpy(), c2_out, decimal=5)
+ print("==> Passed")
+
+
+if __name__ == '__main__':
+ main()
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/onnx_optimize.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/onnx_optimize.py
new file mode 100644
index 0000000000000000000000000000000000000000..ee20bbf9f0f9473370489512eb96ca0b570b5388
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/onnx_optimize.py
@@ -0,0 +1,84 @@
+""" ONNX optimization script
+
+Run ONNX models through the optimizer to prune unneeded nodes, fuse batchnorm layers into conv, etc.
+
+NOTE: This isn't working consistently in recent PyTorch/ONNX combos (ie PyTorch 1.6 and ONNX 1.7),
+it seems time to switch to using the onnxruntime online optimizer (can also be saved for offline).
+
+Copyright 2020 Ross Wightman
+"""
+import argparse
+import warnings
+
+import onnx
+from onnx import optimizer
+
+
+parser = argparse.ArgumentParser(description="Optimize ONNX model")
+
+parser.add_argument("model", help="The ONNX model")
+parser.add_argument("--output", required=True, help="The optimized model output filename")
+
+
+def traverse_graph(graph, prefix=''):
+ content = []
+ indent = prefix + ' '
+ graphs = []
+ num_nodes = 0
+ for node in graph.node:
+ pn, gs = onnx.helper.printable_node(node, indent, subgraphs=True)
+ assert isinstance(gs, list)
+ content.append(pn)
+ graphs.extend(gs)
+ num_nodes += 1
+ for g in graphs:
+ g_count, g_str = traverse_graph(g)
+ content.append('\n' + g_str)
+ num_nodes += g_count
+ return num_nodes, '\n'.join(content)
+
+
+def main():
+ args = parser.parse_args()
+ onnx_model = onnx.load(args.model)
+ num_original_nodes, original_graph_str = traverse_graph(onnx_model.graph)
+
+ # Optimizer passes to perform
+ passes = [
+ #'eliminate_deadend',
+ 'eliminate_identity',
+ 'eliminate_nop_dropout',
+ 'eliminate_nop_pad',
+ 'eliminate_nop_transpose',
+ 'eliminate_unused_initializer',
+ 'extract_constant_to_initializer',
+ 'fuse_add_bias_into_conv',
+ 'fuse_bn_into_conv',
+ 'fuse_consecutive_concats',
+ 'fuse_consecutive_reduce_unsqueeze',
+ 'fuse_consecutive_squeezes',
+ 'fuse_consecutive_transposes',
+ #'fuse_matmul_add_bias_into_gemm',
+ 'fuse_pad_into_conv',
+ #'fuse_transpose_into_gemm',
+ #'lift_lexical_references',
+ ]
+
+ # Apply the optimization on the original serialized model
+ # WARNING I've had issues with optimizer in recent versions of PyTorch / ONNX causing
+ # 'duplicate definition of name' errors, see: https://github.com/onnx/onnx/issues/2401
+ # It may be better to rely on onnxruntime optimizations, see onnx_validate.py script.
+ warnings.warn("I've had issues with optimizer in recent versions of PyTorch / ONNX."
+ "Try onnxruntime optimization if this doesn't work.")
+ optimized_model = optimizer.optimize(onnx_model, passes)
+
+ num_optimized_nodes, optimzied_graph_str = traverse_graph(optimized_model.graph)
+ print('==> The model after optimization:\n{}\n'.format(optimzied_graph_str))
+ print('==> The optimized model has {} nodes, the original had {}.'.format(num_optimized_nodes, num_original_nodes))
+
+ # Save the ONNX model
+ onnx.save(optimized_model, args.output)
+
+
+if __name__ == "__main__":
+ main()
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/onnx_to_caffe.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/onnx_to_caffe.py
new file mode 100644
index 0000000000000000000000000000000000000000..44399aafababcdf6b84147a0613eb0909730db4b
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/onnx_to_caffe.py
@@ -0,0 +1,27 @@
+import argparse
+
+import onnx
+from caffe2.python.onnx.backend import Caffe2Backend
+
+
+parser = argparse.ArgumentParser(description="Convert ONNX to Caffe2")
+
+parser.add_argument("model", help="The ONNX model")
+parser.add_argument("--c2-prefix", required=True,
+ help="The output file prefix for the caffe2 model init and predict file. ")
+
+
+def main():
+ args = parser.parse_args()
+ onnx_model = onnx.load(args.model)
+ caffe2_init, caffe2_predict = Caffe2Backend.onnx_graph_to_caffe2_net(onnx_model)
+ caffe2_init_str = caffe2_init.SerializeToString()
+ with open(args.c2_prefix + '.init.pb', "wb") as f:
+ f.write(caffe2_init_str)
+ caffe2_predict_str = caffe2_predict.SerializeToString()
+ with open(args.c2_prefix + '.predict.pb', "wb") as f:
+ f.write(caffe2_predict_str)
+
+
+if __name__ == "__main__":
+ main()
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/onnx_validate.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/onnx_validate.py
new file mode 100644
index 0000000000000000000000000000000000000000..ab3e4fb141b6ef660dcc5b447fd9f368a2ea19a0
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/onnx_validate.py
@@ -0,0 +1,112 @@
+""" ONNX-runtime validation script
+
+This script was created to verify accuracy and performance of exported ONNX
+models running with the onnxruntime. It utilizes the PyTorch dataloader/processing
+pipeline for a fair comparison against the originals.
+
+Copyright 2020 Ross Wightman
+"""
+import argparse
+import numpy as np
+import onnxruntime
+from data import create_loader, resolve_data_config, Dataset
+from utils import AverageMeter
+import time
+
+parser = argparse.ArgumentParser(description='Caffe2 ImageNet Validation')
+parser.add_argument('data', metavar='DIR',
+ help='path to dataset')
+parser.add_argument('--onnx-input', default='', type=str, metavar='PATH',
+ help='path to onnx model/weights file')
+parser.add_argument('--onnx-output-opt', default='', type=str, metavar='PATH',
+ help='path to output optimized onnx graph')
+parser.add_argument('--profile', action='store_true', default=False,
+ help='Enable profiler output.')
+parser.add_argument('-j', '--workers', default=2, type=int, metavar='N',
+ help='number of data loading workers (default: 2)')
+parser.add_argument('-b', '--batch-size', default=256, type=int,
+ metavar='N', help='mini-batch size (default: 256)')
+parser.add_argument('--img-size', default=None, type=int,
+ metavar='N', help='Input image dimension, uses model default if empty')
+parser.add_argument('--mean', type=float, nargs='+', default=None, metavar='MEAN',
+ help='Override mean pixel value of dataset')
+parser.add_argument('--std', type=float, nargs='+', default=None, metavar='STD',
+ help='Override std deviation of of dataset')
+parser.add_argument('--crop-pct', type=float, default=None, metavar='PCT',
+ help='Override default crop pct of 0.875')
+parser.add_argument('--interpolation', default='', type=str, metavar='NAME',
+ help='Image resize interpolation type (overrides model)')
+parser.add_argument('--tf-preprocessing', dest='tf_preprocessing', action='store_true',
+ help='use tensorflow mnasnet preporcessing')
+parser.add_argument('--print-freq', '-p', default=10, type=int,
+ metavar='N', help='print frequency (default: 10)')
+
+
+def main():
+ args = parser.parse_args()
+ args.gpu_id = 0
+
+ # Set graph optimization level
+ sess_options = onnxruntime.SessionOptions()
+ sess_options.graph_optimization_level = onnxruntime.GraphOptimizationLevel.ORT_ENABLE_ALL
+ if args.profile:
+ sess_options.enable_profiling = True
+ if args.onnx_output_opt:
+ sess_options.optimized_model_filepath = args.onnx_output_opt
+
+ session = onnxruntime.InferenceSession(args.onnx_input, sess_options)
+
+ data_config = resolve_data_config(None, args)
+ loader = create_loader(
+ Dataset(args.data, load_bytes=args.tf_preprocessing),
+ input_size=data_config['input_size'],
+ batch_size=args.batch_size,
+ use_prefetcher=False,
+ interpolation=data_config['interpolation'],
+ mean=data_config['mean'],
+ std=data_config['std'],
+ num_workers=args.workers,
+ crop_pct=data_config['crop_pct'],
+ tensorflow_preprocessing=args.tf_preprocessing)
+
+ input_name = session.get_inputs()[0].name
+
+ batch_time = AverageMeter()
+ top1 = AverageMeter()
+ top5 = AverageMeter()
+ end = time.time()
+ for i, (input, target) in enumerate(loader):
+ # run the net and return prediction
+ output = session.run([], {input_name: input.data.numpy()})
+ output = output[0]
+
+ # measure accuracy and record loss
+ prec1, prec5 = accuracy_np(output, target.numpy())
+ top1.update(prec1.item(), input.size(0))
+ top5.update(prec5.item(), input.size(0))
+
+ # measure elapsed time
+ batch_time.update(time.time() - end)
+ end = time.time()
+
+ if i % args.print_freq == 0:
+ print('Test: [{0}/{1}]\t'
+ 'Time {batch_time.val:.3f} ({batch_time.avg:.3f}, {rate_avg:.3f}/s, {ms_avg:.3f} ms/sample) \t'
+ 'Prec@1 {top1.val:.3f} ({top1.avg:.3f})\t'
+ 'Prec@5 {top5.val:.3f} ({top5.avg:.3f})'.format(
+ i, len(loader), batch_time=batch_time, rate_avg=input.size(0) / batch_time.avg,
+ ms_avg=100 * batch_time.avg / input.size(0), top1=top1, top5=top5))
+
+ print(' * Prec@1 {top1.avg:.3f} ({top1a:.3f}) Prec@5 {top5.avg:.3f} ({top5a:.3f})'.format(
+ top1=top1, top1a=100-top1.avg, top5=top5, top5a=100.-top5.avg))
+
+
+def accuracy_np(output, target):
+ max_indices = np.argsort(output, axis=1)[:, ::-1]
+ top5 = 100 * np.equal(max_indices[:, :5], target[:, np.newaxis]).sum(axis=1).mean()
+ top1 = 100 * np.equal(max_indices[:, 0], target).mean()
+ return top1, top5
+
+
+if __name__ == '__main__':
+ main()
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/requirements.txt b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/requirements.txt
new file mode 100644
index 0000000000000000000000000000000000000000..ac3ffc13bae15f9b11f7cbe3705760056ecd7f13
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/requirements.txt
@@ -0,0 +1,2 @@
+torch>=1.2.0
+torchvision>=0.4.0
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/setup.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/setup.py
new file mode 100644
index 0000000000000000000000000000000000000000..023e4c30f98164595964423e3a83eefaf7ffdad6
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/setup.py
@@ -0,0 +1,47 @@
+""" Setup
+"""
+from setuptools import setup, find_packages
+from codecs import open
+from os import path
+
+here = path.abspath(path.dirname(__file__))
+
+# Get the long description from the README file
+with open(path.join(here, 'README.md'), encoding='utf-8') as f:
+ long_description = f.read()
+
+exec(open('geffnet/version.py').read())
+setup(
+ name='geffnet',
+ version=__version__,
+ description='(Generic) EfficientNets for PyTorch',
+ long_description=long_description,
+ long_description_content_type='text/markdown',
+ url='https://github.com/rwightman/gen-efficientnet-pytorch',
+ author='Ross Wightman',
+ author_email='hello@rwightman.com',
+ classifiers=[
+ # How mature is this project? Common values are
+ # 3 - Alpha
+ # 4 - Beta
+ # 5 - Production/Stable
+ 'Development Status :: 3 - Alpha',
+ 'Intended Audience :: Education',
+ 'Intended Audience :: Science/Research',
+ 'License :: OSI Approved :: Apache Software License',
+ 'Programming Language :: Python :: 3.6',
+ 'Programming Language :: Python :: 3.7',
+ 'Programming Language :: Python :: 3.8',
+ 'Topic :: Scientific/Engineering',
+ 'Topic :: Scientific/Engineering :: Artificial Intelligence',
+ 'Topic :: Software Development',
+ 'Topic :: Software Development :: Libraries',
+ 'Topic :: Software Development :: Libraries :: Python Modules',
+ ],
+
+ # Note that this is a string of words separated by whitespace, not a list.
+ keywords='pytorch pretrained models efficientnet mixnet mobilenetv3 mnasnet',
+ packages=find_packages(exclude=['data']),
+ install_requires=['torch >= 1.4', 'torchvision'],
+ python_requires='>=3.6',
+)
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/utils.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/utils.py
new file mode 100644
index 0000000000000000000000000000000000000000..d327e8bd8120c5cd09ae6c15c3991ccbe27f6c1f
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/utils.py
@@ -0,0 +1,52 @@
+import os
+
+
+class AverageMeter:
+ """Computes and stores the average and current value"""
+ def __init__(self):
+ self.reset()
+
+ def reset(self):
+ self.val = 0
+ self.avg = 0
+ self.sum = 0
+ self.count = 0
+
+ def update(self, val, n=1):
+ self.val = val
+ self.sum += val * n
+ self.count += n
+ self.avg = self.sum / self.count
+
+
+def accuracy(output, target, topk=(1,)):
+ """Computes the precision@k for the specified values of k"""
+ maxk = max(topk)
+ batch_size = target.size(0)
+
+ _, pred = output.topk(maxk, 1, True, True)
+ pred = pred.t()
+ correct = pred.eq(target.view(1, -1).expand_as(pred))
+
+ res = []
+ for k in topk:
+ correct_k = correct[:k].reshape(-1).float().sum(0)
+ res.append(correct_k.mul_(100.0 / batch_size))
+ return res
+
+
+def get_outdir(path, *paths, inc=False):
+ outdir = os.path.join(path, *paths)
+ if not os.path.exists(outdir):
+ os.makedirs(outdir)
+ elif inc:
+ count = 1
+ outdir_inc = outdir + '-' + str(count)
+ while os.path.exists(outdir_inc):
+ count = count + 1
+ outdir_inc = outdir + '-' + str(count)
+ assert count < 100
+ outdir = outdir_inc
+ os.makedirs(outdir)
+ return outdir
+
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/validate.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/validate.py
new file mode 100644
index 0000000000000000000000000000000000000000..5fd44fbb3165ef81ef81251b6299f6aaa80bf2c2
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/efficientnet_repo/validate.py
@@ -0,0 +1,166 @@
+from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
+
+import argparse
+import time
+import torch
+import torch.nn as nn
+import torch.nn.parallel
+from contextlib import suppress
+
+import geffnet
+from data import Dataset, create_loader, resolve_data_config
+from utils import accuracy, AverageMeter
+
+has_native_amp = False
+try:
+ if getattr(torch.cuda.amp, 'autocast') is not None:
+ has_native_amp = True
+except AttributeError:
+ pass
+
+torch.backends.cudnn.benchmark = True
+
+parser = argparse.ArgumentParser(description='PyTorch ImageNet Validation')
+parser.add_argument('data', metavar='DIR',
+ help='path to dataset')
+parser.add_argument('--model', '-m', metavar='MODEL', default='spnasnet1_00',
+ help='model architecture (default: dpn92)')
+parser.add_argument('-j', '--workers', default=4, type=int, metavar='N',
+ help='number of data loading workers (default: 2)')
+parser.add_argument('-b', '--batch-size', default=256, type=int,
+ metavar='N', help='mini-batch size (default: 256)')
+parser.add_argument('--img-size', default=None, type=int,
+ metavar='N', help='Input image dimension, uses model default if empty')
+parser.add_argument('--mean', type=float, nargs='+', default=None, metavar='MEAN',
+ help='Override mean pixel value of dataset')
+parser.add_argument('--std', type=float, nargs='+', default=None, metavar='STD',
+ help='Override std deviation of of dataset')
+parser.add_argument('--crop-pct', type=float, default=None, metavar='PCT',
+ help='Override default crop pct of 0.875')
+parser.add_argument('--interpolation', default='', type=str, metavar='NAME',
+ help='Image resize interpolation type (overrides model)')
+parser.add_argument('--num-classes', type=int, default=1000,
+ help='Number classes in dataset')
+parser.add_argument('--print-freq', '-p', default=10, type=int,
+ metavar='N', help='print frequency (default: 10)')
+parser.add_argument('--checkpoint', default='', type=str, metavar='PATH',
+ help='path to latest checkpoint (default: none)')
+parser.add_argument('--pretrained', dest='pretrained', action='store_true',
+ help='use pre-trained model')
+parser.add_argument('--torchscript', dest='torchscript', action='store_true',
+ help='convert model torchscript for inference')
+parser.add_argument('--num-gpu', type=int, default=1,
+ help='Number of GPUS to use')
+parser.add_argument('--tf-preprocessing', dest='tf_preprocessing', action='store_true',
+ help='use tensorflow mnasnet preporcessing')
+parser.add_argument('--no-cuda', dest='no_cuda', action='store_true',
+ help='')
+parser.add_argument('--channels-last', action='store_true', default=False,
+ help='Use channels_last memory layout')
+parser.add_argument('--amp', action='store_true', default=False,
+ help='Use native Torch AMP mixed precision.')
+
+
+def main():
+ args = parser.parse_args()
+
+ if not args.checkpoint and not args.pretrained:
+ args.pretrained = True
+
+ amp_autocast = suppress # do nothing
+ if args.amp:
+ if not has_native_amp:
+ print("Native Torch AMP is not available (requires torch >= 1.6), using FP32.")
+ else:
+ amp_autocast = torch.cuda.amp.autocast
+
+ # create model
+ model = geffnet.create_model(
+ args.model,
+ num_classes=args.num_classes,
+ in_chans=3,
+ pretrained=args.pretrained,
+ checkpoint_path=args.checkpoint,
+ scriptable=args.torchscript)
+
+ if args.channels_last:
+ model = model.to(memory_format=torch.channels_last)
+
+ if args.torchscript:
+ torch.jit.optimized_execution(True)
+ model = torch.jit.script(model)
+
+ print('Model %s created, param count: %d' %
+ (args.model, sum([m.numel() for m in model.parameters()])))
+
+ data_config = resolve_data_config(model, args)
+
+ criterion = nn.CrossEntropyLoss()
+
+ if not args.no_cuda:
+ if args.num_gpu > 1:
+ model = torch.nn.DataParallel(model, device_ids=list(range(args.num_gpu))).cuda()
+ else:
+ model = model.cuda()
+ criterion = criterion.cuda()
+
+ loader = create_loader(
+ Dataset(args.data, load_bytes=args.tf_preprocessing),
+ input_size=data_config['input_size'],
+ batch_size=args.batch_size,
+ use_prefetcher=not args.no_cuda,
+ interpolation=data_config['interpolation'],
+ mean=data_config['mean'],
+ std=data_config['std'],
+ num_workers=args.workers,
+ crop_pct=data_config['crop_pct'],
+ tensorflow_preprocessing=args.tf_preprocessing)
+
+ batch_time = AverageMeter()
+ losses = AverageMeter()
+ top1 = AverageMeter()
+ top5 = AverageMeter()
+
+ model.eval()
+ end = time.time()
+ with torch.no_grad():
+ for i, (input, target) in enumerate(loader):
+ if not args.no_cuda:
+ target = target.cuda()
+ input = input.cuda()
+ if args.channels_last:
+ input = input.contiguous(memory_format=torch.channels_last)
+
+ # compute output
+ with amp_autocast():
+ output = model(input)
+ loss = criterion(output, target)
+
+ # measure accuracy and record loss
+ prec1, prec5 = accuracy(output.data, target, topk=(1, 5))
+ losses.update(loss.item(), input.size(0))
+ top1.update(prec1.item(), input.size(0))
+ top5.update(prec5.item(), input.size(0))
+
+ # measure elapsed time
+ batch_time.update(time.time() - end)
+ end = time.time()
+
+ if i % args.print_freq == 0:
+ print('Test: [{0}/{1}]\t'
+ 'Time {batch_time.val:.3f} ({batch_time.avg:.3f}, {rate_avg:.3f}/s) \t'
+ 'Loss {loss.val:.4f} ({loss.avg:.4f})\t'
+ 'Prec@1 {top1.val:.3f} ({top1.avg:.3f})\t'
+ 'Prec@5 {top5.val:.3f} ({top5.avg:.3f})'.format(
+ i, len(loader), batch_time=batch_time,
+ rate_avg=input.size(0) / batch_time.avg,
+ loss=losses, top1=top1, top5=top5))
+
+ print(' * Prec@1 {top1.avg:.3f} ({top1a:.3f}) Prec@5 {top5.avg:.3f} ({top5a:.3f})'.format(
+ top1=top1, top1a=100-top1.avg, top5=top5, top5a=100.-top5.avg))
+
+
+if __name__ == '__main__':
+ main()
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/encoder.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/encoder.py
new file mode 100644
index 0000000000000000000000000000000000000000..7f7149ca3c0cf2b6e019105af7e645cfbb3eda11
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/encoder.py
@@ -0,0 +1,34 @@
+import os
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+
+
+class Encoder(nn.Module):
+ def __init__(self):
+ super(Encoder, self).__init__()
+
+ basemodel_name = 'tf_efficientnet_b5_ap'
+ print('Loading base model ()...'.format(basemodel_name), end='')
+ repo_path = os.path.join(os.path.dirname(__file__), 'efficientnet_repo')
+ basemodel = torch.hub.load(repo_path, basemodel_name, pretrained=False, source='local')
+ print('Done.')
+
+ # Remove last layer
+ print('Removing last two layers (global_pool & classifier).')
+ basemodel.global_pool = nn.Identity()
+ basemodel.classifier = nn.Identity()
+
+ self.original_model = basemodel
+
+ def forward(self, x):
+ features = [x]
+ for k, v in self.original_model._modules.items():
+ if (k == 'blocks'):
+ for ki, vi in v._modules.items():
+ features.append(vi(features[-1]))
+ else:
+ features.append(v(features[-1]))
+ return features
+
+
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/submodules.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/submodules.py
new file mode 100644
index 0000000000000000000000000000000000000000..409733351bd6ab5d191c800aff1bc05bfa4cb6f8
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/models/submodules/submodules.py
@@ -0,0 +1,140 @@
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+
+
+########################################################################################################################
+
+
+# Upsample + BatchNorm
+class UpSampleBN(nn.Module):
+ def __init__(self, skip_input, output_features):
+ super(UpSampleBN, self).__init__()
+
+ self._net = nn.Sequential(nn.Conv2d(skip_input, output_features, kernel_size=3, stride=1, padding=1),
+ nn.BatchNorm2d(output_features),
+ nn.LeakyReLU(),
+ nn.Conv2d(output_features, output_features, kernel_size=3, stride=1, padding=1),
+ nn.BatchNorm2d(output_features),
+ nn.LeakyReLU())
+
+ def forward(self, x, concat_with):
+ up_x = F.interpolate(x, size=[concat_with.size(2), concat_with.size(3)], mode='bilinear', align_corners=True)
+ f = torch.cat([up_x, concat_with], dim=1)
+ return self._net(f)
+
+
+# Upsample + GroupNorm + Weight Standardization
+class UpSampleGN(nn.Module):
+ def __init__(self, skip_input, output_features):
+ super(UpSampleGN, self).__init__()
+
+ self._net = nn.Sequential(Conv2d(skip_input, output_features, kernel_size=3, stride=1, padding=1),
+ nn.GroupNorm(8, output_features),
+ nn.LeakyReLU(),
+ Conv2d(output_features, output_features, kernel_size=3, stride=1, padding=1),
+ nn.GroupNorm(8, output_features),
+ nn.LeakyReLU())
+
+ def forward(self, x, concat_with):
+ up_x = F.interpolate(x, size=[concat_with.size(2), concat_with.size(3)], mode='bilinear', align_corners=True)
+ f = torch.cat([up_x, concat_with], dim=1)
+ return self._net(f)
+
+
+# Conv2d with weight standardization
+class Conv2d(nn.Conv2d):
+ def __init__(self, in_channels, out_channels, kernel_size, stride=1,
+ padding=0, dilation=1, groups=1, bias=True):
+ super(Conv2d, self).__init__(in_channels, out_channels, kernel_size, stride,
+ padding, dilation, groups, bias)
+
+ def forward(self, x):
+ weight = self.weight
+ weight_mean = weight.mean(dim=1, keepdim=True).mean(dim=2,
+ keepdim=True).mean(dim=3, keepdim=True)
+ weight = weight - weight_mean
+ std = weight.view(weight.size(0), -1).std(dim=1).view(-1, 1, 1, 1) + 1e-5
+ weight = weight / std.expand_as(weight)
+ return F.conv2d(x, weight, self.bias, self.stride,
+ self.padding, self.dilation, self.groups)
+
+
+# normalize
+def norm_normalize(norm_out):
+ min_kappa = 0.01
+ norm_x, norm_y, norm_z, kappa = torch.split(norm_out, 1, dim=1)
+ norm = torch.sqrt(norm_x ** 2.0 + norm_y ** 2.0 + norm_z ** 2.0) + 1e-10
+ kappa = F.elu(kappa) + 1.0 + min_kappa
+ final_out = torch.cat([norm_x / norm, norm_y / norm, norm_z / norm, kappa], dim=1)
+ return final_out
+
+
+# uncertainty-guided sampling (only used during training)
+@torch.no_grad()
+def sample_points(init_normal, gt_norm_mask, sampling_ratio, beta):
+ device = init_normal.device
+ B, _, H, W = init_normal.shape
+ N = int(sampling_ratio * H * W)
+ beta = beta
+
+ # uncertainty map
+ uncertainty_map = -1 * init_normal[:, 3, :, :] # B, H, W
+
+ # gt_invalid_mask (B, H, W)
+ if gt_norm_mask is not None:
+ gt_invalid_mask = F.interpolate(gt_norm_mask.float(), size=[H, W], mode='nearest')
+ gt_invalid_mask = gt_invalid_mask[:, 0, :, :] < 0.5
+ uncertainty_map[gt_invalid_mask] = -1e4
+
+ # (B, H*W)
+ _, idx = uncertainty_map.view(B, -1).sort(1, descending=True)
+
+ # importance sampling
+ if int(beta * N) > 0:
+ importance = idx[:, :int(beta * N)] # B, beta*N
+
+ # remaining
+ remaining = idx[:, int(beta * N):] # B, H*W - beta*N
+
+ # coverage
+ num_coverage = N - int(beta * N)
+
+ if num_coverage <= 0:
+ samples = importance
+ else:
+ coverage_list = []
+ for i in range(B):
+ idx_c = torch.randperm(remaining.size()[1]) # shuffles "H*W - beta*N"
+ coverage_list.append(remaining[i, :][idx_c[:num_coverage]].view(1, -1)) # 1, N-beta*N
+ coverage = torch.cat(coverage_list, dim=0) # B, N-beta*N
+ samples = torch.cat((importance, coverage), dim=1) # B, N
+
+ else:
+ # remaining
+ remaining = idx[:, :] # B, H*W
+
+ # coverage
+ num_coverage = N
+
+ coverage_list = []
+ for i in range(B):
+ idx_c = torch.randperm(remaining.size()[1]) # shuffles "H*W - beta*N"
+ coverage_list.append(remaining[i, :][idx_c[:num_coverage]].view(1, -1)) # 1, N-beta*N
+ coverage = torch.cat(coverage_list, dim=0) # B, N-beta*N
+ samples = coverage
+
+ # point coordinates
+ rows_int = samples // W # 0 for first row, H-1 for last row
+ rows_float = rows_int / float(H-1) # 0 to 1.0
+ rows_float = (rows_float * 2.0) - 1.0 # -1.0 to 1.0
+
+ cols_int = samples % W # 0 for first column, W-1 for last column
+ cols_float = cols_int / float(W-1) # 0 to 1.0
+ cols_float = (cols_float * 2.0) - 1.0 # -1.0 to 1.0
+
+ point_coords = torch.zeros(B, 1, N, 2)
+ point_coords[:, 0, :, 0] = cols_float # x coord
+ point_coords[:, 0, :, 1] = rows_float # y coord
+ point_coords = point_coords.to(device)
+ return point_coords, rows_int, cols_int
\ No newline at end of file
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/utils/losses.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/utils/losses.py
new file mode 100644
index 0000000000000000000000000000000000000000..f47f610f94b6c14c260433bb0cb94c24820f132e
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/utils/losses.py
@@ -0,0 +1,178 @@
+import torch
+import torch.nn as nn
+import numpy as np
+import torch.nn.functional as F
+
+
+# compute loss
+class compute_loss(nn.Module):
+ def __init__(self, args):
+ """args.loss_fn can be one of following:
+ - L1 - L1 loss (no uncertainty)
+ - L2 - L2 loss (no uncertainty)
+ - AL - Angular loss (no uncertainty)
+ - NLL_vMF - NLL of vonMF distribution
+ - NLL_ours - NLL of Angular vonMF distribution
+ - UG_NLL_vMF - NLL of vonMF distribution (+ pixel-wise MLP + uncertainty-guided sampling)
+ - UG_NLL_ours - NLL of Angular vonMF distribution (+ pixel-wise MLP + uncertainty-guided sampling)
+ """
+ super(compute_loss, self).__init__()
+ self.loss_type = args.loss_fn
+ if self.loss_type in ['L1', 'L2', 'AL', 'NLL_vMF', 'NLL_ours']:
+ self.loss_fn = self.forward_R
+ elif self.loss_type in ['UG_NLL_vMF', 'UG_NLL_ours']:
+ self.loss_fn = self.forward_UG
+ else:
+ raise Exception('invalid loss type')
+
+ def forward(self, *args):
+ return self.loss_fn(*args)
+
+ def forward_R(self, norm_out, gt_norm, gt_norm_mask):
+ pred_norm, pred_kappa = norm_out[:, 0:3, :, :], norm_out[:, 3:, :, :]
+
+ if self.loss_type == 'L1':
+ l1 = torch.sum(torch.abs(gt_norm - pred_norm), dim=1, keepdim=True)
+ loss = torch.mean(l1[gt_norm_mask])
+
+ elif self.loss_type == 'L2':
+ l2 = torch.sum(torch.square(gt_norm - pred_norm), dim=1, keepdim=True)
+ loss = torch.mean(l2[gt_norm_mask])
+
+ elif self.loss_type == 'AL':
+ dot = torch.cosine_similarity(pred_norm, gt_norm, dim=1)
+
+ valid_mask = gt_norm_mask[:, 0, :, :].float() \
+ * (dot.detach() < 0.999).float() \
+ * (dot.detach() > -0.999).float()
+ valid_mask = valid_mask > 0.0
+
+ al = torch.acos(dot[valid_mask])
+ loss = torch.mean(al)
+
+ elif self.loss_type == 'NLL_vMF':
+ dot = torch.cosine_similarity(pred_norm, gt_norm, dim=1)
+
+ valid_mask = gt_norm_mask[:, 0, :, :].float() \
+ * (dot.detach() < 0.999).float() \
+ * (dot.detach() > -0.999).float()
+ valid_mask = valid_mask > 0.0
+
+ dot = dot[valid_mask]
+ kappa = pred_kappa[:, 0, :, :][valid_mask]
+
+ loss_pixelwise = - torch.log(kappa) \
+ - (kappa * (dot - 1)) \
+ + torch.log(1 - torch.exp(- 2 * kappa))
+ loss = torch.mean(loss_pixelwise)
+
+ elif self.loss_type == 'NLL_ours':
+ dot = torch.cosine_similarity(pred_norm, gt_norm, dim=1)
+
+ valid_mask = gt_norm_mask[:, 0, :, :].float() \
+ * (dot.detach() < 0.999).float() \
+ * (dot.detach() > -0.999).float()
+ valid_mask = valid_mask > 0.0
+
+ dot = dot[valid_mask]
+ kappa = pred_kappa[:, 0, :, :][valid_mask]
+
+ loss_pixelwise = - torch.log(torch.square(kappa) + 1) \
+ + kappa * torch.acos(dot) \
+ + torch.log(1 + torch.exp(-kappa * np.pi))
+ loss = torch.mean(loss_pixelwise)
+
+ else:
+ raise Exception('invalid loss type')
+
+ return loss
+
+
+ def forward_UG(self, pred_list, coord_list, gt_norm, gt_norm_mask):
+ loss = 0.0
+ for (pred, coord) in zip(pred_list, coord_list):
+ if coord is None:
+ pred = F.interpolate(pred, size=[gt_norm.size(2), gt_norm.size(3)], mode='bilinear', align_corners=True)
+ pred_norm, pred_kappa = pred[:, 0:3, :, :], pred[:, 3:, :, :]
+
+ if self.loss_type == 'UG_NLL_vMF':
+ dot = torch.cosine_similarity(pred_norm, gt_norm, dim=1)
+
+ valid_mask = gt_norm_mask[:, 0, :, :].float() \
+ * (dot.detach() < 0.999).float() \
+ * (dot.detach() > -0.999).float()
+ valid_mask = valid_mask > 0.5
+
+ # mask
+ dot = dot[valid_mask]
+ kappa = pred_kappa[:, 0, :, :][valid_mask]
+
+ loss_pixelwise = - torch.log(kappa) \
+ - (kappa * (dot - 1)) \
+ + torch.log(1 - torch.exp(- 2 * kappa))
+ loss = loss + torch.mean(loss_pixelwise)
+
+ elif self.loss_type == 'UG_NLL_ours':
+ dot = torch.cosine_similarity(pred_norm, gt_norm, dim=1)
+
+ valid_mask = gt_norm_mask[:, 0, :, :].float() \
+ * (dot.detach() < 0.999).float() \
+ * (dot.detach() > -0.999).float()
+ valid_mask = valid_mask > 0.5
+
+ dot = dot[valid_mask]
+ kappa = pred_kappa[:, 0, :, :][valid_mask]
+
+ loss_pixelwise = - torch.log(torch.square(kappa) + 1) \
+ + kappa * torch.acos(dot) \
+ + torch.log(1 + torch.exp(-kappa * np.pi))
+ loss = loss + torch.mean(loss_pixelwise)
+
+ else:
+ raise Exception
+
+ else:
+ # coord: B, 1, N, 2
+ # pred: B, 4, N
+ gt_norm_ = F.grid_sample(gt_norm, coord, mode='nearest', align_corners=True) # (B, 3, 1, N)
+ gt_norm_mask_ = F.grid_sample(gt_norm_mask.float(), coord, mode='nearest', align_corners=True) # (B, 1, 1, N)
+ gt_norm_ = gt_norm_[:, :, 0, :] # (B, 3, N)
+ gt_norm_mask_ = gt_norm_mask_[:, :, 0, :] > 0.5 # (B, 1, N)
+
+ pred_norm, pred_kappa = pred[:, 0:3, :], pred[:, 3:, :]
+
+ if self.loss_type == 'UG_NLL_vMF':
+ dot = torch.cosine_similarity(pred_norm, gt_norm_, dim=1) # (B, N)
+
+ valid_mask = gt_norm_mask_[:, 0, :].float() \
+ * (dot.detach() < 0.999).float() \
+ * (dot.detach() > -0.999).float()
+ valid_mask = valid_mask > 0.5
+
+ dot = dot[valid_mask]
+ kappa = pred_kappa[:, 0, :][valid_mask]
+
+ loss_pixelwise = - torch.log(kappa) \
+ - (kappa * (dot - 1)) \
+ + torch.log(1 - torch.exp(- 2 * kappa))
+ loss = loss + torch.mean(loss_pixelwise)
+
+ elif self.loss_type == 'UG_NLL_ours':
+ dot = torch.cosine_similarity(pred_norm, gt_norm_, dim=1) # (B, N)
+
+ valid_mask = gt_norm_mask_[:, 0, :].float() \
+ * (dot.detach() < 0.999).float() \
+ * (dot.detach() > -0.999).float()
+ valid_mask = valid_mask > 0.5
+
+ dot = dot[valid_mask]
+ kappa = pred_kappa[:, 0, :][valid_mask]
+
+ loss_pixelwise = - torch.log(torch.square(kappa) + 1) \
+ + kappa * torch.acos(dot) \
+ + torch.log(1 + torch.exp(-kappa * np.pi))
+ loss = loss + torch.mean(loss_pixelwise)
+
+ else:
+ raise Exception
+ return loss
diff --git a/ControlNet-v1-1-nightly-main/annotator/normalbae/utils/utils.py b/ControlNet-v1-1-nightly-main/annotator/normalbae/utils/utils.py
new file mode 100644
index 0000000000000000000000000000000000000000..ca7dde89929c1f9f696cc93e995af9a667cf86c8
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/normalbae/utils/utils.py
@@ -0,0 +1,191 @@
+import os
+import numpy as np
+from PIL import Image
+
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+
+import matplotlib
+matplotlib.use('Agg')
+import matplotlib.pyplot as plt
+
+
+
+# convert arg line to args
+def convert_arg_line_to_args(arg_line):
+ for arg in arg_line.split():
+ if not arg.strip():
+ continue
+ yield str(arg)
+
+
+# save args
+def save_args(args, filename):
+ with open(filename, 'w') as f:
+ for arg in vars(args):
+ f.write('{}: {}\n'.format(arg, getattr(args, arg)))
+
+
+# concatenate images
+def concat_image(image_path_list, concat_image_path):
+ imgs = [Image.open(i).convert("RGB").resize((640, 480), resample=Image.BILINEAR) for i in image_path_list]
+ imgs_list = []
+ for i in range(len(imgs)):
+ img = imgs[i]
+ imgs_list.append(np.asarray(img))
+
+ H, W, _ = np.asarray(img).shape
+ imgs_list.append(255 * np.ones((H, 20, 3)).astype('uint8'))
+
+ imgs_comb = np.hstack(imgs_list[:-1])
+ imgs_comb = Image.fromarray(imgs_comb)
+ imgs_comb.save(concat_image_path)
+
+
+# load model
+def load_checkpoint(fpath, model):
+ ckpt = torch.load(fpath, map_location='cpu')['model']
+
+ load_dict = {}
+ for k, v in ckpt.items():
+ if k.startswith('module.'):
+ k_ = k.replace('module.', '')
+ load_dict[k_] = v
+ else:
+ load_dict[k] = v
+
+ model.load_state_dict(load_dict)
+ return model
+
+
+# compute normal errors
+def compute_normal_errors(total_normal_errors):
+ metrics = {
+ 'mean': np.average(total_normal_errors),
+ 'median': np.median(total_normal_errors),
+ 'rmse': np.sqrt(np.sum(total_normal_errors * total_normal_errors) / total_normal_errors.shape),
+ 'a1': 100.0 * (np.sum(total_normal_errors < 5) / total_normal_errors.shape[0]),
+ 'a2': 100.0 * (np.sum(total_normal_errors < 7.5) / total_normal_errors.shape[0]),
+ 'a3': 100.0 * (np.sum(total_normal_errors < 11.25) / total_normal_errors.shape[0]),
+ 'a4': 100.0 * (np.sum(total_normal_errors < 22.5) / total_normal_errors.shape[0]),
+ 'a5': 100.0 * (np.sum(total_normal_errors < 30) / total_normal_errors.shape[0])
+ }
+ return metrics
+
+
+# log normal errors
+def log_normal_errors(metrics, where_to_write, first_line):
+ print(first_line)
+ print("mean median rmse 5 7.5 11.25 22.5 30")
+ print("%.3f %.3f %.3f %.3f %.3f %.3f %.3f %.3f" % (
+ metrics['mean'], metrics['median'], metrics['rmse'],
+ metrics['a1'], metrics['a2'], metrics['a3'], metrics['a4'], metrics['a5']))
+
+ with open(where_to_write, 'a') as f:
+ f.write('%s\n' % first_line)
+ f.write("mean median rmse 5 7.5 11.25 22.5 30\n")
+ f.write("%.3f %.3f %.3f %.3f %.3f %.3f %.3f %.3f\n\n" % (
+ metrics['mean'], metrics['median'], metrics['rmse'],
+ metrics['a1'], metrics['a2'], metrics['a3'], metrics['a4'], metrics['a5']))
+
+
+# makedir
+def makedir(dirpath):
+ if not os.path.exists(dirpath):
+ os.makedirs(dirpath)
+
+
+# makedir from list
+def make_dir_from_list(dirpath_list):
+ for dirpath in dirpath_list:
+ makedir(dirpath)
+
+
+
+########################################################################################################################
+# Visualization
+########################################################################################################################
+
+
+# unnormalize image
+__imagenet_stats = {'mean': [0.485, 0.456, 0.406], 'std': [0.229, 0.224, 0.225]}
+def unnormalize(img_in):
+ img_out = np.zeros(img_in.shape)
+ for ich in range(3):
+ img_out[:, :, ich] = img_in[:, :, ich] * __imagenet_stats['std'][ich]
+ img_out[:, :, ich] += __imagenet_stats['mean'][ich]
+ img_out = (img_out * 255).astype(np.uint8)
+ return img_out
+
+
+# kappa to exp error (only applicable to AngMF distribution)
+def kappa_to_alpha(pred_kappa):
+ alpha = ((2 * pred_kappa) / ((pred_kappa ** 2.0) + 1)) \
+ + ((np.exp(- pred_kappa * np.pi) * np.pi) / (1 + np.exp(- pred_kappa * np.pi)))
+ alpha = np.degrees(alpha)
+ return alpha
+
+
+# normal vector to rgb values
+def norm_to_rgb(norm):
+ # norm: (B, H, W, 3)
+ norm_rgb = ((norm[0, ...] + 1) * 0.5) * 255
+ norm_rgb = np.clip(norm_rgb, a_min=0, a_max=255)
+ norm_rgb = norm_rgb.astype(np.uint8)
+ return norm_rgb
+
+
+# visualize during training
+def visualize(args, img, gt_norm, gt_norm_mask, norm_out_list, total_iter):
+ B, _, H, W = gt_norm.shape
+
+ pred_norm_list = []
+ pred_kappa_list = []
+ for norm_out in norm_out_list:
+ norm_out = F.interpolate(norm_out, size=[gt_norm.size(2), gt_norm.size(3)], mode='nearest')
+ pred_norm = norm_out[:, :3, :, :] # (B, 3, H, W)
+ pred_norm = pred_norm.detach().cpu().permute(0, 2, 3, 1).numpy() # (B, H, W, 3)
+ pred_norm_list.append(pred_norm)
+
+ pred_kappa = norm_out[:, 3:, :, :] # (B, 1, H, W)
+ pred_kappa = pred_kappa.detach().cpu().permute(0, 2, 3, 1).numpy() # (B, H, W, 1)
+ pred_kappa_list.append(pred_kappa)
+
+ # to numpy arrays
+ img = img.detach().cpu().permute(0, 2, 3, 1).numpy() # (B, H, W, 3)
+ gt_norm = gt_norm.detach().cpu().permute(0, 2, 3, 1).numpy() # (B, H, W, 3)
+ gt_norm_mask = gt_norm_mask.detach().cpu().permute(0, 2, 3, 1).numpy() # (B, H, W, 1)
+
+ # input image
+ target_path = '%s/%08d_img.jpg' % (args.exp_vis_dir, total_iter)
+ img = unnormalize(img[0, ...])
+ plt.imsave(target_path, img)
+
+ # gt norm
+ gt_norm_rgb = ((gt_norm[0, ...] + 1) * 0.5) * 255
+ gt_norm_rgb = np.clip(gt_norm_rgb, a_min=0, a_max=255)
+ gt_norm_rgb = gt_norm_rgb.astype(np.uint8)
+
+ target_path = '%s/%08d_gt_norm.jpg' % (args.exp_vis_dir, total_iter)
+ plt.imsave(target_path, gt_norm_rgb * gt_norm_mask[0, ...])
+
+ # pred_norm
+ for i in range(len(pred_norm_list)):
+ pred_norm = pred_norm_list[i]
+ pred_norm_rgb = norm_to_rgb(pred_norm)
+ target_path = '%s/%08d_pred_norm_%d.jpg' % (args.exp_vis_dir, total_iter, i)
+ plt.imsave(target_path, pred_norm_rgb)
+
+ pred_kappa = pred_kappa_list[i]
+ pred_alpha = kappa_to_alpha(pred_kappa)
+ target_path = '%s/%08d_pred_alpha_%d.jpg' % (args.exp_vis_dir, total_iter, i)
+ plt.imsave(target_path, pred_alpha[0, :, :, 0], vmin=0, vmax=60, cmap='jet')
+
+ # error in angles
+ DP = np.sum(gt_norm * pred_norm, axis=3, keepdims=True) # (B, H, W, 1)
+ DP = np.clip(DP, -1, 1)
+ E = np.degrees(np.arccos(DP)) # (B, H, W, 1)
+ E = E * gt_norm_mask
+ target_path = '%s/%08d_pred_error_%d.jpg' % (args.exp_vis_dir, total_iter, i)
+ plt.imsave(target_path, E[0, :, :, 0], vmin=0, vmax=60, cmap='jet')
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/LICENSE b/ControlNet-v1-1-nightly-main/annotator/oneformer/LICENSE
new file mode 100644
index 0000000000000000000000000000000000000000..16a9d56a3d4c15e4f34ac5426459c58487b01520
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/LICENSE
@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2022 Caroline Chan
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
\ No newline at end of file
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/__init__.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..57ff0c6581e552df750abe5bb92ed4f39a7dfa46
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/__init__.py
@@ -0,0 +1,34 @@
+# https://github.com/SHI-Labs/OneFormer
+
+import os
+from annotator.util import annotator_ckpts_path
+from .api import make_detectron2_model, semantic_run
+
+
+class OneformerCOCODetector:
+ def __init__(self):
+ remote_model_path = "https://huggingface.co/lllyasviel/Annotators/resolve/main/150_16_swin_l_oneformer_coco_100ep.pth"
+ modelpath = os.path.join(annotator_ckpts_path, "150_16_swin_l_oneformer_coco_100ep.pth")
+ if not os.path.exists(modelpath):
+ from basicsr.utils.download_util import load_file_from_url
+ load_file_from_url(remote_model_path, model_dir=annotator_ckpts_path)
+ config = os.path.join(os.path.dirname(__file__), 'configs/coco/oneformer_swin_large_IN21k_384_bs16_100ep.yaml')
+ self.model, self.meta = make_detectron2_model(config, modelpath)
+
+ def __call__(self, img):
+ return semantic_run(img, self.model, self.meta)
+
+
+class OneformerADE20kDetector:
+ def __init__(self):
+ remote_model_path = "https://huggingface.co/lllyasviel/Annotators/resolve/main/250_16_swin_l_oneformer_ade20k_160k.pth"
+ modelpath = os.path.join(annotator_ckpts_path, "250_16_swin_l_oneformer_ade20k_160k.pth")
+ if not os.path.exists(modelpath):
+ from basicsr.utils.download_util import load_file_from_url
+ load_file_from_url(remote_model_path, model_dir=annotator_ckpts_path)
+ config = os.path.join(os.path.dirname(__file__), 'configs/ade20k/oneformer_swin_large_IN21k_384_bs16_160k.yaml')
+ self.model, self.meta = make_detectron2_model(config, modelpath)
+
+ def __call__(self, img):
+ return semantic_run(img, self.model, self.meta)
+
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/api.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/api.py
new file mode 100644
index 0000000000000000000000000000000000000000..e96d7d6e32d7e52fae776792d810a19dfee18015
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/api.py
@@ -0,0 +1,43 @@
+import os
+os.environ["KMP_DUPLICATE_LIB_OK"] = "TRUE"
+
+import torch
+
+from annotator.oneformer.detectron2.config import get_cfg
+from annotator.oneformer.detectron2.projects.deeplab import add_deeplab_config
+from annotator.oneformer.detectron2.data import MetadataCatalog
+
+from annotator.oneformer.oneformer import (
+ add_oneformer_config,
+ add_common_config,
+ add_swin_config,
+ add_dinat_config,
+)
+
+from annotator.oneformer.oneformer.demo.defaults import DefaultPredictor
+from annotator.oneformer.oneformer.demo.visualizer import Visualizer, ColorMode
+
+
+def make_detectron2_model(config_path, ckpt_path):
+ cfg = get_cfg()
+ add_deeplab_config(cfg)
+ add_common_config(cfg)
+ add_swin_config(cfg)
+ add_oneformer_config(cfg)
+ add_dinat_config(cfg)
+ cfg.merge_from_file(config_path)
+ if torch.cuda.is_available():
+ cfg.MODEL.DEVICE = 'cuda'
+ else:
+ cfg.MODEL.DEVICE = 'cpu'
+ cfg.MODEL.WEIGHTS = ckpt_path
+ cfg.freeze()
+ metadata = MetadataCatalog.get(cfg.DATASETS.TEST_PANOPTIC[0] if len(cfg.DATASETS.TEST_PANOPTIC) else "__unused")
+ return DefaultPredictor(cfg), metadata
+
+
+def semantic_run(img, predictor, metadata):
+ predictions = predictor(img[:, :, ::-1], "semantic") # Predictor of OneFormer must use BGR image !!!
+ visualizer_map = Visualizer(img, is_img=False, metadata=metadata, instance_mode=ColorMode.IMAGE)
+ out_map = visualizer_map.draw_sem_seg(predictions["sem_seg"].argmax(dim=0).cpu(), alpha=1, is_text=False).get_image()
+ return out_map
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/configs/ade20k/Base-ADE20K-UnifiedSegmentation.yaml b/ControlNet-v1-1-nightly-main/annotator/oneformer/configs/ade20k/Base-ADE20K-UnifiedSegmentation.yaml
new file mode 100644
index 0000000000000000000000000000000000000000..31eab45b878433fc844a13dbdd54f97c936d9b89
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/configs/ade20k/Base-ADE20K-UnifiedSegmentation.yaml
@@ -0,0 +1,68 @@
+MODEL:
+ BACKBONE:
+ FREEZE_AT: 0
+ NAME: "build_resnet_backbone"
+ WEIGHTS: "detectron2://ImageNetPretrained/torchvision/R-50.pkl"
+ PIXEL_MEAN: [123.675, 116.280, 103.530]
+ PIXEL_STD: [58.395, 57.120, 57.375]
+ RESNETS:
+ DEPTH: 50
+ STEM_TYPE: "basic" # not used
+ STEM_OUT_CHANNELS: 64
+ STRIDE_IN_1X1: False
+ OUT_FEATURES: ["res2", "res3", "res4", "res5"]
+ # NORM: "SyncBN"
+ RES5_MULTI_GRID: [1, 1, 1] # not used
+DATASETS:
+ TRAIN: ("ade20k_panoptic_train",)
+ TEST_PANOPTIC: ("ade20k_panoptic_val",)
+ TEST_INSTANCE: ("ade20k_instance_val",)
+ TEST_SEMANTIC: ("ade20k_sem_seg_val",)
+SOLVER:
+ IMS_PER_BATCH: 16
+ BASE_LR: 0.0001
+ MAX_ITER: 160000
+ WARMUP_FACTOR: 1.0
+ WARMUP_ITERS: 0
+ WEIGHT_DECAY: 0.05
+ OPTIMIZER: "ADAMW"
+ LR_SCHEDULER_NAME: "WarmupPolyLR"
+ BACKBONE_MULTIPLIER: 0.1
+ CLIP_GRADIENTS:
+ ENABLED: True
+ CLIP_TYPE: "full_model"
+ CLIP_VALUE: 0.01
+ NORM_TYPE: 2.0
+ AMP:
+ ENABLED: True
+INPUT:
+ MIN_SIZE_TRAIN: !!python/object/apply:eval ["[int(x * 0.1 * 512) for x in range(5, 21)]"]
+ MIN_SIZE_TRAIN_SAMPLING: "choice"
+ MIN_SIZE_TEST: 512
+ MAX_SIZE_TRAIN: 2048
+ MAX_SIZE_TEST: 2048
+ CROP:
+ ENABLED: True
+ TYPE: "absolute"
+ SIZE: (512, 512)
+ SINGLE_CATEGORY_MAX_AREA: 1.0
+ COLOR_AUG_SSD: True
+ SIZE_DIVISIBILITY: 512 # used in dataset mapper
+ FORMAT: "RGB"
+ DATASET_MAPPER_NAME: "oneformer_unified"
+ MAX_SEQ_LEN: 77
+ TASK_SEQ_LEN: 77
+ TASK_PROB:
+ SEMANTIC: 0.33
+ INSTANCE: 0.66
+TEST:
+ EVAL_PERIOD: 5000
+ AUG:
+ ENABLED: False
+ MIN_SIZES: [256, 384, 512, 640, 768, 896]
+ MAX_SIZE: 3584
+ FLIP: True
+DATALOADER:
+ FILTER_EMPTY_ANNOTATIONS: True
+ NUM_WORKERS: 4
+VERSION: 2
\ No newline at end of file
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/configs/ade20k/oneformer_R50_bs16_160k.yaml b/ControlNet-v1-1-nightly-main/annotator/oneformer/configs/ade20k/oneformer_R50_bs16_160k.yaml
new file mode 100644
index 0000000000000000000000000000000000000000..770ffc81907f8d7c7520e079b1c46060707254b8
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/configs/ade20k/oneformer_R50_bs16_160k.yaml
@@ -0,0 +1,58 @@
+_BASE_: Base-ADE20K-UnifiedSegmentation.yaml
+MODEL:
+ META_ARCHITECTURE: "OneFormer"
+ SEM_SEG_HEAD:
+ NAME: "OneFormerHead"
+ IGNORE_VALUE: 255
+ NUM_CLASSES: 150
+ LOSS_WEIGHT: 1.0
+ CONVS_DIM: 256
+ MASK_DIM: 256
+ NORM: "GN"
+ # pixel decoder
+ PIXEL_DECODER_NAME: "MSDeformAttnPixelDecoder"
+ IN_FEATURES: ["res2", "res3", "res4", "res5"]
+ DEFORMABLE_TRANSFORMER_ENCODER_IN_FEATURES: ["res3", "res4", "res5"]
+ COMMON_STRIDE: 4
+ TRANSFORMER_ENC_LAYERS: 6
+ ONE_FORMER:
+ TRANSFORMER_DECODER_NAME: "ContrastiveMultiScaleMaskedTransformerDecoder"
+ TRANSFORMER_IN_FEATURE: "multi_scale_pixel_decoder"
+ DEEP_SUPERVISION: True
+ NO_OBJECT_WEIGHT: 0.1
+ CLASS_WEIGHT: 2.0
+ MASK_WEIGHT: 5.0
+ DICE_WEIGHT: 5.0
+ CONTRASTIVE_WEIGHT: 0.5
+ CONTRASTIVE_TEMPERATURE: 0.07
+ HIDDEN_DIM: 256
+ NUM_OBJECT_QUERIES: 150
+ USE_TASK_NORM: True
+ NHEADS: 8
+ DROPOUT: 0.1
+ DIM_FEEDFORWARD: 2048
+ ENC_LAYERS: 0
+ PRE_NORM: False
+ ENFORCE_INPUT_PROJ: False
+ SIZE_DIVISIBILITY: 32
+ CLASS_DEC_LAYERS: 2
+ DEC_LAYERS: 10 # 9 decoder layers, add one for the loss on learnable query
+ TRAIN_NUM_POINTS: 12544
+ OVERSAMPLE_RATIO: 3.0
+ IMPORTANCE_SAMPLE_RATIO: 0.75
+ TEXT_ENCODER:
+ WIDTH: 256
+ CONTEXT_LENGTH: 77
+ NUM_LAYERS: 6
+ VOCAB_SIZE: 49408
+ PROJ_NUM_LAYERS: 2
+ N_CTX: 16
+ TEST:
+ SEMANTIC_ON: True
+ INSTANCE_ON: True
+ PANOPTIC_ON: True
+ OVERLAP_THRESHOLD: 0.8
+ OBJECT_MASK_THRESHOLD: 0.8
+ TASK: "panoptic"
+TEST:
+ DETECTIONS_PER_IMAGE: 150
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/configs/ade20k/oneformer_swin_large_IN21k_384_bs16_160k.yaml b/ControlNet-v1-1-nightly-main/annotator/oneformer/configs/ade20k/oneformer_swin_large_IN21k_384_bs16_160k.yaml
new file mode 100644
index 0000000000000000000000000000000000000000..69c44ade144e4504077c0fe04fa8bb3491a679ed
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/configs/ade20k/oneformer_swin_large_IN21k_384_bs16_160k.yaml
@@ -0,0 +1,40 @@
+_BASE_: oneformer_R50_bs16_160k.yaml
+MODEL:
+ BACKBONE:
+ NAME: "D2SwinTransformer"
+ SWIN:
+ EMBED_DIM: 192
+ DEPTHS: [2, 2, 18, 2]
+ NUM_HEADS: [6, 12, 24, 48]
+ WINDOW_SIZE: 12
+ APE: False
+ DROP_PATH_RATE: 0.3
+ PATCH_NORM: True
+ PRETRAIN_IMG_SIZE: 384
+ WEIGHTS: "swin_large_patch4_window12_384_22k.pkl"
+ PIXEL_MEAN: [123.675, 116.280, 103.530]
+ PIXEL_STD: [58.395, 57.120, 57.375]
+ ONE_FORMER:
+ NUM_OBJECT_QUERIES: 250
+INPUT:
+ MIN_SIZE_TRAIN: !!python/object/apply:eval ["[int(x * 0.1 * 640) for x in range(5, 21)]"]
+ MIN_SIZE_TRAIN_SAMPLING: "choice"
+ MIN_SIZE_TEST: 640
+ MAX_SIZE_TRAIN: 2560
+ MAX_SIZE_TEST: 2560
+ CROP:
+ ENABLED: True
+ TYPE: "absolute"
+ SIZE: (640, 640)
+ SINGLE_CATEGORY_MAX_AREA: 1.0
+ COLOR_AUG_SSD: True
+ SIZE_DIVISIBILITY: 640 # used in dataset mapper
+ FORMAT: "RGB"
+TEST:
+ DETECTIONS_PER_IMAGE: 250
+ EVAL_PERIOD: 5000
+ AUG:
+ ENABLED: False
+ MIN_SIZES: [320, 480, 640, 800, 960, 1120]
+ MAX_SIZE: 4480
+ FLIP: True
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/configs/coco/Base-COCO-UnifiedSegmentation.yaml b/ControlNet-v1-1-nightly-main/annotator/oneformer/configs/coco/Base-COCO-UnifiedSegmentation.yaml
new file mode 100644
index 0000000000000000000000000000000000000000..ccd24f348f9bc7d60dcdc4b74d887708e57cb8a8
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/configs/coco/Base-COCO-UnifiedSegmentation.yaml
@@ -0,0 +1,54 @@
+MODEL:
+ BACKBONE:
+ FREEZE_AT: 0
+ NAME: "build_resnet_backbone"
+ WEIGHTS: "detectron2://ImageNetPretrained/torchvision/R-50.pkl"
+ PIXEL_MEAN: [123.675, 116.280, 103.530]
+ PIXEL_STD: [58.395, 57.120, 57.375]
+ RESNETS:
+ DEPTH: 50
+ STEM_TYPE: "basic" # not used
+ STEM_OUT_CHANNELS: 64
+ STRIDE_IN_1X1: False
+ OUT_FEATURES: ["res2", "res3", "res4", "res5"]
+ # NORM: "SyncBN"
+ RES5_MULTI_GRID: [1, 1, 1] # not used
+DATASETS:
+ TRAIN: ("coco_2017_train_panoptic_with_sem_seg",)
+ TEST_PANOPTIC: ("coco_2017_val_panoptic_with_sem_seg",) # to evaluate instance and semantic performance as well
+ TEST_INSTANCE: ("coco_2017_val",)
+ TEST_SEMANTIC: ("coco_2017_val_panoptic_with_sem_seg",)
+SOLVER:
+ IMS_PER_BATCH: 16
+ BASE_LR: 0.0001
+ STEPS: (327778, 355092)
+ MAX_ITER: 368750
+ WARMUP_FACTOR: 1.0
+ WARMUP_ITERS: 10
+ WEIGHT_DECAY: 0.05
+ OPTIMIZER: "ADAMW"
+ BACKBONE_MULTIPLIER: 0.1
+ CLIP_GRADIENTS:
+ ENABLED: True
+ CLIP_TYPE: "full_model"
+ CLIP_VALUE: 0.01
+ NORM_TYPE: 2.0
+ AMP:
+ ENABLED: True
+INPUT:
+ IMAGE_SIZE: 1024
+ MIN_SCALE: 0.1
+ MAX_SCALE: 2.0
+ FORMAT: "RGB"
+ DATASET_MAPPER_NAME: "coco_unified_lsj"
+ MAX_SEQ_LEN: 77
+ TASK_SEQ_LEN: 77
+ TASK_PROB:
+ SEMANTIC: 0.33
+ INSTANCE: 0.66
+TEST:
+ EVAL_PERIOD: 5000
+DATALOADER:
+ FILTER_EMPTY_ANNOTATIONS: True
+ NUM_WORKERS: 4
+VERSION: 2
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/configs/coco/oneformer_R50_bs16_50ep.yaml b/ControlNet-v1-1-nightly-main/annotator/oneformer/configs/coco/oneformer_R50_bs16_50ep.yaml
new file mode 100644
index 0000000000000000000000000000000000000000..f768c8fa8b5e4fc1121e65e050053e0d8870cd73
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/configs/coco/oneformer_R50_bs16_50ep.yaml
@@ -0,0 +1,59 @@
+_BASE_: Base-COCO-UnifiedSegmentation.yaml
+MODEL:
+ META_ARCHITECTURE: "OneFormer"
+ SEM_SEG_HEAD:
+ NAME: "OneFormerHead"
+ IGNORE_VALUE: 255
+ NUM_CLASSES: 133
+ LOSS_WEIGHT: 1.0
+ CONVS_DIM: 256
+ MASK_DIM: 256
+ NORM: "GN"
+ # pixel decoder
+ PIXEL_DECODER_NAME: "MSDeformAttnPixelDecoder"
+ IN_FEATURES: ["res2", "res3", "res4", "res5"]
+ DEFORMABLE_TRANSFORMER_ENCODER_IN_FEATURES: ["res3", "res4", "res5"]
+ COMMON_STRIDE: 4
+ TRANSFORMER_ENC_LAYERS: 6
+ ONE_FORMER:
+ TRANSFORMER_DECODER_NAME: "ContrastiveMultiScaleMaskedTransformerDecoder"
+ TRANSFORMER_IN_FEATURE: "multi_scale_pixel_decoder"
+ DEEP_SUPERVISION: True
+ NO_OBJECT_WEIGHT: 0.1
+ CLASS_WEIGHT: 2.0
+ MASK_WEIGHT: 5.0
+ DICE_WEIGHT: 5.0
+ CONTRASTIVE_WEIGHT: 0.5
+ CONTRASTIVE_TEMPERATURE: 0.07
+ HIDDEN_DIM: 256
+ NUM_OBJECT_QUERIES: 150
+ USE_TASK_NORM: True
+ NHEADS: 8
+ DROPOUT: 0.1
+ DIM_FEEDFORWARD: 2048
+ ENC_LAYERS: 0
+ PRE_NORM: False
+ ENFORCE_INPUT_PROJ: False
+ SIZE_DIVISIBILITY: 32
+ CLASS_DEC_LAYERS: 2
+ DEC_LAYERS: 10 # 9 decoder layers, add one for the loss on learnable query
+ TRAIN_NUM_POINTS: 12544
+ OVERSAMPLE_RATIO: 3.0
+ IMPORTANCE_SAMPLE_RATIO: 0.75
+ TEXT_ENCODER:
+ WIDTH: 256
+ CONTEXT_LENGTH: 77
+ NUM_LAYERS: 6
+ VOCAB_SIZE: 49408
+ PROJ_NUM_LAYERS: 2
+ N_CTX: 16
+ TEST:
+ SEMANTIC_ON: True
+ INSTANCE_ON: True
+ PANOPTIC_ON: True
+ DETECTION_ON: False
+ OVERLAP_THRESHOLD: 0.8
+ OBJECT_MASK_THRESHOLD: 0.8
+ TASK: "panoptic"
+TEST:
+ DETECTIONS_PER_IMAGE: 150
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/configs/coco/oneformer_swin_large_IN21k_384_bs16_100ep.yaml b/ControlNet-v1-1-nightly-main/annotator/oneformer/configs/coco/oneformer_swin_large_IN21k_384_bs16_100ep.yaml
new file mode 100644
index 0000000000000000000000000000000000000000..faae655317c52d90b9f756417f8b1a1adcbe78f2
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/configs/coco/oneformer_swin_large_IN21k_384_bs16_100ep.yaml
@@ -0,0 +1,25 @@
+_BASE_: oneformer_R50_bs16_50ep.yaml
+MODEL:
+ BACKBONE:
+ NAME: "D2SwinTransformer"
+ SWIN:
+ EMBED_DIM: 192
+ DEPTHS: [2, 2, 18, 2]
+ NUM_HEADS: [6, 12, 24, 48]
+ WINDOW_SIZE: 12
+ APE: False
+ DROP_PATH_RATE: 0.3
+ PATCH_NORM: True
+ PRETRAIN_IMG_SIZE: 384
+ WEIGHTS: "swin_large_patch4_window12_384_22k.pkl"
+ PIXEL_MEAN: [123.675, 116.280, 103.530]
+ PIXEL_STD: [58.395, 57.120, 57.375]
+ ONE_FORMER:
+ NUM_OBJECT_QUERIES: 150
+SOLVER:
+ STEPS: (655556, 735184)
+ MAX_ITER: 737500
+ AMP:
+ ENABLED: False
+TEST:
+ DETECTIONS_PER_IMAGE: 150
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/__init__.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..bdd994b49294485c27610772f97f177741f5518f
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/__init__.py
@@ -0,0 +1,10 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+
+from .utils.env import setup_environment
+
+setup_environment()
+
+
+# This line will be programatically read/write by setup.py.
+# Leave them at the bottom of this file and don't touch them.
+__version__ = "0.6"
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/checkpoint/__init__.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/checkpoint/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..99da0469ae7e169d8970e4b642fed3f870076860
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/checkpoint/__init__.py
@@ -0,0 +1,10 @@
+# -*- coding: utf-8 -*-
+# Copyright (c) Facebook, Inc. and its affiliates.
+# File:
+
+
+from . import catalog as _UNUSED # register the handler
+from .detection_checkpoint import DetectionCheckpointer
+from fvcore.common.checkpoint import Checkpointer, PeriodicCheckpointer
+
+__all__ = ["Checkpointer", "PeriodicCheckpointer", "DetectionCheckpointer"]
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/checkpoint/c2_model_loading.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/checkpoint/c2_model_loading.py
new file mode 100644
index 0000000000000000000000000000000000000000..c6de2a3c830089aa7a0d27df96bb4a45fc5a7b0d
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/checkpoint/c2_model_loading.py
@@ -0,0 +1,412 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import copy
+import logging
+import re
+from typing import Dict, List
+import torch
+from tabulate import tabulate
+
+
+def convert_basic_c2_names(original_keys):
+ """
+ Apply some basic name conversion to names in C2 weights.
+ It only deals with typical backbone models.
+
+ Args:
+ original_keys (list[str]):
+ Returns:
+ list[str]: The same number of strings matching those in original_keys.
+ """
+ layer_keys = copy.deepcopy(original_keys)
+ layer_keys = [
+ {"pred_b": "linear_b", "pred_w": "linear_w"}.get(k, k) for k in layer_keys
+ ] # some hard-coded mappings
+
+ layer_keys = [k.replace("_", ".") for k in layer_keys]
+ layer_keys = [re.sub("\\.b$", ".bias", k) for k in layer_keys]
+ layer_keys = [re.sub("\\.w$", ".weight", k) for k in layer_keys]
+ # Uniform both bn and gn names to "norm"
+ layer_keys = [re.sub("bn\\.s$", "norm.weight", k) for k in layer_keys]
+ layer_keys = [re.sub("bn\\.bias$", "norm.bias", k) for k in layer_keys]
+ layer_keys = [re.sub("bn\\.rm", "norm.running_mean", k) for k in layer_keys]
+ layer_keys = [re.sub("bn\\.running.mean$", "norm.running_mean", k) for k in layer_keys]
+ layer_keys = [re.sub("bn\\.riv$", "norm.running_var", k) for k in layer_keys]
+ layer_keys = [re.sub("bn\\.running.var$", "norm.running_var", k) for k in layer_keys]
+ layer_keys = [re.sub("bn\\.gamma$", "norm.weight", k) for k in layer_keys]
+ layer_keys = [re.sub("bn\\.beta$", "norm.bias", k) for k in layer_keys]
+ layer_keys = [re.sub("gn\\.s$", "norm.weight", k) for k in layer_keys]
+ layer_keys = [re.sub("gn\\.bias$", "norm.bias", k) for k in layer_keys]
+
+ # stem
+ layer_keys = [re.sub("^res\\.conv1\\.norm\\.", "conv1.norm.", k) for k in layer_keys]
+ # to avoid mis-matching with "conv1" in other components (e.g. detection head)
+ layer_keys = [re.sub("^conv1\\.", "stem.conv1.", k) for k in layer_keys]
+
+ # layer1-4 is used by torchvision, however we follow the C2 naming strategy (res2-5)
+ # layer_keys = [re.sub("^res2.", "layer1.", k) for k in layer_keys]
+ # layer_keys = [re.sub("^res3.", "layer2.", k) for k in layer_keys]
+ # layer_keys = [re.sub("^res4.", "layer3.", k) for k in layer_keys]
+ # layer_keys = [re.sub("^res5.", "layer4.", k) for k in layer_keys]
+
+ # blocks
+ layer_keys = [k.replace(".branch1.", ".shortcut.") for k in layer_keys]
+ layer_keys = [k.replace(".branch2a.", ".conv1.") for k in layer_keys]
+ layer_keys = [k.replace(".branch2b.", ".conv2.") for k in layer_keys]
+ layer_keys = [k.replace(".branch2c.", ".conv3.") for k in layer_keys]
+
+ # DensePose substitutions
+ layer_keys = [re.sub("^body.conv.fcn", "body_conv_fcn", k) for k in layer_keys]
+ layer_keys = [k.replace("AnnIndex.lowres", "ann_index_lowres") for k in layer_keys]
+ layer_keys = [k.replace("Index.UV.lowres", "index_uv_lowres") for k in layer_keys]
+ layer_keys = [k.replace("U.lowres", "u_lowres") for k in layer_keys]
+ layer_keys = [k.replace("V.lowres", "v_lowres") for k in layer_keys]
+ return layer_keys
+
+
+def convert_c2_detectron_names(weights):
+ """
+ Map Caffe2 Detectron weight names to Detectron2 names.
+
+ Args:
+ weights (dict): name -> tensor
+
+ Returns:
+ dict: detectron2 names -> tensor
+ dict: detectron2 names -> C2 names
+ """
+ logger = logging.getLogger(__name__)
+ logger.info("Renaming Caffe2 weights ......")
+ original_keys = sorted(weights.keys())
+ layer_keys = copy.deepcopy(original_keys)
+
+ layer_keys = convert_basic_c2_names(layer_keys)
+
+ # --------------------------------------------------------------------------
+ # RPN hidden representation conv
+ # --------------------------------------------------------------------------
+ # FPN case
+ # In the C2 model, the RPN hidden layer conv is defined for FPN level 2 and then
+ # shared for all other levels, hence the appearance of "fpn2"
+ layer_keys = [
+ k.replace("conv.rpn.fpn2", "proposal_generator.rpn_head.conv") for k in layer_keys
+ ]
+ # Non-FPN case
+ layer_keys = [k.replace("conv.rpn", "proposal_generator.rpn_head.conv") for k in layer_keys]
+
+ # --------------------------------------------------------------------------
+ # RPN box transformation conv
+ # --------------------------------------------------------------------------
+ # FPN case (see note above about "fpn2")
+ layer_keys = [
+ k.replace("rpn.bbox.pred.fpn2", "proposal_generator.rpn_head.anchor_deltas")
+ for k in layer_keys
+ ]
+ layer_keys = [
+ k.replace("rpn.cls.logits.fpn2", "proposal_generator.rpn_head.objectness_logits")
+ for k in layer_keys
+ ]
+ # Non-FPN case
+ layer_keys = [
+ k.replace("rpn.bbox.pred", "proposal_generator.rpn_head.anchor_deltas") for k in layer_keys
+ ]
+ layer_keys = [
+ k.replace("rpn.cls.logits", "proposal_generator.rpn_head.objectness_logits")
+ for k in layer_keys
+ ]
+
+ # --------------------------------------------------------------------------
+ # Fast R-CNN box head
+ # --------------------------------------------------------------------------
+ layer_keys = [re.sub("^bbox\\.pred", "bbox_pred", k) for k in layer_keys]
+ layer_keys = [re.sub("^cls\\.score", "cls_score", k) for k in layer_keys]
+ layer_keys = [re.sub("^fc6\\.", "box_head.fc1.", k) for k in layer_keys]
+ layer_keys = [re.sub("^fc7\\.", "box_head.fc2.", k) for k in layer_keys]
+ # 4conv1fc head tensor names: head_conv1_w, head_conv1_gn_s
+ layer_keys = [re.sub("^head\\.conv", "box_head.conv", k) for k in layer_keys]
+
+ # --------------------------------------------------------------------------
+ # FPN lateral and output convolutions
+ # --------------------------------------------------------------------------
+ def fpn_map(name):
+ """
+ Look for keys with the following patterns:
+ 1) Starts with "fpn.inner."
+ Example: "fpn.inner.res2.2.sum.lateral.weight"
+ Meaning: These are lateral pathway convolutions
+ 2) Starts with "fpn.res"
+ Example: "fpn.res2.2.sum.weight"
+ Meaning: These are FPN output convolutions
+ """
+ splits = name.split(".")
+ norm = ".norm" if "norm" in splits else ""
+ if name.startswith("fpn.inner."):
+ # splits example: ['fpn', 'inner', 'res2', '2', 'sum', 'lateral', 'weight']
+ stage = int(splits[2][len("res") :])
+ return "fpn_lateral{}{}.{}".format(stage, norm, splits[-1])
+ elif name.startswith("fpn.res"):
+ # splits example: ['fpn', 'res2', '2', 'sum', 'weight']
+ stage = int(splits[1][len("res") :])
+ return "fpn_output{}{}.{}".format(stage, norm, splits[-1])
+ return name
+
+ layer_keys = [fpn_map(k) for k in layer_keys]
+
+ # --------------------------------------------------------------------------
+ # Mask R-CNN mask head
+ # --------------------------------------------------------------------------
+ # roi_heads.StandardROIHeads case
+ layer_keys = [k.replace(".[mask].fcn", "mask_head.mask_fcn") for k in layer_keys]
+ layer_keys = [re.sub("^\\.mask\\.fcn", "mask_head.mask_fcn", k) for k in layer_keys]
+ layer_keys = [k.replace("mask.fcn.logits", "mask_head.predictor") for k in layer_keys]
+ # roi_heads.Res5ROIHeads case
+ layer_keys = [k.replace("conv5.mask", "mask_head.deconv") for k in layer_keys]
+
+ # --------------------------------------------------------------------------
+ # Keypoint R-CNN head
+ # --------------------------------------------------------------------------
+ # interestingly, the keypoint head convs have blob names that are simply "conv_fcnX"
+ layer_keys = [k.replace("conv.fcn", "roi_heads.keypoint_head.conv_fcn") for k in layer_keys]
+ layer_keys = [
+ k.replace("kps.score.lowres", "roi_heads.keypoint_head.score_lowres") for k in layer_keys
+ ]
+ layer_keys = [k.replace("kps.score.", "roi_heads.keypoint_head.score.") for k in layer_keys]
+
+ # --------------------------------------------------------------------------
+ # Done with replacements
+ # --------------------------------------------------------------------------
+ assert len(set(layer_keys)) == len(layer_keys)
+ assert len(original_keys) == len(layer_keys)
+
+ new_weights = {}
+ new_keys_to_original_keys = {}
+ for orig, renamed in zip(original_keys, layer_keys):
+ new_keys_to_original_keys[renamed] = orig
+ if renamed.startswith("bbox_pred.") or renamed.startswith("mask_head.predictor."):
+ # remove the meaningless prediction weight for background class
+ new_start_idx = 4 if renamed.startswith("bbox_pred.") else 1
+ new_weights[renamed] = weights[orig][new_start_idx:]
+ logger.info(
+ "Remove prediction weight for background class in {}. The shape changes from "
+ "{} to {}.".format(
+ renamed, tuple(weights[orig].shape), tuple(new_weights[renamed].shape)
+ )
+ )
+ elif renamed.startswith("cls_score."):
+ # move weights of bg class from original index 0 to last index
+ logger.info(
+ "Move classification weights for background class in {} from index 0 to "
+ "index {}.".format(renamed, weights[orig].shape[0] - 1)
+ )
+ new_weights[renamed] = torch.cat([weights[orig][1:], weights[orig][:1]])
+ else:
+ new_weights[renamed] = weights[orig]
+
+ return new_weights, new_keys_to_original_keys
+
+
+# Note the current matching is not symmetric.
+# it assumes model_state_dict will have longer names.
+def align_and_update_state_dicts(model_state_dict, ckpt_state_dict, c2_conversion=True):
+ """
+ Match names between the two state-dict, and returns a new chkpt_state_dict with names
+ converted to match model_state_dict with heuristics. The returned dict can be later
+ loaded with fvcore checkpointer.
+ If `c2_conversion==True`, `ckpt_state_dict` is assumed to be a Caffe2
+ model and will be renamed at first.
+
+ Strategy: suppose that the models that we will create will have prefixes appended
+ to each of its keys, for example due to an extra level of nesting that the original
+ pre-trained weights from ImageNet won't contain. For example, model.state_dict()
+ might return backbone[0].body.res2.conv1.weight, while the pre-trained model contains
+ res2.conv1.weight. We thus want to match both parameters together.
+ For that, we look for each model weight, look among all loaded keys if there is one
+ that is a suffix of the current weight name, and use it if that's the case.
+ If multiple matches exist, take the one with longest size
+ of the corresponding name. For example, for the same model as before, the pretrained
+ weight file can contain both res2.conv1.weight, as well as conv1.weight. In this case,
+ we want to match backbone[0].body.conv1.weight to conv1.weight, and
+ backbone[0].body.res2.conv1.weight to res2.conv1.weight.
+ """
+ model_keys = sorted(model_state_dict.keys())
+ if c2_conversion:
+ ckpt_state_dict, original_keys = convert_c2_detectron_names(ckpt_state_dict)
+ # original_keys: the name in the original dict (before renaming)
+ else:
+ original_keys = {x: x for x in ckpt_state_dict.keys()}
+ ckpt_keys = sorted(ckpt_state_dict.keys())
+
+ def match(a, b):
+ # Matched ckpt_key should be a complete (starts with '.') suffix.
+ # For example, roi_heads.mesh_head.whatever_conv1 does not match conv1,
+ # but matches whatever_conv1 or mesh_head.whatever_conv1.
+ return a == b or a.endswith("." + b)
+
+ # get a matrix of string matches, where each (i, j) entry correspond to the size of the
+ # ckpt_key string, if it matches
+ match_matrix = [len(j) if match(i, j) else 0 for i in model_keys for j in ckpt_keys]
+ match_matrix = torch.as_tensor(match_matrix).view(len(model_keys), len(ckpt_keys))
+ # use the matched one with longest size in case of multiple matches
+ max_match_size, idxs = match_matrix.max(1)
+ # remove indices that correspond to no-match
+ idxs[max_match_size == 0] = -1
+
+ logger = logging.getLogger(__name__)
+ # matched_pairs (matched checkpoint key --> matched model key)
+ matched_keys = {}
+ result_state_dict = {}
+ for idx_model, idx_ckpt in enumerate(idxs.tolist()):
+ if idx_ckpt == -1:
+ continue
+ key_model = model_keys[idx_model]
+ key_ckpt = ckpt_keys[idx_ckpt]
+ value_ckpt = ckpt_state_dict[key_ckpt]
+ shape_in_model = model_state_dict[key_model].shape
+
+ if shape_in_model != value_ckpt.shape:
+ logger.warning(
+ "Shape of {} in checkpoint is {}, while shape of {} in model is {}.".format(
+ key_ckpt, value_ckpt.shape, key_model, shape_in_model
+ )
+ )
+ logger.warning(
+ "{} will not be loaded. Please double check and see if this is desired.".format(
+ key_ckpt
+ )
+ )
+ continue
+
+ assert key_model not in result_state_dict
+ result_state_dict[key_model] = value_ckpt
+ if key_ckpt in matched_keys: # already added to matched_keys
+ logger.error(
+ "Ambiguity found for {} in checkpoint!"
+ "It matches at least two keys in the model ({} and {}).".format(
+ key_ckpt, key_model, matched_keys[key_ckpt]
+ )
+ )
+ raise ValueError("Cannot match one checkpoint key to multiple keys in the model.")
+
+ matched_keys[key_ckpt] = key_model
+
+ # logging:
+ matched_model_keys = sorted(matched_keys.values())
+ if len(matched_model_keys) == 0:
+ logger.warning("No weights in checkpoint matched with model.")
+ return ckpt_state_dict
+ common_prefix = _longest_common_prefix(matched_model_keys)
+ rev_matched_keys = {v: k for k, v in matched_keys.items()}
+ original_keys = {k: original_keys[rev_matched_keys[k]] for k in matched_model_keys}
+
+ model_key_groups = _group_keys_by_module(matched_model_keys, original_keys)
+ table = []
+ memo = set()
+ for key_model in matched_model_keys:
+ if key_model in memo:
+ continue
+ if key_model in model_key_groups:
+ group = model_key_groups[key_model]
+ memo |= set(group)
+ shapes = [tuple(model_state_dict[k].shape) for k in group]
+ table.append(
+ (
+ _longest_common_prefix([k[len(common_prefix) :] for k in group]) + "*",
+ _group_str([original_keys[k] for k in group]),
+ " ".join([str(x).replace(" ", "") for x in shapes]),
+ )
+ )
+ else:
+ key_checkpoint = original_keys[key_model]
+ shape = str(tuple(model_state_dict[key_model].shape))
+ table.append((key_model[len(common_prefix) :], key_checkpoint, shape))
+ table_str = tabulate(
+ table, tablefmt="pipe", headers=["Names in Model", "Names in Checkpoint", "Shapes"]
+ )
+ logger.info(
+ "Following weights matched with "
+ + (f"submodule {common_prefix[:-1]}" if common_prefix else "model")
+ + ":\n"
+ + table_str
+ )
+
+ unmatched_ckpt_keys = [k for k in ckpt_keys if k not in set(matched_keys.keys())]
+ for k in unmatched_ckpt_keys:
+ result_state_dict[k] = ckpt_state_dict[k]
+ return result_state_dict
+
+
+def _group_keys_by_module(keys: List[str], original_names: Dict[str, str]):
+ """
+ Params in the same submodule are grouped together.
+
+ Args:
+ keys: names of all parameters
+ original_names: mapping from parameter name to their name in the checkpoint
+
+ Returns:
+ dict[name -> all other names in the same group]
+ """
+
+ def _submodule_name(key):
+ pos = key.rfind(".")
+ if pos < 0:
+ return None
+ prefix = key[: pos + 1]
+ return prefix
+
+ all_submodules = [_submodule_name(k) for k in keys]
+ all_submodules = [x for x in all_submodules if x]
+ all_submodules = sorted(all_submodules, key=len)
+
+ ret = {}
+ for prefix in all_submodules:
+ group = [k for k in keys if k.startswith(prefix)]
+ if len(group) <= 1:
+ continue
+ original_name_lcp = _longest_common_prefix_str([original_names[k] for k in group])
+ if len(original_name_lcp) == 0:
+ # don't group weights if original names don't share prefix
+ continue
+
+ for k in group:
+ if k in ret:
+ continue
+ ret[k] = group
+ return ret
+
+
+def _longest_common_prefix(names: List[str]) -> str:
+ """
+ ["abc.zfg", "abc.zef"] -> "abc."
+ """
+ names = [n.split(".") for n in names]
+ m1, m2 = min(names), max(names)
+ ret = [a for a, b in zip(m1, m2) if a == b]
+ ret = ".".join(ret) + "." if len(ret) else ""
+ return ret
+
+
+def _longest_common_prefix_str(names: List[str]) -> str:
+ m1, m2 = min(names), max(names)
+ lcp = []
+ for a, b in zip(m1, m2):
+ if a == b:
+ lcp.append(a)
+ else:
+ break
+ lcp = "".join(lcp)
+ return lcp
+
+
+def _group_str(names: List[str]) -> str:
+ """
+ Turn "common1", "common2", "common3" into "common{1,2,3}"
+ """
+ lcp = _longest_common_prefix_str(names)
+ rest = [x[len(lcp) :] for x in names]
+ rest = "{" + ",".join(rest) + "}"
+ ret = lcp + rest
+
+ # add some simplification for BN specifically
+ ret = ret.replace("bn_{beta,running_mean,running_var,gamma}", "bn_*")
+ ret = ret.replace("bn_beta,bn_running_mean,bn_running_var,bn_gamma", "bn_*")
+ return ret
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/checkpoint/catalog.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/checkpoint/catalog.py
new file mode 100644
index 0000000000000000000000000000000000000000..b5641858fea4936ad10b07a4237faba78dda77ff
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/checkpoint/catalog.py
@@ -0,0 +1,115 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import logging
+
+from annotator.oneformer.detectron2.utils.file_io import PathHandler, PathManager
+
+
+class ModelCatalog(object):
+ """
+ Store mappings from names to third-party models.
+ """
+
+ S3_C2_DETECTRON_PREFIX = "https://dl.fbaipublicfiles.com/detectron"
+
+ # MSRA models have STRIDE_IN_1X1=True. False otherwise.
+ # NOTE: all BN models here have fused BN into an affine layer.
+ # As a result, you should only load them to a model with "FrozenBN".
+ # Loading them to a model with regular BN or SyncBN is wrong.
+ # Even when loaded to FrozenBN, it is still different from affine by an epsilon,
+ # which should be negligible for training.
+ # NOTE: all models here uses PIXEL_STD=[1,1,1]
+ # NOTE: Most of the BN models here are no longer used. We use the
+ # re-converted pre-trained models under detectron2 model zoo instead.
+ C2_IMAGENET_MODELS = {
+ "MSRA/R-50": "ImageNetPretrained/MSRA/R-50.pkl",
+ "MSRA/R-101": "ImageNetPretrained/MSRA/R-101.pkl",
+ "FAIR/R-50-GN": "ImageNetPretrained/47261647/R-50-GN.pkl",
+ "FAIR/R-101-GN": "ImageNetPretrained/47592356/R-101-GN.pkl",
+ "FAIR/X-101-32x8d": "ImageNetPretrained/20171220/X-101-32x8d.pkl",
+ "FAIR/X-101-64x4d": "ImageNetPretrained/FBResNeXt/X-101-64x4d.pkl",
+ "FAIR/X-152-32x8d-IN5k": "ImageNetPretrained/25093814/X-152-32x8d-IN5k.pkl",
+ }
+
+ C2_DETECTRON_PATH_FORMAT = (
+ "{prefix}/{url}/output/train/{dataset}/{type}/model_final.pkl" # noqa B950
+ )
+
+ C2_DATASET_COCO = "coco_2014_train%3Acoco_2014_valminusminival"
+ C2_DATASET_COCO_KEYPOINTS = "keypoints_coco_2014_train%3Akeypoints_coco_2014_valminusminival"
+
+ # format: {model_name} -> part of the url
+ C2_DETECTRON_MODELS = {
+ "35857197/e2e_faster_rcnn_R-50-C4_1x": "35857197/12_2017_baselines/e2e_faster_rcnn_R-50-C4_1x.yaml.01_33_49.iAX0mXvW", # noqa B950
+ "35857345/e2e_faster_rcnn_R-50-FPN_1x": "35857345/12_2017_baselines/e2e_faster_rcnn_R-50-FPN_1x.yaml.01_36_30.cUF7QR7I", # noqa B950
+ "35857890/e2e_faster_rcnn_R-101-FPN_1x": "35857890/12_2017_baselines/e2e_faster_rcnn_R-101-FPN_1x.yaml.01_38_50.sNxI7sX7", # noqa B950
+ "36761737/e2e_faster_rcnn_X-101-32x8d-FPN_1x": "36761737/12_2017_baselines/e2e_faster_rcnn_X-101-32x8d-FPN_1x.yaml.06_31_39.5MIHi1fZ", # noqa B950
+ "35858791/e2e_mask_rcnn_R-50-C4_1x": "35858791/12_2017_baselines/e2e_mask_rcnn_R-50-C4_1x.yaml.01_45_57.ZgkA7hPB", # noqa B950
+ "35858933/e2e_mask_rcnn_R-50-FPN_1x": "35858933/12_2017_baselines/e2e_mask_rcnn_R-50-FPN_1x.yaml.01_48_14.DzEQe4wC", # noqa B950
+ "35861795/e2e_mask_rcnn_R-101-FPN_1x": "35861795/12_2017_baselines/e2e_mask_rcnn_R-101-FPN_1x.yaml.02_31_37.KqyEK4tT", # noqa B950
+ "36761843/e2e_mask_rcnn_X-101-32x8d-FPN_1x": "36761843/12_2017_baselines/e2e_mask_rcnn_X-101-32x8d-FPN_1x.yaml.06_35_59.RZotkLKI", # noqa B950
+ "48616381/e2e_mask_rcnn_R-50-FPN_2x_gn": "GN/48616381/04_2018_gn_baselines/e2e_mask_rcnn_R-50-FPN_2x_gn_0416.13_23_38.bTlTI97Q", # noqa B950
+ "37697547/e2e_keypoint_rcnn_R-50-FPN_1x": "37697547/12_2017_baselines/e2e_keypoint_rcnn_R-50-FPN_1x.yaml.08_42_54.kdzV35ao", # noqa B950
+ "35998355/rpn_R-50-C4_1x": "35998355/12_2017_baselines/rpn_R-50-C4_1x.yaml.08_00_43.njH5oD9L", # noqa B950
+ "35998814/rpn_R-50-FPN_1x": "35998814/12_2017_baselines/rpn_R-50-FPN_1x.yaml.08_06_03.Axg0r179", # noqa B950
+ "36225147/fast_R-50-FPN_1x": "36225147/12_2017_baselines/fast_rcnn_R-50-FPN_1x.yaml.08_39_09.L3obSdQ2", # noqa B950
+ }
+
+ @staticmethod
+ def get(name):
+ if name.startswith("Caffe2Detectron/COCO"):
+ return ModelCatalog._get_c2_detectron_baseline(name)
+ if name.startswith("ImageNetPretrained/"):
+ return ModelCatalog._get_c2_imagenet_pretrained(name)
+ raise RuntimeError("model not present in the catalog: {}".format(name))
+
+ @staticmethod
+ def _get_c2_imagenet_pretrained(name):
+ prefix = ModelCatalog.S3_C2_DETECTRON_PREFIX
+ name = name[len("ImageNetPretrained/") :]
+ name = ModelCatalog.C2_IMAGENET_MODELS[name]
+ url = "/".join([prefix, name])
+ return url
+
+ @staticmethod
+ def _get_c2_detectron_baseline(name):
+ name = name[len("Caffe2Detectron/COCO/") :]
+ url = ModelCatalog.C2_DETECTRON_MODELS[name]
+ if "keypoint_rcnn" in name:
+ dataset = ModelCatalog.C2_DATASET_COCO_KEYPOINTS
+ else:
+ dataset = ModelCatalog.C2_DATASET_COCO
+
+ if "35998355/rpn_R-50-C4_1x" in name:
+ # this one model is somehow different from others ..
+ type = "rpn"
+ else:
+ type = "generalized_rcnn"
+
+ # Detectron C2 models are stored in the structure defined in `C2_DETECTRON_PATH_FORMAT`.
+ url = ModelCatalog.C2_DETECTRON_PATH_FORMAT.format(
+ prefix=ModelCatalog.S3_C2_DETECTRON_PREFIX, url=url, type=type, dataset=dataset
+ )
+ return url
+
+
+class ModelCatalogHandler(PathHandler):
+ """
+ Resolve URL like catalog://.
+ """
+
+ PREFIX = "catalog://"
+
+ def _get_supported_prefixes(self):
+ return [self.PREFIX]
+
+ def _get_local_path(self, path, **kwargs):
+ logger = logging.getLogger(__name__)
+ catalog_path = ModelCatalog.get(path[len(self.PREFIX) :])
+ logger.info("Catalog entry {} points to {}".format(path, catalog_path))
+ return PathManager.get_local_path(catalog_path, **kwargs)
+
+ def _open(self, path, mode="r", **kwargs):
+ return PathManager.open(self._get_local_path(path), mode, **kwargs)
+
+
+PathManager.register_handler(ModelCatalogHandler())
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/checkpoint/detection_checkpoint.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/checkpoint/detection_checkpoint.py
new file mode 100644
index 0000000000000000000000000000000000000000..7d411e54bd5e004504423ba052db6f85ec511f72
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/checkpoint/detection_checkpoint.py
@@ -0,0 +1,145 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import logging
+import os
+import pickle
+from urllib.parse import parse_qs, urlparse
+import torch
+from fvcore.common.checkpoint import Checkpointer
+from torch.nn.parallel import DistributedDataParallel
+
+import annotator.oneformer.detectron2.utils.comm as comm
+from annotator.oneformer.detectron2.utils.file_io import PathManager
+
+from .c2_model_loading import align_and_update_state_dicts
+
+
+class DetectionCheckpointer(Checkpointer):
+ """
+ Same as :class:`Checkpointer`, but is able to:
+ 1. handle models in detectron & detectron2 model zoo, and apply conversions for legacy models.
+ 2. correctly load checkpoints that are only available on the master worker
+ """
+
+ def __init__(self, model, save_dir="", *, save_to_disk=None, **checkpointables):
+ is_main_process = comm.is_main_process()
+ super().__init__(
+ model,
+ save_dir,
+ save_to_disk=is_main_process if save_to_disk is None else save_to_disk,
+ **checkpointables,
+ )
+ self.path_manager = PathManager
+ self._parsed_url_during_load = None
+
+ def load(self, path, *args, **kwargs):
+ assert self._parsed_url_during_load is None
+ need_sync = False
+ logger = logging.getLogger(__name__)
+ logger.info("[DetectionCheckpointer] Loading from {} ...".format(path))
+
+ if path and isinstance(self.model, DistributedDataParallel):
+ path = self.path_manager.get_local_path(path)
+ has_file = os.path.isfile(path)
+ all_has_file = comm.all_gather(has_file)
+ if not all_has_file[0]:
+ raise OSError(f"File {path} not found on main worker.")
+ if not all(all_has_file):
+ logger.warning(
+ f"Not all workers can read checkpoint {path}. "
+ "Training may fail to fully resume."
+ )
+ # TODO: broadcast the checkpoint file contents from main
+ # worker, and load from it instead.
+ need_sync = True
+ if not has_file:
+ path = None # don't load if not readable
+
+ if path:
+ parsed_url = urlparse(path)
+ self._parsed_url_during_load = parsed_url
+ path = parsed_url._replace(query="").geturl() # remove query from filename
+ path = self.path_manager.get_local_path(path)
+
+ self.logger.setLevel('CRITICAL')
+ ret = super().load(path, *args, **kwargs)
+
+ if need_sync:
+ logger.info("Broadcasting model states from main worker ...")
+ self.model._sync_params_and_buffers()
+ self._parsed_url_during_load = None # reset to None
+ return ret
+
+ def _load_file(self, filename):
+ if filename.endswith(".pkl"):
+ with PathManager.open(filename, "rb") as f:
+ data = pickle.load(f, encoding="latin1")
+ if "model" in data and "__author__" in data:
+ # file is in Detectron2 model zoo format
+ self.logger.info("Reading a file from '{}'".format(data["__author__"]))
+ return data
+ else:
+ # assume file is from Caffe2 / Detectron1 model zoo
+ if "blobs" in data:
+ # Detection models have "blobs", but ImageNet models don't
+ data = data["blobs"]
+ data = {k: v for k, v in data.items() if not k.endswith("_momentum")}
+ return {"model": data, "__author__": "Caffe2", "matching_heuristics": True}
+ elif filename.endswith(".pyth"):
+ # assume file is from pycls; no one else seems to use the ".pyth" extension
+ with PathManager.open(filename, "rb") as f:
+ data = torch.load(f)
+ assert (
+ "model_state" in data
+ ), f"Cannot load .pyth file {filename}; pycls checkpoints must contain 'model_state'."
+ model_state = {
+ k: v
+ for k, v in data["model_state"].items()
+ if not k.endswith("num_batches_tracked")
+ }
+ return {"model": model_state, "__author__": "pycls", "matching_heuristics": True}
+
+ loaded = self._torch_load(filename)
+ if "model" not in loaded:
+ loaded = {"model": loaded}
+ assert self._parsed_url_during_load is not None, "`_load_file` must be called inside `load`"
+ parsed_url = self._parsed_url_during_load
+ queries = parse_qs(parsed_url.query)
+ if queries.pop("matching_heuristics", "False") == ["True"]:
+ loaded["matching_heuristics"] = True
+ if len(queries) > 0:
+ raise ValueError(
+ f"Unsupported query remaining: f{queries}, orginal filename: {parsed_url.geturl()}"
+ )
+ return loaded
+
+ def _torch_load(self, f):
+ return super()._load_file(f)
+
+ def _load_model(self, checkpoint):
+ if checkpoint.get("matching_heuristics", False):
+ self._convert_ndarray_to_tensor(checkpoint["model"])
+ # convert weights by name-matching heuristics
+ checkpoint["model"] = align_and_update_state_dicts(
+ self.model.state_dict(),
+ checkpoint["model"],
+ c2_conversion=checkpoint.get("__author__", None) == "Caffe2",
+ )
+ # for non-caffe2 models, use standard ways to load it
+ incompatible = super()._load_model(checkpoint)
+
+ model_buffers = dict(self.model.named_buffers(recurse=False))
+ for k in ["pixel_mean", "pixel_std"]:
+ # Ignore missing key message about pixel_mean/std.
+ # Though they may be missing in old checkpoints, they will be correctly
+ # initialized from config anyway.
+ if k in model_buffers:
+ try:
+ incompatible.missing_keys.remove(k)
+ except ValueError:
+ pass
+ for k in incompatible.unexpected_keys[:]:
+ # Ignore unexpected keys about cell anchors. They exist in old checkpoints
+ # but now they are non-persistent buffers and will not be in new checkpoints.
+ if "anchor_generator.cell_anchors" in k:
+ incompatible.unexpected_keys.remove(k)
+ return incompatible
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/config/__init__.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/config/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..a78ed118685fcfd869f7a72caf6b94621530196a
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/config/__init__.py
@@ -0,0 +1,24 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+from .compat import downgrade_config, upgrade_config
+from .config import CfgNode, get_cfg, global_cfg, set_global_cfg, configurable
+from .instantiate import instantiate
+from .lazy import LazyCall, LazyConfig
+
+__all__ = [
+ "CfgNode",
+ "get_cfg",
+ "global_cfg",
+ "set_global_cfg",
+ "downgrade_config",
+ "upgrade_config",
+ "configurable",
+ "instantiate",
+ "LazyCall",
+ "LazyConfig",
+]
+
+
+from annotator.oneformer.detectron2.utils.env import fixup_module_metadata
+
+fixup_module_metadata(__name__, globals(), __all__)
+del fixup_module_metadata
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/config/compat.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/config/compat.py
new file mode 100644
index 0000000000000000000000000000000000000000..11a08c439bf14defd880e37a938fab8a08e68eeb
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/config/compat.py
@@ -0,0 +1,229 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+"""
+Backward compatibility of configs.
+
+Instructions to bump version:
++ It's not needed to bump version if new keys are added.
+ It's only needed when backward-incompatible changes happen
+ (i.e., some existing keys disappear, or the meaning of a key changes)
++ To bump version, do the following:
+ 1. Increment _C.VERSION in defaults.py
+ 2. Add a converter in this file.
+
+ Each ConverterVX has a function "upgrade" which in-place upgrades config from X-1 to X,
+ and a function "downgrade" which in-place downgrades config from X to X-1
+
+ In each function, VERSION is left unchanged.
+
+ Each converter assumes that its input has the relevant keys
+ (i.e., the input is not a partial config).
+ 3. Run the tests (test_config.py) to make sure the upgrade & downgrade
+ functions are consistent.
+"""
+
+import logging
+from typing import List, Optional, Tuple
+
+from .config import CfgNode as CN
+from .defaults import _C
+
+__all__ = ["upgrade_config", "downgrade_config"]
+
+
+def upgrade_config(cfg: CN, to_version: Optional[int] = None) -> CN:
+ """
+ Upgrade a config from its current version to a newer version.
+
+ Args:
+ cfg (CfgNode):
+ to_version (int): defaults to the latest version.
+ """
+ cfg = cfg.clone()
+ if to_version is None:
+ to_version = _C.VERSION
+
+ assert cfg.VERSION <= to_version, "Cannot upgrade from v{} to v{}!".format(
+ cfg.VERSION, to_version
+ )
+ for k in range(cfg.VERSION, to_version):
+ converter = globals()["ConverterV" + str(k + 1)]
+ converter.upgrade(cfg)
+ cfg.VERSION = k + 1
+ return cfg
+
+
+def downgrade_config(cfg: CN, to_version: int) -> CN:
+ """
+ Downgrade a config from its current version to an older version.
+
+ Args:
+ cfg (CfgNode):
+ to_version (int):
+
+ Note:
+ A general downgrade of arbitrary configs is not always possible due to the
+ different functionalities in different versions.
+ The purpose of downgrade is only to recover the defaults in old versions,
+ allowing it to load an old partial yaml config.
+ Therefore, the implementation only needs to fill in the default values
+ in the old version when a general downgrade is not possible.
+ """
+ cfg = cfg.clone()
+ assert cfg.VERSION >= to_version, "Cannot downgrade from v{} to v{}!".format(
+ cfg.VERSION, to_version
+ )
+ for k in range(cfg.VERSION, to_version, -1):
+ converter = globals()["ConverterV" + str(k)]
+ converter.downgrade(cfg)
+ cfg.VERSION = k - 1
+ return cfg
+
+
+def guess_version(cfg: CN, filename: str) -> int:
+ """
+ Guess the version of a partial config where the VERSION field is not specified.
+ Returns the version, or the latest if cannot make a guess.
+
+ This makes it easier for users to migrate.
+ """
+ logger = logging.getLogger(__name__)
+
+ def _has(name: str) -> bool:
+ cur = cfg
+ for n in name.split("."):
+ if n not in cur:
+ return False
+ cur = cur[n]
+ return True
+
+ # Most users' partial configs have "MODEL.WEIGHT", so guess on it
+ ret = None
+ if _has("MODEL.WEIGHT") or _has("TEST.AUG_ON"):
+ ret = 1
+
+ if ret is not None:
+ logger.warning("Config '{}' has no VERSION. Assuming it to be v{}.".format(filename, ret))
+ else:
+ ret = _C.VERSION
+ logger.warning(
+ "Config '{}' has no VERSION. Assuming it to be compatible with latest v{}.".format(
+ filename, ret
+ )
+ )
+ return ret
+
+
+def _rename(cfg: CN, old: str, new: str) -> None:
+ old_keys = old.split(".")
+ new_keys = new.split(".")
+
+ def _set(key_seq: List[str], val: str) -> None:
+ cur = cfg
+ for k in key_seq[:-1]:
+ if k not in cur:
+ cur[k] = CN()
+ cur = cur[k]
+ cur[key_seq[-1]] = val
+
+ def _get(key_seq: List[str]) -> CN:
+ cur = cfg
+ for k in key_seq:
+ cur = cur[k]
+ return cur
+
+ def _del(key_seq: List[str]) -> None:
+ cur = cfg
+ for k in key_seq[:-1]:
+ cur = cur[k]
+ del cur[key_seq[-1]]
+ if len(cur) == 0 and len(key_seq) > 1:
+ _del(key_seq[:-1])
+
+ _set(new_keys, _get(old_keys))
+ _del(old_keys)
+
+
+class _RenameConverter:
+ """
+ A converter that handles simple rename.
+ """
+
+ RENAME: List[Tuple[str, str]] = [] # list of tuples of (old name, new name)
+
+ @classmethod
+ def upgrade(cls, cfg: CN) -> None:
+ for old, new in cls.RENAME:
+ _rename(cfg, old, new)
+
+ @classmethod
+ def downgrade(cls, cfg: CN) -> None:
+ for old, new in cls.RENAME[::-1]:
+ _rename(cfg, new, old)
+
+
+class ConverterV1(_RenameConverter):
+ RENAME = [("MODEL.RPN_HEAD.NAME", "MODEL.RPN.HEAD_NAME")]
+
+
+class ConverterV2(_RenameConverter):
+ """
+ A large bulk of rename, before public release.
+ """
+
+ RENAME = [
+ ("MODEL.WEIGHT", "MODEL.WEIGHTS"),
+ ("MODEL.PANOPTIC_FPN.SEMANTIC_LOSS_SCALE", "MODEL.SEM_SEG_HEAD.LOSS_WEIGHT"),
+ ("MODEL.PANOPTIC_FPN.RPN_LOSS_SCALE", "MODEL.RPN.LOSS_WEIGHT"),
+ ("MODEL.PANOPTIC_FPN.INSTANCE_LOSS_SCALE", "MODEL.PANOPTIC_FPN.INSTANCE_LOSS_WEIGHT"),
+ ("MODEL.PANOPTIC_FPN.COMBINE_ON", "MODEL.PANOPTIC_FPN.COMBINE.ENABLED"),
+ (
+ "MODEL.PANOPTIC_FPN.COMBINE_OVERLAP_THRESHOLD",
+ "MODEL.PANOPTIC_FPN.COMBINE.OVERLAP_THRESH",
+ ),
+ (
+ "MODEL.PANOPTIC_FPN.COMBINE_STUFF_AREA_LIMIT",
+ "MODEL.PANOPTIC_FPN.COMBINE.STUFF_AREA_LIMIT",
+ ),
+ (
+ "MODEL.PANOPTIC_FPN.COMBINE_INSTANCES_CONFIDENCE_THRESHOLD",
+ "MODEL.PANOPTIC_FPN.COMBINE.INSTANCES_CONFIDENCE_THRESH",
+ ),
+ ("MODEL.ROI_HEADS.SCORE_THRESH", "MODEL.ROI_HEADS.SCORE_THRESH_TEST"),
+ ("MODEL.ROI_HEADS.NMS", "MODEL.ROI_HEADS.NMS_THRESH_TEST"),
+ ("MODEL.RETINANET.INFERENCE_SCORE_THRESHOLD", "MODEL.RETINANET.SCORE_THRESH_TEST"),
+ ("MODEL.RETINANET.INFERENCE_TOPK_CANDIDATES", "MODEL.RETINANET.TOPK_CANDIDATES_TEST"),
+ ("MODEL.RETINANET.INFERENCE_NMS_THRESHOLD", "MODEL.RETINANET.NMS_THRESH_TEST"),
+ ("TEST.DETECTIONS_PER_IMG", "TEST.DETECTIONS_PER_IMAGE"),
+ ("TEST.AUG_ON", "TEST.AUG.ENABLED"),
+ ("TEST.AUG_MIN_SIZES", "TEST.AUG.MIN_SIZES"),
+ ("TEST.AUG_MAX_SIZE", "TEST.AUG.MAX_SIZE"),
+ ("TEST.AUG_FLIP", "TEST.AUG.FLIP"),
+ ]
+
+ @classmethod
+ def upgrade(cls, cfg: CN) -> None:
+ super().upgrade(cfg)
+
+ if cfg.MODEL.META_ARCHITECTURE == "RetinaNet":
+ _rename(
+ cfg, "MODEL.RETINANET.ANCHOR_ASPECT_RATIOS", "MODEL.ANCHOR_GENERATOR.ASPECT_RATIOS"
+ )
+ _rename(cfg, "MODEL.RETINANET.ANCHOR_SIZES", "MODEL.ANCHOR_GENERATOR.SIZES")
+ del cfg["MODEL"]["RPN"]["ANCHOR_SIZES"]
+ del cfg["MODEL"]["RPN"]["ANCHOR_ASPECT_RATIOS"]
+ else:
+ _rename(cfg, "MODEL.RPN.ANCHOR_ASPECT_RATIOS", "MODEL.ANCHOR_GENERATOR.ASPECT_RATIOS")
+ _rename(cfg, "MODEL.RPN.ANCHOR_SIZES", "MODEL.ANCHOR_GENERATOR.SIZES")
+ del cfg["MODEL"]["RETINANET"]["ANCHOR_SIZES"]
+ del cfg["MODEL"]["RETINANET"]["ANCHOR_ASPECT_RATIOS"]
+ del cfg["MODEL"]["RETINANET"]["ANCHOR_STRIDES"]
+
+ @classmethod
+ def downgrade(cls, cfg: CN) -> None:
+ super().downgrade(cfg)
+
+ _rename(cfg, "MODEL.ANCHOR_GENERATOR.ASPECT_RATIOS", "MODEL.RPN.ANCHOR_ASPECT_RATIOS")
+ _rename(cfg, "MODEL.ANCHOR_GENERATOR.SIZES", "MODEL.RPN.ANCHOR_SIZES")
+ cfg.MODEL.RETINANET.ANCHOR_ASPECT_RATIOS = cfg.MODEL.RPN.ANCHOR_ASPECT_RATIOS
+ cfg.MODEL.RETINANET.ANCHOR_SIZES = cfg.MODEL.RPN.ANCHOR_SIZES
+ cfg.MODEL.RETINANET.ANCHOR_STRIDES = [] # this is not used anywhere in any version
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/config/config.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/config/config.py
new file mode 100644
index 0000000000000000000000000000000000000000..c5b1303422481dc7adb3ee5221377770e0c01a81
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/config/config.py
@@ -0,0 +1,265 @@
+# -*- coding: utf-8 -*-
+# Copyright (c) Facebook, Inc. and its affiliates.
+
+import functools
+import inspect
+import logging
+from fvcore.common.config import CfgNode as _CfgNode
+
+from annotator.oneformer.detectron2.utils.file_io import PathManager
+
+
+class CfgNode(_CfgNode):
+ """
+ The same as `fvcore.common.config.CfgNode`, but different in:
+
+ 1. Use unsafe yaml loading by default.
+ Note that this may lead to arbitrary code execution: you must not
+ load a config file from untrusted sources before manually inspecting
+ the content of the file.
+ 2. Support config versioning.
+ When attempting to merge an old config, it will convert the old config automatically.
+
+ .. automethod:: clone
+ .. automethod:: freeze
+ .. automethod:: defrost
+ .. automethod:: is_frozen
+ .. automethod:: load_yaml_with_base
+ .. automethod:: merge_from_list
+ .. automethod:: merge_from_other_cfg
+ """
+
+ @classmethod
+ def _open_cfg(cls, filename):
+ return PathManager.open(filename, "r")
+
+ # Note that the default value of allow_unsafe is changed to True
+ def merge_from_file(self, cfg_filename: str, allow_unsafe: bool = True) -> None:
+ """
+ Load content from the given config file and merge it into self.
+
+ Args:
+ cfg_filename: config filename
+ allow_unsafe: allow unsafe yaml syntax
+ """
+ assert PathManager.isfile(cfg_filename), f"Config file '{cfg_filename}' does not exist!"
+ loaded_cfg = self.load_yaml_with_base(cfg_filename, allow_unsafe=allow_unsafe)
+ loaded_cfg = type(self)(loaded_cfg)
+
+ # defaults.py needs to import CfgNode
+ from .defaults import _C
+
+ latest_ver = _C.VERSION
+ assert (
+ latest_ver == self.VERSION
+ ), "CfgNode.merge_from_file is only allowed on a config object of latest version!"
+
+ logger = logging.getLogger(__name__)
+
+ loaded_ver = loaded_cfg.get("VERSION", None)
+ if loaded_ver is None:
+ from .compat import guess_version
+
+ loaded_ver = guess_version(loaded_cfg, cfg_filename)
+ assert loaded_ver <= self.VERSION, "Cannot merge a v{} config into a v{} config.".format(
+ loaded_ver, self.VERSION
+ )
+
+ if loaded_ver == self.VERSION:
+ self.merge_from_other_cfg(loaded_cfg)
+ else:
+ # compat.py needs to import CfgNode
+ from .compat import upgrade_config, downgrade_config
+
+ logger.warning(
+ "Loading an old v{} config file '{}' by automatically upgrading to v{}. "
+ "See docs/CHANGELOG.md for instructions to update your files.".format(
+ loaded_ver, cfg_filename, self.VERSION
+ )
+ )
+ # To convert, first obtain a full config at an old version
+ old_self = downgrade_config(self, to_version=loaded_ver)
+ old_self.merge_from_other_cfg(loaded_cfg)
+ new_config = upgrade_config(old_self)
+ self.clear()
+ self.update(new_config)
+
+ def dump(self, *args, **kwargs):
+ """
+ Returns:
+ str: a yaml string representation of the config
+ """
+ # to make it show up in docs
+ return super().dump(*args, **kwargs)
+
+
+global_cfg = CfgNode()
+
+
+def get_cfg() -> CfgNode:
+ """
+ Get a copy of the default config.
+
+ Returns:
+ a detectron2 CfgNode instance.
+ """
+ from .defaults import _C
+
+ return _C.clone()
+
+
+def set_global_cfg(cfg: CfgNode) -> None:
+ """
+ Let the global config point to the given cfg.
+
+ Assume that the given "cfg" has the key "KEY", after calling
+ `set_global_cfg(cfg)`, the key can be accessed by:
+ ::
+ from annotator.oneformer.detectron2.config import global_cfg
+ print(global_cfg.KEY)
+
+ By using a hacky global config, you can access these configs anywhere,
+ without having to pass the config object or the values deep into the code.
+ This is a hacky feature introduced for quick prototyping / research exploration.
+ """
+ global global_cfg
+ global_cfg.clear()
+ global_cfg.update(cfg)
+
+
+def configurable(init_func=None, *, from_config=None):
+ """
+ Decorate a function or a class's __init__ method so that it can be called
+ with a :class:`CfgNode` object using a :func:`from_config` function that translates
+ :class:`CfgNode` to arguments.
+
+ Examples:
+ ::
+ # Usage 1: Decorator on __init__:
+ class A:
+ @configurable
+ def __init__(self, a, b=2, c=3):
+ pass
+
+ @classmethod
+ def from_config(cls, cfg): # 'cfg' must be the first argument
+ # Returns kwargs to be passed to __init__
+ return {"a": cfg.A, "b": cfg.B}
+
+ a1 = A(a=1, b=2) # regular construction
+ a2 = A(cfg) # construct with a cfg
+ a3 = A(cfg, b=3, c=4) # construct with extra overwrite
+
+ # Usage 2: Decorator on any function. Needs an extra from_config argument:
+ @configurable(from_config=lambda cfg: {"a: cfg.A, "b": cfg.B})
+ def a_func(a, b=2, c=3):
+ pass
+
+ a1 = a_func(a=1, b=2) # regular call
+ a2 = a_func(cfg) # call with a cfg
+ a3 = a_func(cfg, b=3, c=4) # call with extra overwrite
+
+ Args:
+ init_func (callable): a class's ``__init__`` method in usage 1. The
+ class must have a ``from_config`` classmethod which takes `cfg` as
+ the first argument.
+ from_config (callable): the from_config function in usage 2. It must take `cfg`
+ as its first argument.
+ """
+
+ if init_func is not None:
+ assert (
+ inspect.isfunction(init_func)
+ and from_config is None
+ and init_func.__name__ == "__init__"
+ ), "Incorrect use of @configurable. Check API documentation for examples."
+
+ @functools.wraps(init_func)
+ def wrapped(self, *args, **kwargs):
+ try:
+ from_config_func = type(self).from_config
+ except AttributeError as e:
+ raise AttributeError(
+ "Class with @configurable must have a 'from_config' classmethod."
+ ) from e
+ if not inspect.ismethod(from_config_func):
+ raise TypeError("Class with @configurable must have a 'from_config' classmethod.")
+
+ if _called_with_cfg(*args, **kwargs):
+ explicit_args = _get_args_from_config(from_config_func, *args, **kwargs)
+ init_func(self, **explicit_args)
+ else:
+ init_func(self, *args, **kwargs)
+
+ return wrapped
+
+ else:
+ if from_config is None:
+ return configurable # @configurable() is made equivalent to @configurable
+ assert inspect.isfunction(
+ from_config
+ ), "from_config argument of configurable must be a function!"
+
+ def wrapper(orig_func):
+ @functools.wraps(orig_func)
+ def wrapped(*args, **kwargs):
+ if _called_with_cfg(*args, **kwargs):
+ explicit_args = _get_args_from_config(from_config, *args, **kwargs)
+ return orig_func(**explicit_args)
+ else:
+ return orig_func(*args, **kwargs)
+
+ wrapped.from_config = from_config
+ return wrapped
+
+ return wrapper
+
+
+def _get_args_from_config(from_config_func, *args, **kwargs):
+ """
+ Use `from_config` to obtain explicit arguments.
+
+ Returns:
+ dict: arguments to be used for cls.__init__
+ """
+ signature = inspect.signature(from_config_func)
+ if list(signature.parameters.keys())[0] != "cfg":
+ if inspect.isfunction(from_config_func):
+ name = from_config_func.__name__
+ else:
+ name = f"{from_config_func.__self__}.from_config"
+ raise TypeError(f"{name} must take 'cfg' as the first argument!")
+ support_var_arg = any(
+ param.kind in [param.VAR_POSITIONAL, param.VAR_KEYWORD]
+ for param in signature.parameters.values()
+ )
+ if support_var_arg: # forward all arguments to from_config, if from_config accepts them
+ ret = from_config_func(*args, **kwargs)
+ else:
+ # forward supported arguments to from_config
+ supported_arg_names = set(signature.parameters.keys())
+ extra_kwargs = {}
+ for name in list(kwargs.keys()):
+ if name not in supported_arg_names:
+ extra_kwargs[name] = kwargs.pop(name)
+ ret = from_config_func(*args, **kwargs)
+ # forward the other arguments to __init__
+ ret.update(extra_kwargs)
+ return ret
+
+
+def _called_with_cfg(*args, **kwargs):
+ """
+ Returns:
+ bool: whether the arguments contain CfgNode and should be considered
+ forwarded to from_config.
+ """
+ from omegaconf import DictConfig
+
+ if len(args) and isinstance(args[0], (_CfgNode, DictConfig)):
+ return True
+ if isinstance(kwargs.pop("cfg", None), (_CfgNode, DictConfig)):
+ return True
+ # `from_config`'s first argument is forced to be "cfg".
+ # So the above check covers all cases.
+ return False
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/config/defaults.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/config/defaults.py
new file mode 100644
index 0000000000000000000000000000000000000000..ffb79e763f076c9ae982c727309e19b8e0ef170f
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/config/defaults.py
@@ -0,0 +1,650 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+from .config import CfgNode as CN
+
+# NOTE: given the new config system
+# (https://detectron2.readthedocs.io/en/latest/tutorials/lazyconfigs.html),
+# we will stop adding new functionalities to default CfgNode.
+
+# -----------------------------------------------------------------------------
+# Convention about Training / Test specific parameters
+# -----------------------------------------------------------------------------
+# Whenever an argument can be either used for training or for testing, the
+# corresponding name will be post-fixed by a _TRAIN for a training parameter,
+# or _TEST for a test-specific parameter.
+# For example, the number of images during training will be
+# IMAGES_PER_BATCH_TRAIN, while the number of images for testing will be
+# IMAGES_PER_BATCH_TEST
+
+# -----------------------------------------------------------------------------
+# Config definition
+# -----------------------------------------------------------------------------
+
+_C = CN()
+
+# The version number, to upgrade from old configs to new ones if any
+# changes happen. It's recommended to keep a VERSION in your config file.
+_C.VERSION = 2
+
+_C.MODEL = CN()
+_C.MODEL.LOAD_PROPOSALS = False
+_C.MODEL.MASK_ON = False
+_C.MODEL.KEYPOINT_ON = False
+_C.MODEL.DEVICE = "cuda"
+_C.MODEL.META_ARCHITECTURE = "GeneralizedRCNN"
+
+# Path (a file path, or URL like detectron2://.., https://..) to a checkpoint file
+# to be loaded to the model. You can find available models in the model zoo.
+_C.MODEL.WEIGHTS = ""
+
+# Values to be used for image normalization (BGR order, since INPUT.FORMAT defaults to BGR).
+# To train on images of different number of channels, just set different mean & std.
+# Default values are the mean pixel value from ImageNet: [103.53, 116.28, 123.675]
+_C.MODEL.PIXEL_MEAN = [103.530, 116.280, 123.675]
+# When using pre-trained models in Detectron1 or any MSRA models,
+# std has been absorbed into its conv1 weights, so the std needs to be set 1.
+# Otherwise, you can use [57.375, 57.120, 58.395] (ImageNet std)
+_C.MODEL.PIXEL_STD = [1.0, 1.0, 1.0]
+
+
+# -----------------------------------------------------------------------------
+# INPUT
+# -----------------------------------------------------------------------------
+_C.INPUT = CN()
+# By default, {MIN,MAX}_SIZE options are used in transforms.ResizeShortestEdge.
+# Please refer to ResizeShortestEdge for detailed definition.
+# Size of the smallest side of the image during training
+_C.INPUT.MIN_SIZE_TRAIN = (800,)
+# Sample size of smallest side by choice or random selection from range give by
+# INPUT.MIN_SIZE_TRAIN
+_C.INPUT.MIN_SIZE_TRAIN_SAMPLING = "choice"
+# Maximum size of the side of the image during training
+_C.INPUT.MAX_SIZE_TRAIN = 1333
+# Size of the smallest side of the image during testing. Set to zero to disable resize in testing.
+_C.INPUT.MIN_SIZE_TEST = 800
+# Maximum size of the side of the image during testing
+_C.INPUT.MAX_SIZE_TEST = 1333
+# Mode for flipping images used in data augmentation during training
+# choose one of ["horizontal, "vertical", "none"]
+_C.INPUT.RANDOM_FLIP = "horizontal"
+
+# `True` if cropping is used for data augmentation during training
+_C.INPUT.CROP = CN({"ENABLED": False})
+# Cropping type. See documentation of `detectron2.data.transforms.RandomCrop` for explanation.
+_C.INPUT.CROP.TYPE = "relative_range"
+# Size of crop in range (0, 1] if CROP.TYPE is "relative" or "relative_range" and in number of
+# pixels if CROP.TYPE is "absolute"
+_C.INPUT.CROP.SIZE = [0.9, 0.9]
+
+
+# Whether the model needs RGB, YUV, HSV etc.
+# Should be one of the modes defined here, as we use PIL to read the image:
+# https://pillow.readthedocs.io/en/stable/handbook/concepts.html#concept-modes
+# with BGR being the one exception. One can set image format to BGR, we will
+# internally use RGB for conversion and flip the channels over
+_C.INPUT.FORMAT = "BGR"
+# The ground truth mask format that the model will use.
+# Mask R-CNN supports either "polygon" or "bitmask" as ground truth.
+_C.INPUT.MASK_FORMAT = "polygon" # alternative: "bitmask"
+
+
+# -----------------------------------------------------------------------------
+# Dataset
+# -----------------------------------------------------------------------------
+_C.DATASETS = CN()
+# List of the dataset names for training. Must be registered in DatasetCatalog
+# Samples from these datasets will be merged and used as one dataset.
+_C.DATASETS.TRAIN = ()
+# List of the pre-computed proposal files for training, which must be consistent
+# with datasets listed in DATASETS.TRAIN.
+_C.DATASETS.PROPOSAL_FILES_TRAIN = ()
+# Number of top scoring precomputed proposals to keep for training
+_C.DATASETS.PRECOMPUTED_PROPOSAL_TOPK_TRAIN = 2000
+# List of the dataset names for testing. Must be registered in DatasetCatalog
+_C.DATASETS.TEST = ()
+# List of the pre-computed proposal files for test, which must be consistent
+# with datasets listed in DATASETS.TEST.
+_C.DATASETS.PROPOSAL_FILES_TEST = ()
+# Number of top scoring precomputed proposals to keep for test
+_C.DATASETS.PRECOMPUTED_PROPOSAL_TOPK_TEST = 1000
+
+# -----------------------------------------------------------------------------
+# DataLoader
+# -----------------------------------------------------------------------------
+_C.DATALOADER = CN()
+# Number of data loading threads
+_C.DATALOADER.NUM_WORKERS = 4
+# If True, each batch should contain only images for which the aspect ratio
+# is compatible. This groups portrait images together, and landscape images
+# are not batched with portrait images.
+_C.DATALOADER.ASPECT_RATIO_GROUPING = True
+# Options: TrainingSampler, RepeatFactorTrainingSampler
+_C.DATALOADER.SAMPLER_TRAIN = "TrainingSampler"
+# Repeat threshold for RepeatFactorTrainingSampler
+_C.DATALOADER.REPEAT_THRESHOLD = 0.0
+# Tf True, when working on datasets that have instance annotations, the
+# training dataloader will filter out images without associated annotations
+_C.DATALOADER.FILTER_EMPTY_ANNOTATIONS = True
+
+# ---------------------------------------------------------------------------- #
+# Backbone options
+# ---------------------------------------------------------------------------- #
+_C.MODEL.BACKBONE = CN()
+
+_C.MODEL.BACKBONE.NAME = "build_resnet_backbone"
+# Freeze the first several stages so they are not trained.
+# There are 5 stages in ResNet. The first is a convolution, and the following
+# stages are each group of residual blocks.
+_C.MODEL.BACKBONE.FREEZE_AT = 2
+
+
+# ---------------------------------------------------------------------------- #
+# FPN options
+# ---------------------------------------------------------------------------- #
+_C.MODEL.FPN = CN()
+# Names of the input feature maps to be used by FPN
+# They must have contiguous power of 2 strides
+# e.g., ["res2", "res3", "res4", "res5"]
+_C.MODEL.FPN.IN_FEATURES = []
+_C.MODEL.FPN.OUT_CHANNELS = 256
+
+# Options: "" (no norm), "GN"
+_C.MODEL.FPN.NORM = ""
+
+# Types for fusing the FPN top-down and lateral features. Can be either "sum" or "avg"
+_C.MODEL.FPN.FUSE_TYPE = "sum"
+
+
+# ---------------------------------------------------------------------------- #
+# Proposal generator options
+# ---------------------------------------------------------------------------- #
+_C.MODEL.PROPOSAL_GENERATOR = CN()
+# Current proposal generators include "RPN", "RRPN" and "PrecomputedProposals"
+_C.MODEL.PROPOSAL_GENERATOR.NAME = "RPN"
+# Proposal height and width both need to be greater than MIN_SIZE
+# (a the scale used during training or inference)
+_C.MODEL.PROPOSAL_GENERATOR.MIN_SIZE = 0
+
+
+# ---------------------------------------------------------------------------- #
+# Anchor generator options
+# ---------------------------------------------------------------------------- #
+_C.MODEL.ANCHOR_GENERATOR = CN()
+# The generator can be any name in the ANCHOR_GENERATOR registry
+_C.MODEL.ANCHOR_GENERATOR.NAME = "DefaultAnchorGenerator"
+# Anchor sizes (i.e. sqrt of area) in absolute pixels w.r.t. the network input.
+# Format: list[list[float]]. SIZES[i] specifies the list of sizes to use for
+# IN_FEATURES[i]; len(SIZES) must be equal to len(IN_FEATURES) or 1.
+# When len(SIZES) == 1, SIZES[0] is used for all IN_FEATURES.
+_C.MODEL.ANCHOR_GENERATOR.SIZES = [[32, 64, 128, 256, 512]]
+# Anchor aspect ratios. For each area given in `SIZES`, anchors with different aspect
+# ratios are generated by an anchor generator.
+# Format: list[list[float]]. ASPECT_RATIOS[i] specifies the list of aspect ratios (H/W)
+# to use for IN_FEATURES[i]; len(ASPECT_RATIOS) == len(IN_FEATURES) must be true,
+# or len(ASPECT_RATIOS) == 1 is true and aspect ratio list ASPECT_RATIOS[0] is used
+# for all IN_FEATURES.
+_C.MODEL.ANCHOR_GENERATOR.ASPECT_RATIOS = [[0.5, 1.0, 2.0]]
+# Anchor angles.
+# list[list[float]], the angle in degrees, for each input feature map.
+# ANGLES[i] specifies the list of angles for IN_FEATURES[i].
+_C.MODEL.ANCHOR_GENERATOR.ANGLES = [[-90, 0, 90]]
+# Relative offset between the center of the first anchor and the top-left corner of the image
+# Value has to be in [0, 1). Recommend to use 0.5, which means half stride.
+# The value is not expected to affect model accuracy.
+_C.MODEL.ANCHOR_GENERATOR.OFFSET = 0.0
+
+# ---------------------------------------------------------------------------- #
+# RPN options
+# ---------------------------------------------------------------------------- #
+_C.MODEL.RPN = CN()
+_C.MODEL.RPN.HEAD_NAME = "StandardRPNHead" # used by RPN_HEAD_REGISTRY
+
+# Names of the input feature maps to be used by RPN
+# e.g., ["p2", "p3", "p4", "p5", "p6"] for FPN
+_C.MODEL.RPN.IN_FEATURES = ["res4"]
+# Remove RPN anchors that go outside the image by BOUNDARY_THRESH pixels
+# Set to -1 or a large value, e.g. 100000, to disable pruning anchors
+_C.MODEL.RPN.BOUNDARY_THRESH = -1
+# IOU overlap ratios [BG_IOU_THRESHOLD, FG_IOU_THRESHOLD]
+# Minimum overlap required between an anchor and ground-truth box for the
+# (anchor, gt box) pair to be a positive example (IoU >= FG_IOU_THRESHOLD
+# ==> positive RPN example: 1)
+# Maximum overlap allowed between an anchor and ground-truth box for the
+# (anchor, gt box) pair to be a negative examples (IoU < BG_IOU_THRESHOLD
+# ==> negative RPN example: 0)
+# Anchors with overlap in between (BG_IOU_THRESHOLD <= IoU < FG_IOU_THRESHOLD)
+# are ignored (-1)
+_C.MODEL.RPN.IOU_THRESHOLDS = [0.3, 0.7]
+_C.MODEL.RPN.IOU_LABELS = [0, -1, 1]
+# Number of regions per image used to train RPN
+_C.MODEL.RPN.BATCH_SIZE_PER_IMAGE = 256
+# Target fraction of foreground (positive) examples per RPN minibatch
+_C.MODEL.RPN.POSITIVE_FRACTION = 0.5
+# Options are: "smooth_l1", "giou", "diou", "ciou"
+_C.MODEL.RPN.BBOX_REG_LOSS_TYPE = "smooth_l1"
+_C.MODEL.RPN.BBOX_REG_LOSS_WEIGHT = 1.0
+# Weights on (dx, dy, dw, dh) for normalizing RPN anchor regression targets
+_C.MODEL.RPN.BBOX_REG_WEIGHTS = (1.0, 1.0, 1.0, 1.0)
+# The transition point from L1 to L2 loss. Set to 0.0 to make the loss simply L1.
+_C.MODEL.RPN.SMOOTH_L1_BETA = 0.0
+_C.MODEL.RPN.LOSS_WEIGHT = 1.0
+# Number of top scoring RPN proposals to keep before applying NMS
+# When FPN is used, this is *per FPN level* (not total)
+_C.MODEL.RPN.PRE_NMS_TOPK_TRAIN = 12000
+_C.MODEL.RPN.PRE_NMS_TOPK_TEST = 6000
+# Number of top scoring RPN proposals to keep after applying NMS
+# When FPN is used, this limit is applied per level and then again to the union
+# of proposals from all levels
+# NOTE: When FPN is used, the meaning of this config is different from Detectron1.
+# It means per-batch topk in Detectron1, but per-image topk here.
+# See the "find_top_rpn_proposals" function for details.
+_C.MODEL.RPN.POST_NMS_TOPK_TRAIN = 2000
+_C.MODEL.RPN.POST_NMS_TOPK_TEST = 1000
+# NMS threshold used on RPN proposals
+_C.MODEL.RPN.NMS_THRESH = 0.7
+# Set this to -1 to use the same number of output channels as input channels.
+_C.MODEL.RPN.CONV_DIMS = [-1]
+
+# ---------------------------------------------------------------------------- #
+# ROI HEADS options
+# ---------------------------------------------------------------------------- #
+_C.MODEL.ROI_HEADS = CN()
+_C.MODEL.ROI_HEADS.NAME = "Res5ROIHeads"
+# Number of foreground classes
+_C.MODEL.ROI_HEADS.NUM_CLASSES = 80
+# Names of the input feature maps to be used by ROI heads
+# Currently all heads (box, mask, ...) use the same input feature map list
+# e.g., ["p2", "p3", "p4", "p5"] is commonly used for FPN
+_C.MODEL.ROI_HEADS.IN_FEATURES = ["res4"]
+# IOU overlap ratios [IOU_THRESHOLD]
+# Overlap threshold for an RoI to be considered background (if < IOU_THRESHOLD)
+# Overlap threshold for an RoI to be considered foreground (if >= IOU_THRESHOLD)
+_C.MODEL.ROI_HEADS.IOU_THRESHOLDS = [0.5]
+_C.MODEL.ROI_HEADS.IOU_LABELS = [0, 1]
+# RoI minibatch size *per image* (number of regions of interest [ROIs]) during training
+# Total number of RoIs per training minibatch =
+# ROI_HEADS.BATCH_SIZE_PER_IMAGE * SOLVER.IMS_PER_BATCH
+# E.g., a common configuration is: 512 * 16 = 8192
+_C.MODEL.ROI_HEADS.BATCH_SIZE_PER_IMAGE = 512
+# Target fraction of RoI minibatch that is labeled foreground (i.e. class > 0)
+_C.MODEL.ROI_HEADS.POSITIVE_FRACTION = 0.25
+
+# Only used on test mode
+
+# Minimum score threshold (assuming scores in a [0, 1] range); a value chosen to
+# balance obtaining high recall with not having too many low precision
+# detections that will slow down inference post processing steps (like NMS)
+# A default threshold of 0.0 increases AP by ~0.2-0.3 but significantly slows down
+# inference.
+_C.MODEL.ROI_HEADS.SCORE_THRESH_TEST = 0.05
+# Overlap threshold used for non-maximum suppression (suppress boxes with
+# IoU >= this threshold)
+_C.MODEL.ROI_HEADS.NMS_THRESH_TEST = 0.5
+# If True, augment proposals with ground-truth boxes before sampling proposals to
+# train ROI heads.
+_C.MODEL.ROI_HEADS.PROPOSAL_APPEND_GT = True
+
+# ---------------------------------------------------------------------------- #
+# Box Head
+# ---------------------------------------------------------------------------- #
+_C.MODEL.ROI_BOX_HEAD = CN()
+# C4 don't use head name option
+# Options for non-C4 models: FastRCNNConvFCHead,
+_C.MODEL.ROI_BOX_HEAD.NAME = ""
+# Options are: "smooth_l1", "giou", "diou", "ciou"
+_C.MODEL.ROI_BOX_HEAD.BBOX_REG_LOSS_TYPE = "smooth_l1"
+# The final scaling coefficient on the box regression loss, used to balance the magnitude of its
+# gradients with other losses in the model. See also `MODEL.ROI_KEYPOINT_HEAD.LOSS_WEIGHT`.
+_C.MODEL.ROI_BOX_HEAD.BBOX_REG_LOSS_WEIGHT = 1.0
+# Default weights on (dx, dy, dw, dh) for normalizing bbox regression targets
+# These are empirically chosen to approximately lead to unit variance targets
+_C.MODEL.ROI_BOX_HEAD.BBOX_REG_WEIGHTS = (10.0, 10.0, 5.0, 5.0)
+# The transition point from L1 to L2 loss. Set to 0.0 to make the loss simply L1.
+_C.MODEL.ROI_BOX_HEAD.SMOOTH_L1_BETA = 0.0
+_C.MODEL.ROI_BOX_HEAD.POOLER_RESOLUTION = 14
+_C.MODEL.ROI_BOX_HEAD.POOLER_SAMPLING_RATIO = 0
+# Type of pooling operation applied to the incoming feature map for each RoI
+_C.MODEL.ROI_BOX_HEAD.POOLER_TYPE = "ROIAlignV2"
+
+_C.MODEL.ROI_BOX_HEAD.NUM_FC = 0
+# Hidden layer dimension for FC layers in the RoI box head
+_C.MODEL.ROI_BOX_HEAD.FC_DIM = 1024
+_C.MODEL.ROI_BOX_HEAD.NUM_CONV = 0
+# Channel dimension for Conv layers in the RoI box head
+_C.MODEL.ROI_BOX_HEAD.CONV_DIM = 256
+# Normalization method for the convolution layers.
+# Options: "" (no norm), "GN", "SyncBN".
+_C.MODEL.ROI_BOX_HEAD.NORM = ""
+# Whether to use class agnostic for bbox regression
+_C.MODEL.ROI_BOX_HEAD.CLS_AGNOSTIC_BBOX_REG = False
+# If true, RoI heads use bounding boxes predicted by the box head rather than proposal boxes.
+_C.MODEL.ROI_BOX_HEAD.TRAIN_ON_PRED_BOXES = False
+
+# Federated loss can be used to improve the training of LVIS
+_C.MODEL.ROI_BOX_HEAD.USE_FED_LOSS = False
+# Sigmoid cross entrophy is used with federated loss
+_C.MODEL.ROI_BOX_HEAD.USE_SIGMOID_CE = False
+# The power value applied to image_count when calcualting frequency weight
+_C.MODEL.ROI_BOX_HEAD.FED_LOSS_FREQ_WEIGHT_POWER = 0.5
+# Number of classes to keep in total
+_C.MODEL.ROI_BOX_HEAD.FED_LOSS_NUM_CLASSES = 50
+
+# ---------------------------------------------------------------------------- #
+# Cascaded Box Head
+# ---------------------------------------------------------------------------- #
+_C.MODEL.ROI_BOX_CASCADE_HEAD = CN()
+# The number of cascade stages is implicitly defined by the length of the following two configs.
+_C.MODEL.ROI_BOX_CASCADE_HEAD.BBOX_REG_WEIGHTS = (
+ (10.0, 10.0, 5.0, 5.0),
+ (20.0, 20.0, 10.0, 10.0),
+ (30.0, 30.0, 15.0, 15.0),
+)
+_C.MODEL.ROI_BOX_CASCADE_HEAD.IOUS = (0.5, 0.6, 0.7)
+
+
+# ---------------------------------------------------------------------------- #
+# Mask Head
+# ---------------------------------------------------------------------------- #
+_C.MODEL.ROI_MASK_HEAD = CN()
+_C.MODEL.ROI_MASK_HEAD.NAME = "MaskRCNNConvUpsampleHead"
+_C.MODEL.ROI_MASK_HEAD.POOLER_RESOLUTION = 14
+_C.MODEL.ROI_MASK_HEAD.POOLER_SAMPLING_RATIO = 0
+_C.MODEL.ROI_MASK_HEAD.NUM_CONV = 0 # The number of convs in the mask head
+_C.MODEL.ROI_MASK_HEAD.CONV_DIM = 256
+# Normalization method for the convolution layers.
+# Options: "" (no norm), "GN", "SyncBN".
+_C.MODEL.ROI_MASK_HEAD.NORM = ""
+# Whether to use class agnostic for mask prediction
+_C.MODEL.ROI_MASK_HEAD.CLS_AGNOSTIC_MASK = False
+# Type of pooling operation applied to the incoming feature map for each RoI
+_C.MODEL.ROI_MASK_HEAD.POOLER_TYPE = "ROIAlignV2"
+
+
+# ---------------------------------------------------------------------------- #
+# Keypoint Head
+# ---------------------------------------------------------------------------- #
+_C.MODEL.ROI_KEYPOINT_HEAD = CN()
+_C.MODEL.ROI_KEYPOINT_HEAD.NAME = "KRCNNConvDeconvUpsampleHead"
+_C.MODEL.ROI_KEYPOINT_HEAD.POOLER_RESOLUTION = 14
+_C.MODEL.ROI_KEYPOINT_HEAD.POOLER_SAMPLING_RATIO = 0
+_C.MODEL.ROI_KEYPOINT_HEAD.CONV_DIMS = tuple(512 for _ in range(8))
+_C.MODEL.ROI_KEYPOINT_HEAD.NUM_KEYPOINTS = 17 # 17 is the number of keypoints in COCO.
+
+# Images with too few (or no) keypoints are excluded from training.
+_C.MODEL.ROI_KEYPOINT_HEAD.MIN_KEYPOINTS_PER_IMAGE = 1
+# Normalize by the total number of visible keypoints in the minibatch if True.
+# Otherwise, normalize by the total number of keypoints that could ever exist
+# in the minibatch.
+# The keypoint softmax loss is only calculated on visible keypoints.
+# Since the number of visible keypoints can vary significantly between
+# minibatches, this has the effect of up-weighting the importance of
+# minibatches with few visible keypoints. (Imagine the extreme case of
+# only one visible keypoint versus N: in the case of N, each one
+# contributes 1/N to the gradient compared to the single keypoint
+# determining the gradient direction). Instead, we can normalize the
+# loss by the total number of keypoints, if it were the case that all
+# keypoints were visible in a full minibatch. (Returning to the example,
+# this means that the one visible keypoint contributes as much as each
+# of the N keypoints.)
+_C.MODEL.ROI_KEYPOINT_HEAD.NORMALIZE_LOSS_BY_VISIBLE_KEYPOINTS = True
+# Multi-task loss weight to use for keypoints
+# Recommended values:
+# - use 1.0 if NORMALIZE_LOSS_BY_VISIBLE_KEYPOINTS is True
+# - use 4.0 if NORMALIZE_LOSS_BY_VISIBLE_KEYPOINTS is False
+_C.MODEL.ROI_KEYPOINT_HEAD.LOSS_WEIGHT = 1.0
+# Type of pooling operation applied to the incoming feature map for each RoI
+_C.MODEL.ROI_KEYPOINT_HEAD.POOLER_TYPE = "ROIAlignV2"
+
+# ---------------------------------------------------------------------------- #
+# Semantic Segmentation Head
+# ---------------------------------------------------------------------------- #
+_C.MODEL.SEM_SEG_HEAD = CN()
+_C.MODEL.SEM_SEG_HEAD.NAME = "SemSegFPNHead"
+_C.MODEL.SEM_SEG_HEAD.IN_FEATURES = ["p2", "p3", "p4", "p5"]
+# Label in the semantic segmentation ground truth that is ignored, i.e., no loss is calculated for
+# the correposnding pixel.
+_C.MODEL.SEM_SEG_HEAD.IGNORE_VALUE = 255
+# Number of classes in the semantic segmentation head
+_C.MODEL.SEM_SEG_HEAD.NUM_CLASSES = 54
+# Number of channels in the 3x3 convs inside semantic-FPN heads.
+_C.MODEL.SEM_SEG_HEAD.CONVS_DIM = 128
+# Outputs from semantic-FPN heads are up-scaled to the COMMON_STRIDE stride.
+_C.MODEL.SEM_SEG_HEAD.COMMON_STRIDE = 4
+# Normalization method for the convolution layers. Options: "" (no norm), "GN".
+_C.MODEL.SEM_SEG_HEAD.NORM = "GN"
+_C.MODEL.SEM_SEG_HEAD.LOSS_WEIGHT = 1.0
+
+_C.MODEL.PANOPTIC_FPN = CN()
+# Scaling of all losses from instance detection / segmentation head.
+_C.MODEL.PANOPTIC_FPN.INSTANCE_LOSS_WEIGHT = 1.0
+
+# options when combining instance & semantic segmentation outputs
+_C.MODEL.PANOPTIC_FPN.COMBINE = CN({"ENABLED": True}) # "COMBINE.ENABLED" is deprecated & not used
+_C.MODEL.PANOPTIC_FPN.COMBINE.OVERLAP_THRESH = 0.5
+_C.MODEL.PANOPTIC_FPN.COMBINE.STUFF_AREA_LIMIT = 4096
+_C.MODEL.PANOPTIC_FPN.COMBINE.INSTANCES_CONFIDENCE_THRESH = 0.5
+
+
+# ---------------------------------------------------------------------------- #
+# RetinaNet Head
+# ---------------------------------------------------------------------------- #
+_C.MODEL.RETINANET = CN()
+
+# This is the number of foreground classes.
+_C.MODEL.RETINANET.NUM_CLASSES = 80
+
+_C.MODEL.RETINANET.IN_FEATURES = ["p3", "p4", "p5", "p6", "p7"]
+
+# Convolutions to use in the cls and bbox tower
+# NOTE: this doesn't include the last conv for logits
+_C.MODEL.RETINANET.NUM_CONVS = 4
+
+# IoU overlap ratio [bg, fg] for labeling anchors.
+# Anchors with < bg are labeled negative (0)
+# Anchors with >= bg and < fg are ignored (-1)
+# Anchors with >= fg are labeled positive (1)
+_C.MODEL.RETINANET.IOU_THRESHOLDS = [0.4, 0.5]
+_C.MODEL.RETINANET.IOU_LABELS = [0, -1, 1]
+
+# Prior prob for rare case (i.e. foreground) at the beginning of training.
+# This is used to set the bias for the logits layer of the classifier subnet.
+# This improves training stability in the case of heavy class imbalance.
+_C.MODEL.RETINANET.PRIOR_PROB = 0.01
+
+# Inference cls score threshold, only anchors with score > INFERENCE_TH are
+# considered for inference (to improve speed)
+_C.MODEL.RETINANET.SCORE_THRESH_TEST = 0.05
+# Select topk candidates before NMS
+_C.MODEL.RETINANET.TOPK_CANDIDATES_TEST = 1000
+_C.MODEL.RETINANET.NMS_THRESH_TEST = 0.5
+
+# Weights on (dx, dy, dw, dh) for normalizing Retinanet anchor regression targets
+_C.MODEL.RETINANET.BBOX_REG_WEIGHTS = (1.0, 1.0, 1.0, 1.0)
+
+# Loss parameters
+_C.MODEL.RETINANET.FOCAL_LOSS_GAMMA = 2.0
+_C.MODEL.RETINANET.FOCAL_LOSS_ALPHA = 0.25
+_C.MODEL.RETINANET.SMOOTH_L1_LOSS_BETA = 0.1
+# Options are: "smooth_l1", "giou", "diou", "ciou"
+_C.MODEL.RETINANET.BBOX_REG_LOSS_TYPE = "smooth_l1"
+
+# One of BN, SyncBN, FrozenBN, GN
+# Only supports GN until unshared norm is implemented
+_C.MODEL.RETINANET.NORM = ""
+
+
+# ---------------------------------------------------------------------------- #
+# ResNe[X]t options (ResNets = {ResNet, ResNeXt}
+# Note that parts of a resnet may be used for both the backbone and the head
+# These options apply to both
+# ---------------------------------------------------------------------------- #
+_C.MODEL.RESNETS = CN()
+
+_C.MODEL.RESNETS.DEPTH = 50
+_C.MODEL.RESNETS.OUT_FEATURES = ["res4"] # res4 for C4 backbone, res2..5 for FPN backbone
+
+# Number of groups to use; 1 ==> ResNet; > 1 ==> ResNeXt
+_C.MODEL.RESNETS.NUM_GROUPS = 1
+
+# Options: FrozenBN, GN, "SyncBN", "BN"
+_C.MODEL.RESNETS.NORM = "FrozenBN"
+
+# Baseline width of each group.
+# Scaling this parameters will scale the width of all bottleneck layers.
+_C.MODEL.RESNETS.WIDTH_PER_GROUP = 64
+
+# Place the stride 2 conv on the 1x1 filter
+# Use True only for the original MSRA ResNet; use False for C2 and Torch models
+_C.MODEL.RESNETS.STRIDE_IN_1X1 = True
+
+# Apply dilation in stage "res5"
+_C.MODEL.RESNETS.RES5_DILATION = 1
+
+# Output width of res2. Scaling this parameters will scale the width of all 1x1 convs in ResNet
+# For R18 and R34, this needs to be set to 64
+_C.MODEL.RESNETS.RES2_OUT_CHANNELS = 256
+_C.MODEL.RESNETS.STEM_OUT_CHANNELS = 64
+
+# Apply Deformable Convolution in stages
+# Specify if apply deform_conv on Res2, Res3, Res4, Res5
+_C.MODEL.RESNETS.DEFORM_ON_PER_STAGE = [False, False, False, False]
+# Use True to use modulated deform_conv (DeformableV2, https://arxiv.org/abs/1811.11168);
+# Use False for DeformableV1.
+_C.MODEL.RESNETS.DEFORM_MODULATED = False
+# Number of groups in deformable conv.
+_C.MODEL.RESNETS.DEFORM_NUM_GROUPS = 1
+
+
+# ---------------------------------------------------------------------------- #
+# Solver
+# ---------------------------------------------------------------------------- #
+_C.SOLVER = CN()
+
+# Options: WarmupMultiStepLR, WarmupCosineLR.
+# See detectron2/solver/build.py for definition.
+_C.SOLVER.LR_SCHEDULER_NAME = "WarmupMultiStepLR"
+
+_C.SOLVER.MAX_ITER = 40000
+
+_C.SOLVER.BASE_LR = 0.001
+# The end lr, only used by WarmupCosineLR
+_C.SOLVER.BASE_LR_END = 0.0
+
+_C.SOLVER.MOMENTUM = 0.9
+
+_C.SOLVER.NESTEROV = False
+
+_C.SOLVER.WEIGHT_DECAY = 0.0001
+# The weight decay that's applied to parameters of normalization layers
+# (typically the affine transformation)
+_C.SOLVER.WEIGHT_DECAY_NORM = 0.0
+
+_C.SOLVER.GAMMA = 0.1
+# The iteration number to decrease learning rate by GAMMA.
+_C.SOLVER.STEPS = (30000,)
+# Number of decays in WarmupStepWithFixedGammaLR schedule
+_C.SOLVER.NUM_DECAYS = 3
+
+_C.SOLVER.WARMUP_FACTOR = 1.0 / 1000
+_C.SOLVER.WARMUP_ITERS = 1000
+_C.SOLVER.WARMUP_METHOD = "linear"
+# Whether to rescale the interval for the learning schedule after warmup
+_C.SOLVER.RESCALE_INTERVAL = False
+
+# Save a checkpoint after every this number of iterations
+_C.SOLVER.CHECKPOINT_PERIOD = 5000
+
+# Number of images per batch across all machines. This is also the number
+# of training images per step (i.e. per iteration). If we use 16 GPUs
+# and IMS_PER_BATCH = 32, each GPU will see 2 images per batch.
+# May be adjusted automatically if REFERENCE_WORLD_SIZE is set.
+_C.SOLVER.IMS_PER_BATCH = 16
+
+# The reference number of workers (GPUs) this config is meant to train with.
+# It takes no effect when set to 0.
+# With a non-zero value, it will be used by DefaultTrainer to compute a desired
+# per-worker batch size, and then scale the other related configs (total batch size,
+# learning rate, etc) to match the per-worker batch size.
+# See documentation of `DefaultTrainer.auto_scale_workers` for details:
+_C.SOLVER.REFERENCE_WORLD_SIZE = 0
+
+# Detectron v1 (and previous detection code) used a 2x higher LR and 0 WD for
+# biases. This is not useful (at least for recent models). You should avoid
+# changing these and they exist only to reproduce Detectron v1 training if
+# desired.
+_C.SOLVER.BIAS_LR_FACTOR = 1.0
+_C.SOLVER.WEIGHT_DECAY_BIAS = None # None means following WEIGHT_DECAY
+
+# Gradient clipping
+_C.SOLVER.CLIP_GRADIENTS = CN({"ENABLED": False})
+# Type of gradient clipping, currently 2 values are supported:
+# - "value": the absolute values of elements of each gradients are clipped
+# - "norm": the norm of the gradient for each parameter is clipped thus
+# affecting all elements in the parameter
+_C.SOLVER.CLIP_GRADIENTS.CLIP_TYPE = "value"
+# Maximum absolute value used for clipping gradients
+_C.SOLVER.CLIP_GRADIENTS.CLIP_VALUE = 1.0
+# Floating point number p for L-p norm to be used with the "norm"
+# gradient clipping type; for L-inf, please specify .inf
+_C.SOLVER.CLIP_GRADIENTS.NORM_TYPE = 2.0
+
+# Enable automatic mixed precision for training
+# Note that this does not change model's inference behavior.
+# To use AMP in inference, run inference under autocast()
+_C.SOLVER.AMP = CN({"ENABLED": False})
+
+# ---------------------------------------------------------------------------- #
+# Specific test options
+# ---------------------------------------------------------------------------- #
+_C.TEST = CN()
+# For end-to-end tests to verify the expected accuracy.
+# Each item is [task, metric, value, tolerance]
+# e.g.: [['bbox', 'AP', 38.5, 0.2]]
+_C.TEST.EXPECTED_RESULTS = []
+# The period (in terms of steps) to evaluate the model during training.
+# Set to 0 to disable.
+_C.TEST.EVAL_PERIOD = 0
+# The sigmas used to calculate keypoint OKS. See http://cocodataset.org/#keypoints-eval
+# When empty, it will use the defaults in COCO.
+# Otherwise it should be a list[float] with the same length as ROI_KEYPOINT_HEAD.NUM_KEYPOINTS.
+_C.TEST.KEYPOINT_OKS_SIGMAS = []
+# Maximum number of detections to return per image during inference (100 is
+# based on the limit established for the COCO dataset).
+_C.TEST.DETECTIONS_PER_IMAGE = 100
+
+_C.TEST.AUG = CN({"ENABLED": False})
+_C.TEST.AUG.MIN_SIZES = (400, 500, 600, 700, 800, 900, 1000, 1100, 1200)
+_C.TEST.AUG.MAX_SIZE = 4000
+_C.TEST.AUG.FLIP = True
+
+_C.TEST.PRECISE_BN = CN({"ENABLED": False})
+_C.TEST.PRECISE_BN.NUM_ITER = 200
+
+# ---------------------------------------------------------------------------- #
+# Misc options
+# ---------------------------------------------------------------------------- #
+# Directory where output files are written
+_C.OUTPUT_DIR = "./output"
+# Set seed to negative to fully randomize everything.
+# Set seed to positive to use a fixed seed. Note that a fixed seed increases
+# reproducibility but does not guarantee fully deterministic behavior.
+# Disabling all parallelism further increases reproducibility.
+_C.SEED = -1
+# Benchmark different cudnn algorithms.
+# If input images have very different sizes, this option will have large overhead
+# for about 10k iterations. It usually hurts total time, but can benefit for certain models.
+# If input images have the same or similar sizes, benchmark is often helpful.
+_C.CUDNN_BENCHMARK = False
+# The period (in terms of steps) for minibatch visualization at train time.
+# Set to 0 to disable.
+_C.VIS_PERIOD = 0
+
+# global config is for quick hack purposes.
+# You can set them in command line or config files,
+# and access it with:
+#
+# from annotator.oneformer.detectron2.config import global_cfg
+# print(global_cfg.HACK)
+#
+# Do not commit any configs into it.
+_C.GLOBAL = CN()
+_C.GLOBAL.HACK = 1.0
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/config/instantiate.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/config/instantiate.py
new file mode 100644
index 0000000000000000000000000000000000000000..26d191b03f800dae5620128957d137cd4fdb1728
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/config/instantiate.py
@@ -0,0 +1,88 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+
+import collections.abc as abc
+import dataclasses
+import logging
+from typing import Any
+
+from annotator.oneformer.detectron2.utils.registry import _convert_target_to_string, locate
+
+__all__ = ["dump_dataclass", "instantiate"]
+
+
+def dump_dataclass(obj: Any):
+ """
+ Dump a dataclass recursively into a dict that can be later instantiated.
+
+ Args:
+ obj: a dataclass object
+
+ Returns:
+ dict
+ """
+ assert dataclasses.is_dataclass(obj) and not isinstance(
+ obj, type
+ ), "dump_dataclass() requires an instance of a dataclass."
+ ret = {"_target_": _convert_target_to_string(type(obj))}
+ for f in dataclasses.fields(obj):
+ v = getattr(obj, f.name)
+ if dataclasses.is_dataclass(v):
+ v = dump_dataclass(v)
+ if isinstance(v, (list, tuple)):
+ v = [dump_dataclass(x) if dataclasses.is_dataclass(x) else x for x in v]
+ ret[f.name] = v
+ return ret
+
+
+def instantiate(cfg):
+ """
+ Recursively instantiate objects defined in dictionaries by
+ "_target_" and arguments.
+
+ Args:
+ cfg: a dict-like object with "_target_" that defines the caller, and
+ other keys that define the arguments
+
+ Returns:
+ object instantiated by cfg
+ """
+ from omegaconf import ListConfig, DictConfig, OmegaConf
+
+ if isinstance(cfg, ListConfig):
+ lst = [instantiate(x) for x in cfg]
+ return ListConfig(lst, flags={"allow_objects": True})
+ if isinstance(cfg, list):
+ # Specialize for list, because many classes take
+ # list[objects] as arguments, such as ResNet, DatasetMapper
+ return [instantiate(x) for x in cfg]
+
+ # If input is a DictConfig backed by dataclasses (i.e. omegaconf's structured config),
+ # instantiate it to the actual dataclass.
+ if isinstance(cfg, DictConfig) and dataclasses.is_dataclass(cfg._metadata.object_type):
+ return OmegaConf.to_object(cfg)
+
+ if isinstance(cfg, abc.Mapping) and "_target_" in cfg:
+ # conceptually equivalent to hydra.utils.instantiate(cfg) with _convert_=all,
+ # but faster: https://github.com/facebookresearch/hydra/issues/1200
+ cfg = {k: instantiate(v) for k, v in cfg.items()}
+ cls = cfg.pop("_target_")
+ cls = instantiate(cls)
+
+ if isinstance(cls, str):
+ cls_name = cls
+ cls = locate(cls_name)
+ assert cls is not None, cls_name
+ else:
+ try:
+ cls_name = cls.__module__ + "." + cls.__qualname__
+ except Exception:
+ # target could be anything, so the above could fail
+ cls_name = str(cls)
+ assert callable(cls), f"_target_ {cls} does not define a callable object"
+ try:
+ return cls(**cfg)
+ except TypeError:
+ logger = logging.getLogger(__name__)
+ logger.error(f"Error when instantiating {cls_name}!")
+ raise
+ return cfg # return as-is if don't know what to do
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/config/lazy.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/config/lazy.py
new file mode 100644
index 0000000000000000000000000000000000000000..72a3e5c036f9f78a2cdf3ef0975639da3299d694
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/config/lazy.py
@@ -0,0 +1,435 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+
+import ast
+import builtins
+import collections.abc as abc
+import importlib
+import inspect
+import logging
+import os
+import uuid
+from contextlib import contextmanager
+from copy import deepcopy
+from dataclasses import is_dataclass
+from typing import List, Tuple, Union
+import yaml
+from omegaconf import DictConfig, ListConfig, OmegaConf, SCMode
+
+from annotator.oneformer.detectron2.utils.file_io import PathManager
+from annotator.oneformer.detectron2.utils.registry import _convert_target_to_string
+
+__all__ = ["LazyCall", "LazyConfig"]
+
+
+class LazyCall:
+ """
+ Wrap a callable so that when it's called, the call will not be executed,
+ but returns a dict that describes the call.
+
+ LazyCall object has to be called with only keyword arguments. Positional
+ arguments are not yet supported.
+
+ Examples:
+ ::
+ from annotator.oneformer.detectron2.config import instantiate, LazyCall
+
+ layer_cfg = LazyCall(nn.Conv2d)(in_channels=32, out_channels=32)
+ layer_cfg.out_channels = 64 # can edit it afterwards
+ layer = instantiate(layer_cfg)
+ """
+
+ def __init__(self, target):
+ if not (callable(target) or isinstance(target, (str, abc.Mapping))):
+ raise TypeError(
+ f"target of LazyCall must be a callable or defines a callable! Got {target}"
+ )
+ self._target = target
+
+ def __call__(self, **kwargs):
+ if is_dataclass(self._target):
+ # omegaconf object cannot hold dataclass type
+ # https://github.com/omry/omegaconf/issues/784
+ target = _convert_target_to_string(self._target)
+ else:
+ target = self._target
+ kwargs["_target_"] = target
+
+ return DictConfig(content=kwargs, flags={"allow_objects": True})
+
+
+def _visit_dict_config(cfg, func):
+ """
+ Apply func recursively to all DictConfig in cfg.
+ """
+ if isinstance(cfg, DictConfig):
+ func(cfg)
+ for v in cfg.values():
+ _visit_dict_config(v, func)
+ elif isinstance(cfg, ListConfig):
+ for v in cfg:
+ _visit_dict_config(v, func)
+
+
+def _validate_py_syntax(filename):
+ # see also https://github.com/open-mmlab/mmcv/blob/master/mmcv/utils/config.py
+ with PathManager.open(filename, "r") as f:
+ content = f.read()
+ try:
+ ast.parse(content)
+ except SyntaxError as e:
+ raise SyntaxError(f"Config file {filename} has syntax error!") from e
+
+
+def _cast_to_config(obj):
+ # if given a dict, return DictConfig instead
+ if isinstance(obj, dict):
+ return DictConfig(obj, flags={"allow_objects": True})
+ return obj
+
+
+_CFG_PACKAGE_NAME = "detectron2._cfg_loader"
+"""
+A namespace to put all imported config into.
+"""
+
+
+def _random_package_name(filename):
+ # generate a random package name when loading config files
+ return _CFG_PACKAGE_NAME + str(uuid.uuid4())[:4] + "." + os.path.basename(filename)
+
+
+@contextmanager
+def _patch_import():
+ """
+ Enhance relative import statements in config files, so that they:
+ 1. locate files purely based on relative location, regardless of packages.
+ e.g. you can import file without having __init__
+ 2. do not cache modules globally; modifications of module states has no side effect
+ 3. support other storage system through PathManager, so config files can be in the cloud
+ 4. imported dict are turned into omegaconf.DictConfig automatically
+ """
+ old_import = builtins.__import__
+
+ def find_relative_file(original_file, relative_import_path, level):
+ # NOTE: "from . import x" is not handled. Because then it's unclear
+ # if such import should produce `x` as a python module or DictConfig.
+ # This can be discussed further if needed.
+ relative_import_err = """
+Relative import of directories is not allowed within config files.
+Within a config file, relative import can only import other config files.
+""".replace(
+ "\n", " "
+ )
+ if not len(relative_import_path):
+ raise ImportError(relative_import_err)
+
+ cur_file = os.path.dirname(original_file)
+ for _ in range(level - 1):
+ cur_file = os.path.dirname(cur_file)
+ cur_name = relative_import_path.lstrip(".")
+ for part in cur_name.split("."):
+ cur_file = os.path.join(cur_file, part)
+ if not cur_file.endswith(".py"):
+ cur_file += ".py"
+ if not PathManager.isfile(cur_file):
+ cur_file_no_suffix = cur_file[: -len(".py")]
+ if PathManager.isdir(cur_file_no_suffix):
+ raise ImportError(f"Cannot import from {cur_file_no_suffix}." + relative_import_err)
+ else:
+ raise ImportError(
+ f"Cannot import name {relative_import_path} from "
+ f"{original_file}: {cur_file} does not exist."
+ )
+ return cur_file
+
+ def new_import(name, globals=None, locals=None, fromlist=(), level=0):
+ if (
+ # Only deal with relative imports inside config files
+ level != 0
+ and globals is not None
+ and (globals.get("__package__", "") or "").startswith(_CFG_PACKAGE_NAME)
+ ):
+ cur_file = find_relative_file(globals["__file__"], name, level)
+ _validate_py_syntax(cur_file)
+ spec = importlib.machinery.ModuleSpec(
+ _random_package_name(cur_file), None, origin=cur_file
+ )
+ module = importlib.util.module_from_spec(spec)
+ module.__file__ = cur_file
+ with PathManager.open(cur_file) as f:
+ content = f.read()
+ exec(compile(content, cur_file, "exec"), module.__dict__)
+ for name in fromlist: # turn imported dict into DictConfig automatically
+ val = _cast_to_config(module.__dict__[name])
+ module.__dict__[name] = val
+ return module
+ return old_import(name, globals, locals, fromlist=fromlist, level=level)
+
+ builtins.__import__ = new_import
+ yield new_import
+ builtins.__import__ = old_import
+
+
+class LazyConfig:
+ """
+ Provide methods to save, load, and overrides an omegaconf config object
+ which may contain definition of lazily-constructed objects.
+ """
+
+ @staticmethod
+ def load_rel(filename: str, keys: Union[None, str, Tuple[str, ...]] = None):
+ """
+ Similar to :meth:`load()`, but load path relative to the caller's
+ source file.
+
+ This has the same functionality as a relative import, except that this method
+ accepts filename as a string, so more characters are allowed in the filename.
+ """
+ caller_frame = inspect.stack()[1]
+ caller_fname = caller_frame[0].f_code.co_filename
+ assert caller_fname != "", "load_rel Unable to find caller"
+ caller_dir = os.path.dirname(caller_fname)
+ filename = os.path.join(caller_dir, filename)
+ return LazyConfig.load(filename, keys)
+
+ @staticmethod
+ def load(filename: str, keys: Union[None, str, Tuple[str, ...]] = None):
+ """
+ Load a config file.
+
+ Args:
+ filename: absolute path or relative path w.r.t. the current working directory
+ keys: keys to load and return. If not given, return all keys
+ (whose values are config objects) in a dict.
+ """
+ has_keys = keys is not None
+ filename = filename.replace("/./", "/") # redundant
+ if os.path.splitext(filename)[1] not in [".py", ".yaml", ".yml"]:
+ raise ValueError(f"Config file {filename} has to be a python or yaml file.")
+ if filename.endswith(".py"):
+ _validate_py_syntax(filename)
+
+ with _patch_import():
+ # Record the filename
+ module_namespace = {
+ "__file__": filename,
+ "__package__": _random_package_name(filename),
+ }
+ with PathManager.open(filename) as f:
+ content = f.read()
+ # Compile first with filename to:
+ # 1. make filename appears in stacktrace
+ # 2. make load_rel able to find its parent's (possibly remote) location
+ exec(compile(content, filename, "exec"), module_namespace)
+
+ ret = module_namespace
+ else:
+ with PathManager.open(filename) as f:
+ obj = yaml.unsafe_load(f)
+ ret = OmegaConf.create(obj, flags={"allow_objects": True})
+
+ if has_keys:
+ if isinstance(keys, str):
+ return _cast_to_config(ret[keys])
+ else:
+ return tuple(_cast_to_config(ret[a]) for a in keys)
+ else:
+ if filename.endswith(".py"):
+ # when not specified, only load those that are config objects
+ ret = DictConfig(
+ {
+ name: _cast_to_config(value)
+ for name, value in ret.items()
+ if isinstance(value, (DictConfig, ListConfig, dict))
+ and not name.startswith("_")
+ },
+ flags={"allow_objects": True},
+ )
+ return ret
+
+ @staticmethod
+ def save(cfg, filename: str):
+ """
+ Save a config object to a yaml file.
+ Note that when the config dictionary contains complex objects (e.g. lambda),
+ it can't be saved to yaml. In that case we will print an error and
+ attempt to save to a pkl file instead.
+
+ Args:
+ cfg: an omegaconf config object
+ filename: yaml file name to save the config file
+ """
+ logger = logging.getLogger(__name__)
+ try:
+ cfg = deepcopy(cfg)
+ except Exception:
+ pass
+ else:
+ # if it's deep-copyable, then...
+ def _replace_type_by_name(x):
+ if "_target_" in x and callable(x._target_):
+ try:
+ x._target_ = _convert_target_to_string(x._target_)
+ except AttributeError:
+ pass
+
+ # not necessary, but makes yaml looks nicer
+ _visit_dict_config(cfg, _replace_type_by_name)
+
+ save_pkl = False
+ try:
+ dict = OmegaConf.to_container(
+ cfg,
+ # Do not resolve interpolation when saving, i.e. do not turn ${a} into
+ # actual values when saving.
+ resolve=False,
+ # Save structures (dataclasses) in a format that can be instantiated later.
+ # Without this option, the type information of the dataclass will be erased.
+ structured_config_mode=SCMode.INSTANTIATE,
+ )
+ dumped = yaml.dump(dict, default_flow_style=None, allow_unicode=True, width=9999)
+ with PathManager.open(filename, "w") as f:
+ f.write(dumped)
+
+ try:
+ _ = yaml.unsafe_load(dumped) # test that it is loadable
+ except Exception:
+ logger.warning(
+ "The config contains objects that cannot serialize to a valid yaml. "
+ f"{filename} is human-readable but cannot be loaded."
+ )
+ save_pkl = True
+ except Exception:
+ logger.exception("Unable to serialize the config to yaml. Error:")
+ save_pkl = True
+
+ if save_pkl:
+ new_filename = filename + ".pkl"
+ # try:
+ # # retry by pickle
+ # with PathManager.open(new_filename, "wb") as f:
+ # cloudpickle.dump(cfg, f)
+ # logger.warning(f"Config is saved using cloudpickle at {new_filename}.")
+ # except Exception:
+ # pass
+
+ @staticmethod
+ def apply_overrides(cfg, overrides: List[str]):
+ """
+ In-place override contents of cfg.
+
+ Args:
+ cfg: an omegaconf config object
+ overrides: list of strings in the format of "a=b" to override configs.
+ See https://hydra.cc/docs/next/advanced/override_grammar/basic/
+ for syntax.
+
+ Returns:
+ the cfg object
+ """
+
+ def safe_update(cfg, key, value):
+ parts = key.split(".")
+ for idx in range(1, len(parts)):
+ prefix = ".".join(parts[:idx])
+ v = OmegaConf.select(cfg, prefix, default=None)
+ if v is None:
+ break
+ if not OmegaConf.is_config(v):
+ raise KeyError(
+ f"Trying to update key {key}, but {prefix} "
+ f"is not a config, but has type {type(v)}."
+ )
+ OmegaConf.update(cfg, key, value, merge=True)
+
+ try:
+ from hydra.core.override_parser.overrides_parser import OverridesParser
+
+ has_hydra = True
+ except ImportError:
+ has_hydra = False
+
+ if has_hydra:
+ parser = OverridesParser.create()
+ overrides = parser.parse_overrides(overrides)
+ for o in overrides:
+ key = o.key_or_group
+ value = o.value()
+ if o.is_delete():
+ # TODO support this
+ raise NotImplementedError("deletion is not yet a supported override")
+ safe_update(cfg, key, value)
+ else:
+ # Fallback. Does not support all the features and error checking like hydra.
+ for o in overrides:
+ key, value = o.split("=")
+ try:
+ value = eval(value, {})
+ except NameError:
+ pass
+ safe_update(cfg, key, value)
+ return cfg
+
+ # @staticmethod
+ # def to_py(cfg, prefix: str = "cfg."):
+ # """
+ # Try to convert a config object into Python-like psuedo code.
+ #
+ # Note that perfect conversion is not always possible. So the returned
+ # results are mainly meant to be human-readable, and not meant to be executed.
+ #
+ # Args:
+ # cfg: an omegaconf config object
+ # prefix: root name for the resulting code (default: "cfg.")
+ #
+ #
+ # Returns:
+ # str of formatted Python code
+ # """
+ # import black
+ #
+ # cfg = OmegaConf.to_container(cfg, resolve=True)
+ #
+ # def _to_str(obj, prefix=None, inside_call=False):
+ # if prefix is None:
+ # prefix = []
+ # if isinstance(obj, abc.Mapping) and "_target_" in obj:
+ # # Dict representing a function call
+ # target = _convert_target_to_string(obj.pop("_target_"))
+ # args = []
+ # for k, v in sorted(obj.items()):
+ # args.append(f"{k}={_to_str(v, inside_call=True)}")
+ # args = ", ".join(args)
+ # call = f"{target}({args})"
+ # return "".join(prefix) + call
+ # elif isinstance(obj, abc.Mapping) and not inside_call:
+ # # Dict that is not inside a call is a list of top-level config objects that we
+ # # render as one object per line with dot separated prefixes
+ # key_list = []
+ # for k, v in sorted(obj.items()):
+ # if isinstance(v, abc.Mapping) and "_target_" not in v:
+ # key_list.append(_to_str(v, prefix=prefix + [k + "."]))
+ # else:
+ # key = "".join(prefix) + k
+ # key_list.append(f"{key}={_to_str(v)}")
+ # return "\n".join(key_list)
+ # elif isinstance(obj, abc.Mapping):
+ # # Dict that is inside a call is rendered as a regular dict
+ # return (
+ # "{"
+ # + ",".join(
+ # f"{repr(k)}: {_to_str(v, inside_call=inside_call)}"
+ # for k, v in sorted(obj.items())
+ # )
+ # + "}"
+ # )
+ # elif isinstance(obj, list):
+ # return "[" + ",".join(_to_str(x, inside_call=inside_call) for x in obj) + "]"
+ # else:
+ # return repr(obj)
+ #
+ # py_str = _to_str(cfg, prefix=[prefix])
+ # try:
+ # return black.format_str(py_str, mode=black.Mode())
+ # except black.InvalidInput:
+ # return py_str
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/__init__.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..259f669b78bd05815cb8d3351fd6c5fc9a1b85a1
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/__init__.py
@@ -0,0 +1,19 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+from . import transforms # isort:skip
+
+from .build import (
+ build_batch_data_loader,
+ build_detection_test_loader,
+ build_detection_train_loader,
+ get_detection_dataset_dicts,
+ load_proposals_into_dataset,
+ print_instances_class_histogram,
+)
+from .catalog import DatasetCatalog, MetadataCatalog, Metadata
+from .common import DatasetFromList, MapDataset, ToIterableDataset
+from .dataset_mapper import DatasetMapper
+
+# ensure the builtin datasets are registered
+from . import datasets, samplers # isort:skip
+
+__all__ = [k for k in globals().keys() if not k.startswith("_")]
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/benchmark.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/benchmark.py
new file mode 100644
index 0000000000000000000000000000000000000000..bfd650582c83cd032b4fe76303517cdfd9a2a8b4
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/benchmark.py
@@ -0,0 +1,225 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import logging
+import numpy as np
+from itertools import count
+from typing import List, Tuple
+import torch
+import tqdm
+from fvcore.common.timer import Timer
+
+from annotator.oneformer.detectron2.utils import comm
+
+from .build import build_batch_data_loader
+from .common import DatasetFromList, MapDataset
+from .samplers import TrainingSampler
+
+logger = logging.getLogger(__name__)
+
+
+class _EmptyMapDataset(torch.utils.data.Dataset):
+ """
+ Map anything to emptiness.
+ """
+
+ def __init__(self, dataset):
+ self.ds = dataset
+
+ def __len__(self):
+ return len(self.ds)
+
+ def __getitem__(self, idx):
+ _ = self.ds[idx]
+ return [0]
+
+
+def iter_benchmark(
+ iterator, num_iter: int, warmup: int = 5, max_time_seconds: float = 60
+) -> Tuple[float, List[float]]:
+ """
+ Benchmark an iterator/iterable for `num_iter` iterations with an extra
+ `warmup` iterations of warmup.
+ End early if `max_time_seconds` time is spent on iterations.
+
+ Returns:
+ float: average time (seconds) per iteration
+ list[float]: time spent on each iteration. Sometimes useful for further analysis.
+ """
+ num_iter, warmup = int(num_iter), int(warmup)
+
+ iterator = iter(iterator)
+ for _ in range(warmup):
+ next(iterator)
+ timer = Timer()
+ all_times = []
+ for curr_iter in tqdm.trange(num_iter):
+ start = timer.seconds()
+ if start > max_time_seconds:
+ num_iter = curr_iter
+ break
+ next(iterator)
+ all_times.append(timer.seconds() - start)
+ avg = timer.seconds() / num_iter
+ return avg, all_times
+
+
+class DataLoaderBenchmark:
+ """
+ Some common benchmarks that help understand perf bottleneck of a standard dataloader
+ made of dataset, mapper and sampler.
+ """
+
+ def __init__(
+ self,
+ dataset,
+ *,
+ mapper,
+ sampler=None,
+ total_batch_size,
+ num_workers=0,
+ max_time_seconds: int = 90,
+ ):
+ """
+ Args:
+ max_time_seconds (int): maximum time to spent for each benchmark
+ other args: same as in `build.py:build_detection_train_loader`
+ """
+ if isinstance(dataset, list):
+ dataset = DatasetFromList(dataset, copy=False, serialize=True)
+ if sampler is None:
+ sampler = TrainingSampler(len(dataset))
+
+ self.dataset = dataset
+ self.mapper = mapper
+ self.sampler = sampler
+ self.total_batch_size = total_batch_size
+ self.num_workers = num_workers
+ self.per_gpu_batch_size = self.total_batch_size // comm.get_world_size()
+
+ self.max_time_seconds = max_time_seconds
+
+ def _benchmark(self, iterator, num_iter, warmup, msg=None):
+ avg, all_times = iter_benchmark(iterator, num_iter, warmup, self.max_time_seconds)
+ if msg is not None:
+ self._log_time(msg, avg, all_times)
+ return avg, all_times
+
+ def _log_time(self, msg, avg, all_times, distributed=False):
+ percentiles = [np.percentile(all_times, k, interpolation="nearest") for k in [1, 5, 95, 99]]
+ if not distributed:
+ logger.info(
+ f"{msg}: avg={1.0/avg:.1f} it/s, "
+ f"p1={percentiles[0]:.2g}s, p5={percentiles[1]:.2g}s, "
+ f"p95={percentiles[2]:.2g}s, p99={percentiles[3]:.2g}s."
+ )
+ return
+ avg_per_gpu = comm.all_gather(avg)
+ percentiles_per_gpu = comm.all_gather(percentiles)
+ if comm.get_rank() > 0:
+ return
+ for idx, avg, percentiles in zip(count(), avg_per_gpu, percentiles_per_gpu):
+ logger.info(
+ f"GPU{idx} {msg}: avg={1.0/avg:.1f} it/s, "
+ f"p1={percentiles[0]:.2g}s, p5={percentiles[1]:.2g}s, "
+ f"p95={percentiles[2]:.2g}s, p99={percentiles[3]:.2g}s."
+ )
+
+ def benchmark_dataset(self, num_iter, warmup=5):
+ """
+ Benchmark the speed of taking raw samples from the dataset.
+ """
+
+ def loader():
+ while True:
+ for k in self.sampler:
+ yield self.dataset[k]
+
+ self._benchmark(loader(), num_iter, warmup, "Dataset Alone")
+
+ def benchmark_mapper(self, num_iter, warmup=5):
+ """
+ Benchmark the speed of taking raw samples from the dataset and map
+ them in a single process.
+ """
+
+ def loader():
+ while True:
+ for k in self.sampler:
+ yield self.mapper(self.dataset[k])
+
+ self._benchmark(loader(), num_iter, warmup, "Single Process Mapper (sec/sample)")
+
+ def benchmark_workers(self, num_iter, warmup=10):
+ """
+ Benchmark the dataloader by tuning num_workers to [0, 1, self.num_workers].
+ """
+ candidates = [0, 1]
+ if self.num_workers not in candidates:
+ candidates.append(self.num_workers)
+
+ dataset = MapDataset(self.dataset, self.mapper)
+ for n in candidates:
+ loader = build_batch_data_loader(
+ dataset,
+ self.sampler,
+ self.total_batch_size,
+ num_workers=n,
+ )
+ self._benchmark(
+ iter(loader),
+ num_iter * max(n, 1),
+ warmup * max(n, 1),
+ f"DataLoader ({n} workers, bs={self.per_gpu_batch_size})",
+ )
+ del loader
+
+ def benchmark_IPC(self, num_iter, warmup=10):
+ """
+ Benchmark the dataloader where each worker outputs nothing. This
+ eliminates the IPC overhead compared to the regular dataloader.
+
+ PyTorch multiprocessing's IPC only optimizes for torch tensors.
+ Large numpy arrays or other data structure may incur large IPC overhead.
+ """
+ n = self.num_workers
+ dataset = _EmptyMapDataset(MapDataset(self.dataset, self.mapper))
+ loader = build_batch_data_loader(
+ dataset, self.sampler, self.total_batch_size, num_workers=n
+ )
+ self._benchmark(
+ iter(loader),
+ num_iter * max(n, 1),
+ warmup * max(n, 1),
+ f"DataLoader ({n} workers, bs={self.per_gpu_batch_size}) w/o comm",
+ )
+
+ def benchmark_distributed(self, num_iter, warmup=10):
+ """
+ Benchmark the dataloader in each distributed worker, and log results of
+ all workers. This helps understand the final performance as well as
+ the variances among workers.
+
+ It also prints startup time (first iter) of the dataloader.
+ """
+ gpu = comm.get_world_size()
+ dataset = MapDataset(self.dataset, self.mapper)
+ n = self.num_workers
+ loader = build_batch_data_loader(
+ dataset, self.sampler, self.total_batch_size, num_workers=n
+ )
+
+ timer = Timer()
+ loader = iter(loader)
+ next(loader)
+ startup_time = timer.seconds()
+ logger.info("Dataloader startup time: {:.2f} seconds".format(startup_time))
+
+ comm.synchronize()
+
+ avg, all_times = self._benchmark(loader, num_iter * max(n, 1), warmup * max(n, 1))
+ del loader
+ self._log_time(
+ f"DataLoader ({gpu} GPUs x {n} workers, total bs={self.total_batch_size})",
+ avg,
+ all_times,
+ True,
+ )
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/build.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/build.py
new file mode 100644
index 0000000000000000000000000000000000000000..d03137a9aabfc4a056dd671d4c3d0ba6f349fe03
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/build.py
@@ -0,0 +1,556 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import itertools
+import logging
+import numpy as np
+import operator
+import pickle
+from typing import Any, Callable, Dict, List, Optional, Union
+import torch
+import torch.utils.data as torchdata
+from tabulate import tabulate
+from termcolor import colored
+
+from annotator.oneformer.detectron2.config import configurable
+from annotator.oneformer.detectron2.structures import BoxMode
+from annotator.oneformer.detectron2.utils.comm import get_world_size
+from annotator.oneformer.detectron2.utils.env import seed_all_rng
+from annotator.oneformer.detectron2.utils.file_io import PathManager
+from annotator.oneformer.detectron2.utils.logger import _log_api_usage, log_first_n
+
+from .catalog import DatasetCatalog, MetadataCatalog
+from .common import AspectRatioGroupedDataset, DatasetFromList, MapDataset, ToIterableDataset
+from .dataset_mapper import DatasetMapper
+from .detection_utils import check_metadata_consistency
+from .samplers import (
+ InferenceSampler,
+ RandomSubsetTrainingSampler,
+ RepeatFactorTrainingSampler,
+ TrainingSampler,
+)
+
+"""
+This file contains the default logic to build a dataloader for training or testing.
+"""
+
+__all__ = [
+ "build_batch_data_loader",
+ "build_detection_train_loader",
+ "build_detection_test_loader",
+ "get_detection_dataset_dicts",
+ "load_proposals_into_dataset",
+ "print_instances_class_histogram",
+]
+
+
+def filter_images_with_only_crowd_annotations(dataset_dicts):
+ """
+ Filter out images with none annotations or only crowd annotations
+ (i.e., images without non-crowd annotations).
+ A common training-time preprocessing on COCO dataset.
+
+ Args:
+ dataset_dicts (list[dict]): annotations in Detectron2 Dataset format.
+
+ Returns:
+ list[dict]: the same format, but filtered.
+ """
+ num_before = len(dataset_dicts)
+
+ def valid(anns):
+ for ann in anns:
+ if ann.get("iscrowd", 0) == 0:
+ return True
+ return False
+
+ dataset_dicts = [x for x in dataset_dicts if valid(x["annotations"])]
+ num_after = len(dataset_dicts)
+ logger = logging.getLogger(__name__)
+ logger.info(
+ "Removed {} images with no usable annotations. {} images left.".format(
+ num_before - num_after, num_after
+ )
+ )
+ return dataset_dicts
+
+
+def filter_images_with_few_keypoints(dataset_dicts, min_keypoints_per_image):
+ """
+ Filter out images with too few number of keypoints.
+
+ Args:
+ dataset_dicts (list[dict]): annotations in Detectron2 Dataset format.
+
+ Returns:
+ list[dict]: the same format as dataset_dicts, but filtered.
+ """
+ num_before = len(dataset_dicts)
+
+ def visible_keypoints_in_image(dic):
+ # Each keypoints field has the format [x1, y1, v1, ...], where v is visibility
+ annotations = dic["annotations"]
+ return sum(
+ (np.array(ann["keypoints"][2::3]) > 0).sum()
+ for ann in annotations
+ if "keypoints" in ann
+ )
+
+ dataset_dicts = [
+ x for x in dataset_dicts if visible_keypoints_in_image(x) >= min_keypoints_per_image
+ ]
+ num_after = len(dataset_dicts)
+ logger = logging.getLogger(__name__)
+ logger.info(
+ "Removed {} images with fewer than {} keypoints.".format(
+ num_before - num_after, min_keypoints_per_image
+ )
+ )
+ return dataset_dicts
+
+
+def load_proposals_into_dataset(dataset_dicts, proposal_file):
+ """
+ Load precomputed object proposals into the dataset.
+
+ The proposal file should be a pickled dict with the following keys:
+
+ - "ids": list[int] or list[str], the image ids
+ - "boxes": list[np.ndarray], each is an Nx4 array of boxes corresponding to the image id
+ - "objectness_logits": list[np.ndarray], each is an N sized array of objectness scores
+ corresponding to the boxes.
+ - "bbox_mode": the BoxMode of the boxes array. Defaults to ``BoxMode.XYXY_ABS``.
+
+ Args:
+ dataset_dicts (list[dict]): annotations in Detectron2 Dataset format.
+ proposal_file (str): file path of pre-computed proposals, in pkl format.
+
+ Returns:
+ list[dict]: the same format as dataset_dicts, but added proposal field.
+ """
+ logger = logging.getLogger(__name__)
+ logger.info("Loading proposals from: {}".format(proposal_file))
+
+ with PathManager.open(proposal_file, "rb") as f:
+ proposals = pickle.load(f, encoding="latin1")
+
+ # Rename the key names in D1 proposal files
+ rename_keys = {"indexes": "ids", "scores": "objectness_logits"}
+ for key in rename_keys:
+ if key in proposals:
+ proposals[rename_keys[key]] = proposals.pop(key)
+
+ # Fetch the indexes of all proposals that are in the dataset
+ # Convert image_id to str since they could be int.
+ img_ids = set({str(record["image_id"]) for record in dataset_dicts})
+ id_to_index = {str(id): i for i, id in enumerate(proposals["ids"]) if str(id) in img_ids}
+
+ # Assuming default bbox_mode of precomputed proposals are 'XYXY_ABS'
+ bbox_mode = BoxMode(proposals["bbox_mode"]) if "bbox_mode" in proposals else BoxMode.XYXY_ABS
+
+ for record in dataset_dicts:
+ # Get the index of the proposal
+ i = id_to_index[str(record["image_id"])]
+
+ boxes = proposals["boxes"][i]
+ objectness_logits = proposals["objectness_logits"][i]
+ # Sort the proposals in descending order of the scores
+ inds = objectness_logits.argsort()[::-1]
+ record["proposal_boxes"] = boxes[inds]
+ record["proposal_objectness_logits"] = objectness_logits[inds]
+ record["proposal_bbox_mode"] = bbox_mode
+
+ return dataset_dicts
+
+
+def print_instances_class_histogram(dataset_dicts, class_names):
+ """
+ Args:
+ dataset_dicts (list[dict]): list of dataset dicts.
+ class_names (list[str]): list of class names (zero-indexed).
+ """
+ num_classes = len(class_names)
+ hist_bins = np.arange(num_classes + 1)
+ histogram = np.zeros((num_classes,), dtype=np.int)
+ for entry in dataset_dicts:
+ annos = entry["annotations"]
+ classes = np.asarray(
+ [x["category_id"] for x in annos if not x.get("iscrowd", 0)], dtype=np.int
+ )
+ if len(classes):
+ assert classes.min() >= 0, f"Got an invalid category_id={classes.min()}"
+ assert (
+ classes.max() < num_classes
+ ), f"Got an invalid category_id={classes.max()} for a dataset of {num_classes} classes"
+ histogram += np.histogram(classes, bins=hist_bins)[0]
+
+ N_COLS = min(6, len(class_names) * 2)
+
+ def short_name(x):
+ # make long class names shorter. useful for lvis
+ if len(x) > 13:
+ return x[:11] + ".."
+ return x
+
+ data = list(
+ itertools.chain(*[[short_name(class_names[i]), int(v)] for i, v in enumerate(histogram)])
+ )
+ total_num_instances = sum(data[1::2])
+ data.extend([None] * (N_COLS - (len(data) % N_COLS)))
+ if num_classes > 1:
+ data.extend(["total", total_num_instances])
+ data = itertools.zip_longest(*[data[i::N_COLS] for i in range(N_COLS)])
+ table = tabulate(
+ data,
+ headers=["category", "#instances"] * (N_COLS // 2),
+ tablefmt="pipe",
+ numalign="left",
+ stralign="center",
+ )
+ log_first_n(
+ logging.INFO,
+ "Distribution of instances among all {} categories:\n".format(num_classes)
+ + colored(table, "cyan"),
+ key="message",
+ )
+
+
+def get_detection_dataset_dicts(
+ names,
+ filter_empty=True,
+ min_keypoints=0,
+ proposal_files=None,
+ check_consistency=True,
+):
+ """
+ Load and prepare dataset dicts for instance detection/segmentation and semantic segmentation.
+
+ Args:
+ names (str or list[str]): a dataset name or a list of dataset names
+ filter_empty (bool): whether to filter out images without instance annotations
+ min_keypoints (int): filter out images with fewer keypoints than
+ `min_keypoints`. Set to 0 to do nothing.
+ proposal_files (list[str]): if given, a list of object proposal files
+ that match each dataset in `names`.
+ check_consistency (bool): whether to check if datasets have consistent metadata.
+
+ Returns:
+ list[dict]: a list of dicts following the standard dataset dict format.
+ """
+ if isinstance(names, str):
+ names = [names]
+ assert len(names), names
+ dataset_dicts = [DatasetCatalog.get(dataset_name) for dataset_name in names]
+
+ if isinstance(dataset_dicts[0], torchdata.Dataset):
+ if len(dataset_dicts) > 1:
+ # ConcatDataset does not work for iterable style dataset.
+ # We could support concat for iterable as well, but it's often
+ # not a good idea to concat iterables anyway.
+ return torchdata.ConcatDataset(dataset_dicts)
+ return dataset_dicts[0]
+
+ for dataset_name, dicts in zip(names, dataset_dicts):
+ assert len(dicts), "Dataset '{}' is empty!".format(dataset_name)
+
+ if proposal_files is not None:
+ assert len(names) == len(proposal_files)
+ # load precomputed proposals from proposal files
+ dataset_dicts = [
+ load_proposals_into_dataset(dataset_i_dicts, proposal_file)
+ for dataset_i_dicts, proposal_file in zip(dataset_dicts, proposal_files)
+ ]
+
+ dataset_dicts = list(itertools.chain.from_iterable(dataset_dicts))
+
+ has_instances = "annotations" in dataset_dicts[0]
+ if filter_empty and has_instances:
+ dataset_dicts = filter_images_with_only_crowd_annotations(dataset_dicts)
+ if min_keypoints > 0 and has_instances:
+ dataset_dicts = filter_images_with_few_keypoints(dataset_dicts, min_keypoints)
+
+ if check_consistency and has_instances:
+ try:
+ class_names = MetadataCatalog.get(names[0]).thing_classes
+ check_metadata_consistency("thing_classes", names)
+ print_instances_class_histogram(dataset_dicts, class_names)
+ except AttributeError: # class names are not available for this dataset
+ pass
+
+ assert len(dataset_dicts), "No valid data found in {}.".format(",".join(names))
+ return dataset_dicts
+
+
+def build_batch_data_loader(
+ dataset,
+ sampler,
+ total_batch_size,
+ *,
+ aspect_ratio_grouping=False,
+ num_workers=0,
+ collate_fn=None,
+):
+ """
+ Build a batched dataloader. The main differences from `torch.utils.data.DataLoader` are:
+ 1. support aspect ratio grouping options
+ 2. use no "batch collation", because this is common for detection training
+
+ Args:
+ dataset (torch.utils.data.Dataset): a pytorch map-style or iterable dataset.
+ sampler (torch.utils.data.sampler.Sampler or None): a sampler that produces indices.
+ Must be provided iff. ``dataset`` is a map-style dataset.
+ total_batch_size, aspect_ratio_grouping, num_workers, collate_fn: see
+ :func:`build_detection_train_loader`.
+
+ Returns:
+ iterable[list]. Length of each list is the batch size of the current
+ GPU. Each element in the list comes from the dataset.
+ """
+ world_size = get_world_size()
+ assert (
+ total_batch_size > 0 and total_batch_size % world_size == 0
+ ), "Total batch size ({}) must be divisible by the number of gpus ({}).".format(
+ total_batch_size, world_size
+ )
+ batch_size = total_batch_size // world_size
+
+ if isinstance(dataset, torchdata.IterableDataset):
+ assert sampler is None, "sampler must be None if dataset is IterableDataset"
+ else:
+ dataset = ToIterableDataset(dataset, sampler)
+
+ if aspect_ratio_grouping:
+ data_loader = torchdata.DataLoader(
+ dataset,
+ num_workers=num_workers,
+ collate_fn=operator.itemgetter(0), # don't batch, but yield individual elements
+ worker_init_fn=worker_init_reset_seed,
+ ) # yield individual mapped dict
+ data_loader = AspectRatioGroupedDataset(data_loader, batch_size)
+ if collate_fn is None:
+ return data_loader
+ return MapDataset(data_loader, collate_fn)
+ else:
+ return torchdata.DataLoader(
+ dataset,
+ batch_size=batch_size,
+ drop_last=True,
+ num_workers=num_workers,
+ collate_fn=trivial_batch_collator if collate_fn is None else collate_fn,
+ worker_init_fn=worker_init_reset_seed,
+ )
+
+
+def _train_loader_from_config(cfg, mapper=None, *, dataset=None, sampler=None):
+ if dataset is None:
+ dataset = get_detection_dataset_dicts(
+ cfg.DATASETS.TRAIN,
+ filter_empty=cfg.DATALOADER.FILTER_EMPTY_ANNOTATIONS,
+ min_keypoints=cfg.MODEL.ROI_KEYPOINT_HEAD.MIN_KEYPOINTS_PER_IMAGE
+ if cfg.MODEL.KEYPOINT_ON
+ else 0,
+ proposal_files=cfg.DATASETS.PROPOSAL_FILES_TRAIN if cfg.MODEL.LOAD_PROPOSALS else None,
+ )
+ _log_api_usage("dataset." + cfg.DATASETS.TRAIN[0])
+
+ if mapper is None:
+ mapper = DatasetMapper(cfg, True)
+
+ if sampler is None:
+ sampler_name = cfg.DATALOADER.SAMPLER_TRAIN
+ logger = logging.getLogger(__name__)
+ if isinstance(dataset, torchdata.IterableDataset):
+ logger.info("Not using any sampler since the dataset is IterableDataset.")
+ sampler = None
+ else:
+ logger.info("Using training sampler {}".format(sampler_name))
+ if sampler_name == "TrainingSampler":
+ sampler = TrainingSampler(len(dataset))
+ elif sampler_name == "RepeatFactorTrainingSampler":
+ repeat_factors = RepeatFactorTrainingSampler.repeat_factors_from_category_frequency(
+ dataset, cfg.DATALOADER.REPEAT_THRESHOLD
+ )
+ sampler = RepeatFactorTrainingSampler(repeat_factors)
+ elif sampler_name == "RandomSubsetTrainingSampler":
+ sampler = RandomSubsetTrainingSampler(
+ len(dataset), cfg.DATALOADER.RANDOM_SUBSET_RATIO
+ )
+ else:
+ raise ValueError("Unknown training sampler: {}".format(sampler_name))
+
+ return {
+ "dataset": dataset,
+ "sampler": sampler,
+ "mapper": mapper,
+ "total_batch_size": cfg.SOLVER.IMS_PER_BATCH,
+ "aspect_ratio_grouping": cfg.DATALOADER.ASPECT_RATIO_GROUPING,
+ "num_workers": cfg.DATALOADER.NUM_WORKERS,
+ }
+
+
+@configurable(from_config=_train_loader_from_config)
+def build_detection_train_loader(
+ dataset,
+ *,
+ mapper,
+ sampler=None,
+ total_batch_size,
+ aspect_ratio_grouping=True,
+ num_workers=0,
+ collate_fn=None,
+):
+ """
+ Build a dataloader for object detection with some default features.
+
+ Args:
+ dataset (list or torch.utils.data.Dataset): a list of dataset dicts,
+ or a pytorch dataset (either map-style or iterable). It can be obtained
+ by using :func:`DatasetCatalog.get` or :func:`get_detection_dataset_dicts`.
+ mapper (callable): a callable which takes a sample (dict) from dataset and
+ returns the format to be consumed by the model.
+ When using cfg, the default choice is ``DatasetMapper(cfg, is_train=True)``.
+ sampler (torch.utils.data.sampler.Sampler or None): a sampler that produces
+ indices to be applied on ``dataset``.
+ If ``dataset`` is map-style, the default sampler is a :class:`TrainingSampler`,
+ which coordinates an infinite random shuffle sequence across all workers.
+ Sampler must be None if ``dataset`` is iterable.
+ total_batch_size (int): total batch size across all workers.
+ aspect_ratio_grouping (bool): whether to group images with similar
+ aspect ratio for efficiency. When enabled, it requires each
+ element in dataset be a dict with keys "width" and "height".
+ num_workers (int): number of parallel data loading workers
+ collate_fn: a function that determines how to do batching, same as the argument of
+ `torch.utils.data.DataLoader`. Defaults to do no collation and return a list of
+ data. No collation is OK for small batch size and simple data structures.
+ If your batch size is large and each sample contains too many small tensors,
+ it's more efficient to collate them in data loader.
+
+ Returns:
+ torch.utils.data.DataLoader:
+ a dataloader. Each output from it is a ``list[mapped_element]`` of length
+ ``total_batch_size / num_workers``, where ``mapped_element`` is produced
+ by the ``mapper``.
+ """
+ if isinstance(dataset, list):
+ dataset = DatasetFromList(dataset, copy=False)
+ if mapper is not None:
+ dataset = MapDataset(dataset, mapper)
+
+ if isinstance(dataset, torchdata.IterableDataset):
+ assert sampler is None, "sampler must be None if dataset is IterableDataset"
+ else:
+ if sampler is None:
+ sampler = TrainingSampler(len(dataset))
+ assert isinstance(sampler, torchdata.Sampler), f"Expect a Sampler but got {type(sampler)}"
+ return build_batch_data_loader(
+ dataset,
+ sampler,
+ total_batch_size,
+ aspect_ratio_grouping=aspect_ratio_grouping,
+ num_workers=num_workers,
+ collate_fn=collate_fn,
+ )
+
+
+def _test_loader_from_config(cfg, dataset_name, mapper=None):
+ """
+ Uses the given `dataset_name` argument (instead of the names in cfg), because the
+ standard practice is to evaluate each test set individually (not combining them).
+ """
+ if isinstance(dataset_name, str):
+ dataset_name = [dataset_name]
+
+ dataset = get_detection_dataset_dicts(
+ dataset_name,
+ filter_empty=False,
+ proposal_files=[
+ cfg.DATASETS.PROPOSAL_FILES_TEST[list(cfg.DATASETS.TEST).index(x)] for x in dataset_name
+ ]
+ if cfg.MODEL.LOAD_PROPOSALS
+ else None,
+ )
+ if mapper is None:
+ mapper = DatasetMapper(cfg, False)
+ return {
+ "dataset": dataset,
+ "mapper": mapper,
+ "num_workers": cfg.DATALOADER.NUM_WORKERS,
+ "sampler": InferenceSampler(len(dataset))
+ if not isinstance(dataset, torchdata.IterableDataset)
+ else None,
+ }
+
+
+@configurable(from_config=_test_loader_from_config)
+def build_detection_test_loader(
+ dataset: Union[List[Any], torchdata.Dataset],
+ *,
+ mapper: Callable[[Dict[str, Any]], Any],
+ sampler: Optional[torchdata.Sampler] = None,
+ batch_size: int = 1,
+ num_workers: int = 0,
+ collate_fn: Optional[Callable[[List[Any]], Any]] = None,
+) -> torchdata.DataLoader:
+ """
+ Similar to `build_detection_train_loader`, with default batch size = 1,
+ and sampler = :class:`InferenceSampler`. This sampler coordinates all workers
+ to produce the exact set of all samples.
+
+ Args:
+ dataset: a list of dataset dicts,
+ or a pytorch dataset (either map-style or iterable). They can be obtained
+ by using :func:`DatasetCatalog.get` or :func:`get_detection_dataset_dicts`.
+ mapper: a callable which takes a sample (dict) from dataset
+ and returns the format to be consumed by the model.
+ When using cfg, the default choice is ``DatasetMapper(cfg, is_train=False)``.
+ sampler: a sampler that produces
+ indices to be applied on ``dataset``. Default to :class:`InferenceSampler`,
+ which splits the dataset across all workers. Sampler must be None
+ if `dataset` is iterable.
+ batch_size: the batch size of the data loader to be created.
+ Default to 1 image per worker since this is the standard when reporting
+ inference time in papers.
+ num_workers: number of parallel data loading workers
+ collate_fn: same as the argument of `torch.utils.data.DataLoader`.
+ Defaults to do no collation and return a list of data.
+
+ Returns:
+ DataLoader: a torch DataLoader, that loads the given detection
+ dataset, with test-time transformation and batching.
+
+ Examples:
+ ::
+ data_loader = build_detection_test_loader(
+ DatasetRegistry.get("my_test"),
+ mapper=DatasetMapper(...))
+
+ # or, instantiate with a CfgNode:
+ data_loader = build_detection_test_loader(cfg, "my_test")
+ """
+ if isinstance(dataset, list):
+ dataset = DatasetFromList(dataset, copy=False)
+ if mapper is not None:
+ dataset = MapDataset(dataset, mapper)
+ if isinstance(dataset, torchdata.IterableDataset):
+ assert sampler is None, "sampler must be None if dataset is IterableDataset"
+ else:
+ if sampler is None:
+ sampler = InferenceSampler(len(dataset))
+ return torchdata.DataLoader(
+ dataset,
+ batch_size=batch_size,
+ sampler=sampler,
+ drop_last=False,
+ num_workers=num_workers,
+ collate_fn=trivial_batch_collator if collate_fn is None else collate_fn,
+ )
+
+
+def trivial_batch_collator(batch):
+ """
+ A batch collator that does nothing.
+ """
+ return batch
+
+
+def worker_init_reset_seed(worker_id):
+ initial_seed = torch.initial_seed() % 2**31
+ seed_all_rng(initial_seed + worker_id)
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/catalog.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/catalog.py
new file mode 100644
index 0000000000000000000000000000000000000000..4f5209b5583d01258437bdc9b52a3dd716bdbbf6
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/catalog.py
@@ -0,0 +1,236 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import copy
+import logging
+import types
+from collections import UserDict
+from typing import List
+
+from annotator.oneformer.detectron2.utils.logger import log_first_n
+
+__all__ = ["DatasetCatalog", "MetadataCatalog", "Metadata"]
+
+
+class _DatasetCatalog(UserDict):
+ """
+ A global dictionary that stores information about the datasets and how to obtain them.
+
+ It contains a mapping from strings
+ (which are names that identify a dataset, e.g. "coco_2014_train")
+ to a function which parses the dataset and returns the samples in the
+ format of `list[dict]`.
+
+ The returned dicts should be in Detectron2 Dataset format (See DATASETS.md for details)
+ if used with the data loader functionalities in `data/build.py,data/detection_transform.py`.
+
+ The purpose of having this catalog is to make it easy to choose
+ different datasets, by just using the strings in the config.
+ """
+
+ def register(self, name, func):
+ """
+ Args:
+ name (str): the name that identifies a dataset, e.g. "coco_2014_train".
+ func (callable): a callable which takes no arguments and returns a list of dicts.
+ It must return the same results if called multiple times.
+ """
+ assert callable(func), "You must register a function with `DatasetCatalog.register`!"
+ assert name not in self, "Dataset '{}' is already registered!".format(name)
+ self[name] = func
+
+ def get(self, name):
+ """
+ Call the registered function and return its results.
+
+ Args:
+ name (str): the name that identifies a dataset, e.g. "coco_2014_train".
+
+ Returns:
+ list[dict]: dataset annotations.
+ """
+ try:
+ f = self[name]
+ except KeyError as e:
+ raise KeyError(
+ "Dataset '{}' is not registered! Available datasets are: {}".format(
+ name, ", ".join(list(self.keys()))
+ )
+ ) from e
+ return f()
+
+ def list(self) -> List[str]:
+ """
+ List all registered datasets.
+
+ Returns:
+ list[str]
+ """
+ return list(self.keys())
+
+ def remove(self, name):
+ """
+ Alias of ``pop``.
+ """
+ self.pop(name)
+
+ def __str__(self):
+ return "DatasetCatalog(registered datasets: {})".format(", ".join(self.keys()))
+
+ __repr__ = __str__
+
+
+DatasetCatalog = _DatasetCatalog()
+DatasetCatalog.__doc__ = (
+ _DatasetCatalog.__doc__
+ + """
+ .. automethod:: detectron2.data.catalog.DatasetCatalog.register
+ .. automethod:: detectron2.data.catalog.DatasetCatalog.get
+"""
+)
+
+
+class Metadata(types.SimpleNamespace):
+ """
+ A class that supports simple attribute setter/getter.
+ It is intended for storing metadata of a dataset and make it accessible globally.
+
+ Examples:
+ ::
+ # somewhere when you load the data:
+ MetadataCatalog.get("mydataset").thing_classes = ["person", "dog"]
+
+ # somewhere when you print statistics or visualize:
+ classes = MetadataCatalog.get("mydataset").thing_classes
+ """
+
+ # the name of the dataset
+ # set default to N/A so that `self.name` in the errors will not trigger getattr again
+ name: str = "N/A"
+
+ _RENAMED = {
+ "class_names": "thing_classes",
+ "dataset_id_to_contiguous_id": "thing_dataset_id_to_contiguous_id",
+ "stuff_class_names": "stuff_classes",
+ }
+
+ def __getattr__(self, key):
+ if key in self._RENAMED:
+ log_first_n(
+ logging.WARNING,
+ "Metadata '{}' was renamed to '{}'!".format(key, self._RENAMED[key]),
+ n=10,
+ )
+ return getattr(self, self._RENAMED[key])
+
+ # "name" exists in every metadata
+ if len(self.__dict__) > 1:
+ raise AttributeError(
+ "Attribute '{}' does not exist in the metadata of dataset '{}'. Available "
+ "keys are {}.".format(key, self.name, str(self.__dict__.keys()))
+ )
+ else:
+ raise AttributeError(
+ f"Attribute '{key}' does not exist in the metadata of dataset '{self.name}': "
+ "metadata is empty."
+ )
+
+ def __setattr__(self, key, val):
+ if key in self._RENAMED:
+ log_first_n(
+ logging.WARNING,
+ "Metadata '{}' was renamed to '{}'!".format(key, self._RENAMED[key]),
+ n=10,
+ )
+ setattr(self, self._RENAMED[key], val)
+
+ # Ensure that metadata of the same name stays consistent
+ try:
+ oldval = getattr(self, key)
+ assert oldval == val, (
+ "Attribute '{}' in the metadata of '{}' cannot be set "
+ "to a different value!\n{} != {}".format(key, self.name, oldval, val)
+ )
+ except AttributeError:
+ super().__setattr__(key, val)
+
+ def as_dict(self):
+ """
+ Returns all the metadata as a dict.
+ Note that modifications to the returned dict will not reflect on the Metadata object.
+ """
+ return copy.copy(self.__dict__)
+
+ def set(self, **kwargs):
+ """
+ Set multiple metadata with kwargs.
+ """
+ for k, v in kwargs.items():
+ setattr(self, k, v)
+ return self
+
+ def get(self, key, default=None):
+ """
+ Access an attribute and return its value if exists.
+ Otherwise return default.
+ """
+ try:
+ return getattr(self, key)
+ except AttributeError:
+ return default
+
+
+class _MetadataCatalog(UserDict):
+ """
+ MetadataCatalog is a global dictionary that provides access to
+ :class:`Metadata` of a given dataset.
+
+ The metadata associated with a certain name is a singleton: once created, the
+ metadata will stay alive and will be returned by future calls to ``get(name)``.
+
+ It's like global variables, so don't abuse it.
+ It's meant for storing knowledge that's constant and shared across the execution
+ of the program, e.g.: the class names in COCO.
+ """
+
+ def get(self, name):
+ """
+ Args:
+ name (str): name of a dataset (e.g. coco_2014_train).
+
+ Returns:
+ Metadata: The :class:`Metadata` instance associated with this name,
+ or create an empty one if none is available.
+ """
+ assert len(name)
+ r = super().get(name, None)
+ if r is None:
+ r = self[name] = Metadata(name=name)
+ return r
+
+ def list(self):
+ """
+ List all registered metadata.
+
+ Returns:
+ list[str]: keys (names of datasets) of all registered metadata
+ """
+ return list(self.keys())
+
+ def remove(self, name):
+ """
+ Alias of ``pop``.
+ """
+ self.pop(name)
+
+ def __str__(self):
+ return "MetadataCatalog(registered metadata: {})".format(", ".join(self.keys()))
+
+ __repr__ = __str__
+
+
+MetadataCatalog = _MetadataCatalog()
+MetadataCatalog.__doc__ = (
+ _MetadataCatalog.__doc__
+ + """
+ .. automethod:: detectron2.data.catalog.MetadataCatalog.get
+"""
+)
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/common.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/common.py
new file mode 100644
index 0000000000000000000000000000000000000000..aa69a6a6546030aee818b195a0fbb399d5b776f6
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/common.py
@@ -0,0 +1,301 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import contextlib
+import copy
+import itertools
+import logging
+import numpy as np
+import pickle
+import random
+from typing import Callable, Union
+import torch
+import torch.utils.data as data
+from torch.utils.data.sampler import Sampler
+
+from annotator.oneformer.detectron2.utils.serialize import PicklableWrapper
+
+__all__ = ["MapDataset", "DatasetFromList", "AspectRatioGroupedDataset", "ToIterableDataset"]
+
+logger = logging.getLogger(__name__)
+
+
+def _shard_iterator_dataloader_worker(iterable):
+ # Shard the iterable if we're currently inside pytorch dataloader worker.
+ worker_info = data.get_worker_info()
+ if worker_info is None or worker_info.num_workers == 1:
+ # do nothing
+ yield from iterable
+ else:
+ yield from itertools.islice(iterable, worker_info.id, None, worker_info.num_workers)
+
+
+class _MapIterableDataset(data.IterableDataset):
+ """
+ Map a function over elements in an IterableDataset.
+
+ Similar to pytorch's MapIterDataPipe, but support filtering when map_func
+ returns None.
+
+ This class is not public-facing. Will be called by `MapDataset`.
+ """
+
+ def __init__(self, dataset, map_func):
+ self._dataset = dataset
+ self._map_func = PicklableWrapper(map_func) # wrap so that a lambda will work
+
+ def __len__(self):
+ return len(self._dataset)
+
+ def __iter__(self):
+ for x in map(self._map_func, self._dataset):
+ if x is not None:
+ yield x
+
+
+class MapDataset(data.Dataset):
+ """
+ Map a function over the elements in a dataset.
+ """
+
+ def __init__(self, dataset, map_func):
+ """
+ Args:
+ dataset: a dataset where map function is applied. Can be either
+ map-style or iterable dataset. When given an iterable dataset,
+ the returned object will also be an iterable dataset.
+ map_func: a callable which maps the element in dataset. map_func can
+ return None to skip the data (e.g. in case of errors).
+ How None is handled depends on the style of `dataset`.
+ If `dataset` is map-style, it randomly tries other elements.
+ If `dataset` is iterable, it skips the data and tries the next.
+ """
+ self._dataset = dataset
+ self._map_func = PicklableWrapper(map_func) # wrap so that a lambda will work
+
+ self._rng = random.Random(42)
+ self._fallback_candidates = set(range(len(dataset)))
+
+ def __new__(cls, dataset, map_func):
+ is_iterable = isinstance(dataset, data.IterableDataset)
+ if is_iterable:
+ return _MapIterableDataset(dataset, map_func)
+ else:
+ return super().__new__(cls)
+
+ def __getnewargs__(self):
+ return self._dataset, self._map_func
+
+ def __len__(self):
+ return len(self._dataset)
+
+ def __getitem__(self, idx):
+ retry_count = 0
+ cur_idx = int(idx)
+
+ while True:
+ data = self._map_func(self._dataset[cur_idx])
+ if data is not None:
+ self._fallback_candidates.add(cur_idx)
+ return data
+
+ # _map_func fails for this idx, use a random new index from the pool
+ retry_count += 1
+ self._fallback_candidates.discard(cur_idx)
+ cur_idx = self._rng.sample(self._fallback_candidates, k=1)[0]
+
+ if retry_count >= 3:
+ logger = logging.getLogger(__name__)
+ logger.warning(
+ "Failed to apply `_map_func` for idx: {}, retry count: {}".format(
+ idx, retry_count
+ )
+ )
+
+
+class _TorchSerializedList(object):
+ """
+ A list-like object whose items are serialized and stored in a torch tensor. When
+ launching a process that uses TorchSerializedList with "fork" start method,
+ the subprocess can read the same buffer without triggering copy-on-access. When
+ launching a process that uses TorchSerializedList with "spawn/forkserver" start
+ method, the list will be pickled by a special ForkingPickler registered by PyTorch
+ that moves data to shared memory. In both cases, this allows parent and child
+ processes to share RAM for the list data, hence avoids the issue in
+ https://github.com/pytorch/pytorch/issues/13246.
+
+ See also https://ppwwyyxx.com/blog/2022/Demystify-RAM-Usage-in-Multiprocess-DataLoader/
+ on how it works.
+ """
+
+ def __init__(self, lst: list):
+ self._lst = lst
+
+ def _serialize(data):
+ buffer = pickle.dumps(data, protocol=-1)
+ return np.frombuffer(buffer, dtype=np.uint8)
+
+ logger.info(
+ "Serializing {} elements to byte tensors and concatenating them all ...".format(
+ len(self._lst)
+ )
+ )
+ self._lst = [_serialize(x) for x in self._lst]
+ self._addr = np.asarray([len(x) for x in self._lst], dtype=np.int64)
+ self._addr = torch.from_numpy(np.cumsum(self._addr))
+ self._lst = torch.from_numpy(np.concatenate(self._lst))
+ logger.info("Serialized dataset takes {:.2f} MiB".format(len(self._lst) / 1024**2))
+
+ def __len__(self):
+ return len(self._addr)
+
+ def __getitem__(self, idx):
+ start_addr = 0 if idx == 0 else self._addr[idx - 1].item()
+ end_addr = self._addr[idx].item()
+ bytes = memoryview(self._lst[start_addr:end_addr].numpy())
+
+ # @lint-ignore PYTHONPICKLEISBAD
+ return pickle.loads(bytes)
+
+
+_DEFAULT_DATASET_FROM_LIST_SERIALIZE_METHOD = _TorchSerializedList
+
+
+@contextlib.contextmanager
+def set_default_dataset_from_list_serialize_method(new):
+ """
+ Context manager for using custom serialize function when creating DatasetFromList
+ """
+
+ global _DEFAULT_DATASET_FROM_LIST_SERIALIZE_METHOD
+ orig = _DEFAULT_DATASET_FROM_LIST_SERIALIZE_METHOD
+ _DEFAULT_DATASET_FROM_LIST_SERIALIZE_METHOD = new
+ yield
+ _DEFAULT_DATASET_FROM_LIST_SERIALIZE_METHOD = orig
+
+
+class DatasetFromList(data.Dataset):
+ """
+ Wrap a list to a torch Dataset. It produces elements of the list as data.
+ """
+
+ def __init__(
+ self,
+ lst: list,
+ copy: bool = True,
+ serialize: Union[bool, Callable] = True,
+ ):
+ """
+ Args:
+ lst (list): a list which contains elements to produce.
+ copy (bool): whether to deepcopy the element when producing it,
+ so that the result can be modified in place without affecting the
+ source in the list.
+ serialize (bool or callable): whether to serialize the stroage to other
+ backend. If `True`, the default serialize method will be used, if given
+ a callable, the callable will be used as serialize method.
+ """
+ self._lst = lst
+ self._copy = copy
+ if not isinstance(serialize, (bool, Callable)):
+ raise TypeError(f"Unsupported type for argument `serailzie`: {serialize}")
+ self._serialize = serialize is not False
+
+ if self._serialize:
+ serialize_method = (
+ serialize
+ if isinstance(serialize, Callable)
+ else _DEFAULT_DATASET_FROM_LIST_SERIALIZE_METHOD
+ )
+ logger.info(f"Serializing the dataset using: {serialize_method}")
+ self._lst = serialize_method(self._lst)
+
+ def __len__(self):
+ return len(self._lst)
+
+ def __getitem__(self, idx):
+ if self._copy and not self._serialize:
+ return copy.deepcopy(self._lst[idx])
+ else:
+ return self._lst[idx]
+
+
+class ToIterableDataset(data.IterableDataset):
+ """
+ Convert an old indices-based (also called map-style) dataset
+ to an iterable-style dataset.
+ """
+
+ def __init__(self, dataset: data.Dataset, sampler: Sampler, shard_sampler: bool = True):
+ """
+ Args:
+ dataset: an old-style dataset with ``__getitem__``
+ sampler: a cheap iterable that produces indices to be applied on ``dataset``.
+ shard_sampler: whether to shard the sampler based on the current pytorch data loader
+ worker id. When an IterableDataset is forked by pytorch's DataLoader into multiple
+ workers, it is responsible for sharding its data based on worker id so that workers
+ don't produce identical data.
+
+ Most samplers (like our TrainingSampler) do not shard based on dataloader worker id
+ and this argument should be set to True. But certain samplers may be already
+ sharded, in that case this argument should be set to False.
+ """
+ assert not isinstance(dataset, data.IterableDataset), dataset
+ assert isinstance(sampler, Sampler), sampler
+ self.dataset = dataset
+ self.sampler = sampler
+ self.shard_sampler = shard_sampler
+
+ def __iter__(self):
+ if not self.shard_sampler:
+ sampler = self.sampler
+ else:
+ # With map-style dataset, `DataLoader(dataset, sampler)` runs the
+ # sampler in main process only. But `DataLoader(ToIterableDataset(dataset, sampler))`
+ # will run sampler in every of the N worker. So we should only keep 1/N of the ids on
+ # each worker. The assumption is that sampler is cheap to iterate so it's fine to
+ # discard ids in workers.
+ sampler = _shard_iterator_dataloader_worker(self.sampler)
+ for idx in sampler:
+ yield self.dataset[idx]
+
+ def __len__(self):
+ return len(self.sampler)
+
+
+class AspectRatioGroupedDataset(data.IterableDataset):
+ """
+ Batch data that have similar aspect ratio together.
+ In this implementation, images whose aspect ratio < (or >) 1 will
+ be batched together.
+ This improves training speed because the images then need less padding
+ to form a batch.
+
+ It assumes the underlying dataset produces dicts with "width" and "height" keys.
+ It will then produce a list of original dicts with length = batch_size,
+ all with similar aspect ratios.
+ """
+
+ def __init__(self, dataset, batch_size):
+ """
+ Args:
+ dataset: an iterable. Each element must be a dict with keys
+ "width" and "height", which will be used to batch data.
+ batch_size (int):
+ """
+ self.dataset = dataset
+ self.batch_size = batch_size
+ self._buckets = [[] for _ in range(2)]
+ # Hard-coded two aspect ratio groups: w > h and w < h.
+ # Can add support for more aspect ratio groups, but doesn't seem useful
+
+ def __iter__(self):
+ for d in self.dataset:
+ w, h = d["width"], d["height"]
+ bucket_id = 0 if w > h else 1
+ bucket = self._buckets[bucket_id]
+ bucket.append(d)
+ if len(bucket) == self.batch_size:
+ data = bucket[:]
+ # Clear bucket first, because code after yield is not
+ # guaranteed to execute
+ del bucket[:]
+ yield data
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/dataset_mapper.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/dataset_mapper.py
new file mode 100644
index 0000000000000000000000000000000000000000..3bb6bb1057a68bfb12e55872f391065f02023ed3
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/dataset_mapper.py
@@ -0,0 +1,191 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import copy
+import logging
+import numpy as np
+from typing import List, Optional, Union
+import torch
+
+from annotator.oneformer.detectron2.config import configurable
+
+from . import detection_utils as utils
+from . import transforms as T
+
+"""
+This file contains the default mapping that's applied to "dataset dicts".
+"""
+
+__all__ = ["DatasetMapper"]
+
+
+class DatasetMapper:
+ """
+ A callable which takes a dataset dict in Detectron2 Dataset format,
+ and map it into a format used by the model.
+
+ This is the default callable to be used to map your dataset dict into training data.
+ You may need to follow it to implement your own one for customized logic,
+ such as a different way to read or transform images.
+ See :doc:`/tutorials/data_loading` for details.
+
+ The callable currently does the following:
+
+ 1. Read the image from "file_name"
+ 2. Applies cropping/geometric transforms to the image and annotations
+ 3. Prepare data and annotations to Tensor and :class:`Instances`
+ """
+
+ @configurable
+ def __init__(
+ self,
+ is_train: bool,
+ *,
+ augmentations: List[Union[T.Augmentation, T.Transform]],
+ image_format: str,
+ use_instance_mask: bool = False,
+ use_keypoint: bool = False,
+ instance_mask_format: str = "polygon",
+ keypoint_hflip_indices: Optional[np.ndarray] = None,
+ precomputed_proposal_topk: Optional[int] = None,
+ recompute_boxes: bool = False,
+ ):
+ """
+ NOTE: this interface is experimental.
+
+ Args:
+ is_train: whether it's used in training or inference
+ augmentations: a list of augmentations or deterministic transforms to apply
+ image_format: an image format supported by :func:`detection_utils.read_image`.
+ use_instance_mask: whether to process instance segmentation annotations, if available
+ use_keypoint: whether to process keypoint annotations if available
+ instance_mask_format: one of "polygon" or "bitmask". Process instance segmentation
+ masks into this format.
+ keypoint_hflip_indices: see :func:`detection_utils.create_keypoint_hflip_indices`
+ precomputed_proposal_topk: if given, will load pre-computed
+ proposals from dataset_dict and keep the top k proposals for each image.
+ recompute_boxes: whether to overwrite bounding box annotations
+ by computing tight bounding boxes from instance mask annotations.
+ """
+ if recompute_boxes:
+ assert use_instance_mask, "recompute_boxes requires instance masks"
+ # fmt: off
+ self.is_train = is_train
+ self.augmentations = T.AugmentationList(augmentations)
+ self.image_format = image_format
+ self.use_instance_mask = use_instance_mask
+ self.instance_mask_format = instance_mask_format
+ self.use_keypoint = use_keypoint
+ self.keypoint_hflip_indices = keypoint_hflip_indices
+ self.proposal_topk = precomputed_proposal_topk
+ self.recompute_boxes = recompute_boxes
+ # fmt: on
+ logger = logging.getLogger(__name__)
+ mode = "training" if is_train else "inference"
+ logger.info(f"[DatasetMapper] Augmentations used in {mode}: {augmentations}")
+
+ @classmethod
+ def from_config(cls, cfg, is_train: bool = True):
+ augs = utils.build_augmentation(cfg, is_train)
+ if cfg.INPUT.CROP.ENABLED and is_train:
+ augs.insert(0, T.RandomCrop(cfg.INPUT.CROP.TYPE, cfg.INPUT.CROP.SIZE))
+ recompute_boxes = cfg.MODEL.MASK_ON
+ else:
+ recompute_boxes = False
+
+ ret = {
+ "is_train": is_train,
+ "augmentations": augs,
+ "image_format": cfg.INPUT.FORMAT,
+ "use_instance_mask": cfg.MODEL.MASK_ON,
+ "instance_mask_format": cfg.INPUT.MASK_FORMAT,
+ "use_keypoint": cfg.MODEL.KEYPOINT_ON,
+ "recompute_boxes": recompute_boxes,
+ }
+
+ if cfg.MODEL.KEYPOINT_ON:
+ ret["keypoint_hflip_indices"] = utils.create_keypoint_hflip_indices(cfg.DATASETS.TRAIN)
+
+ if cfg.MODEL.LOAD_PROPOSALS:
+ ret["precomputed_proposal_topk"] = (
+ cfg.DATASETS.PRECOMPUTED_PROPOSAL_TOPK_TRAIN
+ if is_train
+ else cfg.DATASETS.PRECOMPUTED_PROPOSAL_TOPK_TEST
+ )
+ return ret
+
+ def _transform_annotations(self, dataset_dict, transforms, image_shape):
+ # USER: Modify this if you want to keep them for some reason.
+ for anno in dataset_dict["annotations"]:
+ if not self.use_instance_mask:
+ anno.pop("segmentation", None)
+ if not self.use_keypoint:
+ anno.pop("keypoints", None)
+
+ # USER: Implement additional transformations if you have other types of data
+ annos = [
+ utils.transform_instance_annotations(
+ obj, transforms, image_shape, keypoint_hflip_indices=self.keypoint_hflip_indices
+ )
+ for obj in dataset_dict.pop("annotations")
+ if obj.get("iscrowd", 0) == 0
+ ]
+ instances = utils.annotations_to_instances(
+ annos, image_shape, mask_format=self.instance_mask_format
+ )
+
+ # After transforms such as cropping are applied, the bounding box may no longer
+ # tightly bound the object. As an example, imagine a triangle object
+ # [(0,0), (2,0), (0,2)] cropped by a box [(1,0),(2,2)] (XYXY format). The tight
+ # bounding box of the cropped triangle should be [(1,0),(2,1)], which is not equal to
+ # the intersection of original bounding box and the cropping box.
+ if self.recompute_boxes:
+ instances.gt_boxes = instances.gt_masks.get_bounding_boxes()
+ dataset_dict["instances"] = utils.filter_empty_instances(instances)
+
+ def __call__(self, dataset_dict):
+ """
+ Args:
+ dataset_dict (dict): Metadata of one image, in Detectron2 Dataset format.
+
+ Returns:
+ dict: a format that builtin models in detectron2 accept
+ """
+ dataset_dict = copy.deepcopy(dataset_dict) # it will be modified by code below
+ # USER: Write your own image loading if it's not from a file
+ image = utils.read_image(dataset_dict["file_name"], format=self.image_format)
+ utils.check_image_size(dataset_dict, image)
+
+ # USER: Remove if you don't do semantic/panoptic segmentation.
+ if "sem_seg_file_name" in dataset_dict:
+ sem_seg_gt = utils.read_image(dataset_dict.pop("sem_seg_file_name"), "L").squeeze(2)
+ else:
+ sem_seg_gt = None
+
+ aug_input = T.AugInput(image, sem_seg=sem_seg_gt)
+ transforms = self.augmentations(aug_input)
+ image, sem_seg_gt = aug_input.image, aug_input.sem_seg
+
+ image_shape = image.shape[:2] # h, w
+ # Pytorch's dataloader is efficient on torch.Tensor due to shared-memory,
+ # but not efficient on large generic data structures due to the use of pickle & mp.Queue.
+ # Therefore it's important to use torch.Tensor.
+ dataset_dict["image"] = torch.as_tensor(np.ascontiguousarray(image.transpose(2, 0, 1)))
+ if sem_seg_gt is not None:
+ dataset_dict["sem_seg"] = torch.as_tensor(sem_seg_gt.astype("long"))
+
+ # USER: Remove if you don't use pre-computed proposals.
+ # Most users would not need this feature.
+ if self.proposal_topk is not None:
+ utils.transform_proposals(
+ dataset_dict, image_shape, transforms, proposal_topk=self.proposal_topk
+ )
+
+ if not self.is_train:
+ # USER: Modify this if you want to keep them for some reason.
+ dataset_dict.pop("annotations", None)
+ dataset_dict.pop("sem_seg_file_name", None)
+ return dataset_dict
+
+ if "annotations" in dataset_dict:
+ self._transform_annotations(dataset_dict, transforms, image_shape)
+
+ return dataset_dict
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/README.md b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..9fb3e4f7afec17137c95c78be6ef06d520ec8032
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/README.md
@@ -0,0 +1,9 @@
+
+
+### Common Datasets
+
+The dataset implemented here do not need to load the data into the final format.
+It should provide the minimal data structure needed to use the dataset, so it can be very efficient.
+
+For example, for an image dataset, just provide the file names and labels, but don't read the images.
+Let the downstream decide how to read.
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/__init__.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..a44bedc15e5f0e762fc4d77efd6f1b07c6ff77d0
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/__init__.py
@@ -0,0 +1,9 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+from .coco import load_coco_json, load_sem_seg, register_coco_instances, convert_to_coco_json
+from .coco_panoptic import register_coco_panoptic, register_coco_panoptic_separated
+from .lvis import load_lvis_json, register_lvis_instances, get_lvis_instances_meta
+from .pascal_voc import load_voc_instances, register_pascal_voc
+from . import builtin as _builtin # ensure the builtin datasets are registered
+
+
+__all__ = [k for k in globals().keys() if not k.startswith("_")]
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/builtin.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/builtin.py
new file mode 100644
index 0000000000000000000000000000000000000000..39bbb1feec64f76705ba32c46f19f89f71be2ca7
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/builtin.py
@@ -0,0 +1,259 @@
+# -*- coding: utf-8 -*-
+# Copyright (c) Facebook, Inc. and its affiliates.
+
+
+"""
+This file registers pre-defined datasets at hard-coded paths, and their metadata.
+
+We hard-code metadata for common datasets. This will enable:
+1. Consistency check when loading the datasets
+2. Use models on these standard datasets directly and run demos,
+ without having to download the dataset annotations
+
+We hard-code some paths to the dataset that's assumed to
+exist in "./datasets/".
+
+Users SHOULD NOT use this file to create new dataset / metadata for new dataset.
+To add new dataset, refer to the tutorial "docs/DATASETS.md".
+"""
+
+import os
+
+from annotator.oneformer.detectron2.data import DatasetCatalog, MetadataCatalog
+
+from .builtin_meta import ADE20K_SEM_SEG_CATEGORIES, _get_builtin_metadata
+from .cityscapes import load_cityscapes_instances, load_cityscapes_semantic
+from .cityscapes_panoptic import register_all_cityscapes_panoptic
+from .coco import load_sem_seg, register_coco_instances
+from .coco_panoptic import register_coco_panoptic, register_coco_panoptic_separated
+from .lvis import get_lvis_instances_meta, register_lvis_instances
+from .pascal_voc import register_pascal_voc
+
+# ==== Predefined datasets and splits for COCO ==========
+
+_PREDEFINED_SPLITS_COCO = {}
+_PREDEFINED_SPLITS_COCO["coco"] = {
+ "coco_2014_train": ("coco/train2014", "coco/annotations/instances_train2014.json"),
+ "coco_2014_val": ("coco/val2014", "coco/annotations/instances_val2014.json"),
+ "coco_2014_minival": ("coco/val2014", "coco/annotations/instances_minival2014.json"),
+ "coco_2014_valminusminival": (
+ "coco/val2014",
+ "coco/annotations/instances_valminusminival2014.json",
+ ),
+ "coco_2017_train": ("coco/train2017", "coco/annotations/instances_train2017.json"),
+ "coco_2017_val": ("coco/val2017", "coco/annotations/instances_val2017.json"),
+ "coco_2017_test": ("coco/test2017", "coco/annotations/image_info_test2017.json"),
+ "coco_2017_test-dev": ("coco/test2017", "coco/annotations/image_info_test-dev2017.json"),
+ "coco_2017_val_100": ("coco/val2017", "coco/annotations/instances_val2017_100.json"),
+}
+
+_PREDEFINED_SPLITS_COCO["coco_person"] = {
+ "keypoints_coco_2014_train": (
+ "coco/train2014",
+ "coco/annotations/person_keypoints_train2014.json",
+ ),
+ "keypoints_coco_2014_val": ("coco/val2014", "coco/annotations/person_keypoints_val2014.json"),
+ "keypoints_coco_2014_minival": (
+ "coco/val2014",
+ "coco/annotations/person_keypoints_minival2014.json",
+ ),
+ "keypoints_coco_2014_valminusminival": (
+ "coco/val2014",
+ "coco/annotations/person_keypoints_valminusminival2014.json",
+ ),
+ "keypoints_coco_2017_train": (
+ "coco/train2017",
+ "coco/annotations/person_keypoints_train2017.json",
+ ),
+ "keypoints_coco_2017_val": ("coco/val2017", "coco/annotations/person_keypoints_val2017.json"),
+ "keypoints_coco_2017_val_100": (
+ "coco/val2017",
+ "coco/annotations/person_keypoints_val2017_100.json",
+ ),
+}
+
+
+_PREDEFINED_SPLITS_COCO_PANOPTIC = {
+ "coco_2017_train_panoptic": (
+ # This is the original panoptic annotation directory
+ "coco/panoptic_train2017",
+ "coco/annotations/panoptic_train2017.json",
+ # This directory contains semantic annotations that are
+ # converted from panoptic annotations.
+ # It is used by PanopticFPN.
+ # You can use the script at detectron2/datasets/prepare_panoptic_fpn.py
+ # to create these directories.
+ "coco/panoptic_stuff_train2017",
+ ),
+ "coco_2017_val_panoptic": (
+ "coco/panoptic_val2017",
+ "coco/annotations/panoptic_val2017.json",
+ "coco/panoptic_stuff_val2017",
+ ),
+ "coco_2017_val_100_panoptic": (
+ "coco/panoptic_val2017_100",
+ "coco/annotations/panoptic_val2017_100.json",
+ "coco/panoptic_stuff_val2017_100",
+ ),
+}
+
+
+def register_all_coco(root):
+ for dataset_name, splits_per_dataset in _PREDEFINED_SPLITS_COCO.items():
+ for key, (image_root, json_file) in splits_per_dataset.items():
+ # Assume pre-defined datasets live in `./datasets`.
+ register_coco_instances(
+ key,
+ _get_builtin_metadata(dataset_name),
+ os.path.join(root, json_file) if "://" not in json_file else json_file,
+ os.path.join(root, image_root),
+ )
+
+ for (
+ prefix,
+ (panoptic_root, panoptic_json, semantic_root),
+ ) in _PREDEFINED_SPLITS_COCO_PANOPTIC.items():
+ prefix_instances = prefix[: -len("_panoptic")]
+ instances_meta = MetadataCatalog.get(prefix_instances)
+ image_root, instances_json = instances_meta.image_root, instances_meta.json_file
+ # The "separated" version of COCO panoptic segmentation dataset,
+ # e.g. used by Panoptic FPN
+ register_coco_panoptic_separated(
+ prefix,
+ _get_builtin_metadata("coco_panoptic_separated"),
+ image_root,
+ os.path.join(root, panoptic_root),
+ os.path.join(root, panoptic_json),
+ os.path.join(root, semantic_root),
+ instances_json,
+ )
+ # The "standard" version of COCO panoptic segmentation dataset,
+ # e.g. used by Panoptic-DeepLab
+ register_coco_panoptic(
+ prefix,
+ _get_builtin_metadata("coco_panoptic_standard"),
+ image_root,
+ os.path.join(root, panoptic_root),
+ os.path.join(root, panoptic_json),
+ instances_json,
+ )
+
+
+# ==== Predefined datasets and splits for LVIS ==========
+
+
+_PREDEFINED_SPLITS_LVIS = {
+ "lvis_v1": {
+ "lvis_v1_train": ("coco/", "lvis/lvis_v1_train.json"),
+ "lvis_v1_val": ("coco/", "lvis/lvis_v1_val.json"),
+ "lvis_v1_test_dev": ("coco/", "lvis/lvis_v1_image_info_test_dev.json"),
+ "lvis_v1_test_challenge": ("coco/", "lvis/lvis_v1_image_info_test_challenge.json"),
+ },
+ "lvis_v0.5": {
+ "lvis_v0.5_train": ("coco/", "lvis/lvis_v0.5_train.json"),
+ "lvis_v0.5_val": ("coco/", "lvis/lvis_v0.5_val.json"),
+ "lvis_v0.5_val_rand_100": ("coco/", "lvis/lvis_v0.5_val_rand_100.json"),
+ "lvis_v0.5_test": ("coco/", "lvis/lvis_v0.5_image_info_test.json"),
+ },
+ "lvis_v0.5_cocofied": {
+ "lvis_v0.5_train_cocofied": ("coco/", "lvis/lvis_v0.5_train_cocofied.json"),
+ "lvis_v0.5_val_cocofied": ("coco/", "lvis/lvis_v0.5_val_cocofied.json"),
+ },
+}
+
+
+def register_all_lvis(root):
+ for dataset_name, splits_per_dataset in _PREDEFINED_SPLITS_LVIS.items():
+ for key, (image_root, json_file) in splits_per_dataset.items():
+ register_lvis_instances(
+ key,
+ get_lvis_instances_meta(dataset_name),
+ os.path.join(root, json_file) if "://" not in json_file else json_file,
+ os.path.join(root, image_root),
+ )
+
+
+# ==== Predefined splits for raw cityscapes images ===========
+_RAW_CITYSCAPES_SPLITS = {
+ "cityscapes_fine_{task}_train": ("cityscapes/leftImg8bit/train/", "cityscapes/gtFine/train/"),
+ "cityscapes_fine_{task}_val": ("cityscapes/leftImg8bit/val/", "cityscapes/gtFine/val/"),
+ "cityscapes_fine_{task}_test": ("cityscapes/leftImg8bit/test/", "cityscapes/gtFine/test/"),
+}
+
+
+def register_all_cityscapes(root):
+ for key, (image_dir, gt_dir) in _RAW_CITYSCAPES_SPLITS.items():
+ meta = _get_builtin_metadata("cityscapes")
+ image_dir = os.path.join(root, image_dir)
+ gt_dir = os.path.join(root, gt_dir)
+
+ inst_key = key.format(task="instance_seg")
+ DatasetCatalog.register(
+ inst_key,
+ lambda x=image_dir, y=gt_dir: load_cityscapes_instances(
+ x, y, from_json=True, to_polygons=True
+ ),
+ )
+ MetadataCatalog.get(inst_key).set(
+ image_dir=image_dir, gt_dir=gt_dir, evaluator_type="cityscapes_instance", **meta
+ )
+
+ sem_key = key.format(task="sem_seg")
+ DatasetCatalog.register(
+ sem_key, lambda x=image_dir, y=gt_dir: load_cityscapes_semantic(x, y)
+ )
+ MetadataCatalog.get(sem_key).set(
+ image_dir=image_dir,
+ gt_dir=gt_dir,
+ evaluator_type="cityscapes_sem_seg",
+ ignore_label=255,
+ **meta,
+ )
+
+
+# ==== Predefined splits for PASCAL VOC ===========
+def register_all_pascal_voc(root):
+ SPLITS = [
+ ("voc_2007_trainval", "VOC2007", "trainval"),
+ ("voc_2007_train", "VOC2007", "train"),
+ ("voc_2007_val", "VOC2007", "val"),
+ ("voc_2007_test", "VOC2007", "test"),
+ ("voc_2012_trainval", "VOC2012", "trainval"),
+ ("voc_2012_train", "VOC2012", "train"),
+ ("voc_2012_val", "VOC2012", "val"),
+ ]
+ for name, dirname, split in SPLITS:
+ year = 2007 if "2007" in name else 2012
+ register_pascal_voc(name, os.path.join(root, dirname), split, year)
+ MetadataCatalog.get(name).evaluator_type = "pascal_voc"
+
+
+def register_all_ade20k(root):
+ root = os.path.join(root, "ADEChallengeData2016")
+ for name, dirname in [("train", "training"), ("val", "validation")]:
+ image_dir = os.path.join(root, "images", dirname)
+ gt_dir = os.path.join(root, "annotations_detectron2", dirname)
+ name = f"ade20k_sem_seg_{name}"
+ DatasetCatalog.register(
+ name, lambda x=image_dir, y=gt_dir: load_sem_seg(y, x, gt_ext="png", image_ext="jpg")
+ )
+ MetadataCatalog.get(name).set(
+ stuff_classes=ADE20K_SEM_SEG_CATEGORIES[:],
+ image_root=image_dir,
+ sem_seg_root=gt_dir,
+ evaluator_type="sem_seg",
+ ignore_label=255,
+ )
+
+
+# True for open source;
+# Internally at fb, we register them elsewhere
+if __name__.endswith(".builtin"):
+ # Assume pre-defined datasets live in `./datasets`.
+ _root = os.path.expanduser(os.getenv("DETECTRON2_DATASETS", "datasets"))
+ register_all_coco(_root)
+ register_all_lvis(_root)
+ register_all_cityscapes(_root)
+ register_all_cityscapes_panoptic(_root)
+ register_all_pascal_voc(_root)
+ register_all_ade20k(_root)
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/builtin_meta.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/builtin_meta.py
new file mode 100644
index 0000000000000000000000000000000000000000..63c7a1a31b31dd89b82011effee26471faccacf5
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/builtin_meta.py
@@ -0,0 +1,350 @@
+# -*- coding: utf-8 -*-
+# Copyright (c) Facebook, Inc. and its affiliates.
+
+"""
+Note:
+For your custom dataset, there is no need to hard-code metadata anywhere in the code.
+For example, for COCO-format dataset, metadata will be obtained automatically
+when calling `load_coco_json`. For other dataset, metadata may also be obtained in other ways
+during loading.
+
+However, we hard-coded metadata for a few common dataset here.
+The only goal is to allow users who don't have these dataset to use pre-trained models.
+Users don't have to download a COCO json (which contains metadata), in order to visualize a
+COCO model (with correct class names and colors).
+"""
+
+
+# All coco categories, together with their nice-looking visualization colors
+# It's from https://github.com/cocodataset/panopticapi/blob/master/panoptic_coco_categories.json
+COCO_CATEGORIES = [
+ {"color": [220, 20, 60], "isthing": 1, "id": 1, "name": "person"},
+ {"color": [119, 11, 32], "isthing": 1, "id": 2, "name": "bicycle"},
+ {"color": [0, 0, 142], "isthing": 1, "id": 3, "name": "car"},
+ {"color": [0, 0, 230], "isthing": 1, "id": 4, "name": "motorcycle"},
+ {"color": [106, 0, 228], "isthing": 1, "id": 5, "name": "airplane"},
+ {"color": [0, 60, 100], "isthing": 1, "id": 6, "name": "bus"},
+ {"color": [0, 80, 100], "isthing": 1, "id": 7, "name": "train"},
+ {"color": [0, 0, 70], "isthing": 1, "id": 8, "name": "truck"},
+ {"color": [0, 0, 192], "isthing": 1, "id": 9, "name": "boat"},
+ {"color": [250, 170, 30], "isthing": 1, "id": 10, "name": "traffic light"},
+ {"color": [100, 170, 30], "isthing": 1, "id": 11, "name": "fire hydrant"},
+ {"color": [220, 220, 0], "isthing": 1, "id": 13, "name": "stop sign"},
+ {"color": [175, 116, 175], "isthing": 1, "id": 14, "name": "parking meter"},
+ {"color": [250, 0, 30], "isthing": 1, "id": 15, "name": "bench"},
+ {"color": [165, 42, 42], "isthing": 1, "id": 16, "name": "bird"},
+ {"color": [255, 77, 255], "isthing": 1, "id": 17, "name": "cat"},
+ {"color": [0, 226, 252], "isthing": 1, "id": 18, "name": "dog"},
+ {"color": [182, 182, 255], "isthing": 1, "id": 19, "name": "horse"},
+ {"color": [0, 82, 0], "isthing": 1, "id": 20, "name": "sheep"},
+ {"color": [120, 166, 157], "isthing": 1, "id": 21, "name": "cow"},
+ {"color": [110, 76, 0], "isthing": 1, "id": 22, "name": "elephant"},
+ {"color": [174, 57, 255], "isthing": 1, "id": 23, "name": "bear"},
+ {"color": [199, 100, 0], "isthing": 1, "id": 24, "name": "zebra"},
+ {"color": [72, 0, 118], "isthing": 1, "id": 25, "name": "giraffe"},
+ {"color": [255, 179, 240], "isthing": 1, "id": 27, "name": "backpack"},
+ {"color": [0, 125, 92], "isthing": 1, "id": 28, "name": "umbrella"},
+ {"color": [209, 0, 151], "isthing": 1, "id": 31, "name": "handbag"},
+ {"color": [188, 208, 182], "isthing": 1, "id": 32, "name": "tie"},
+ {"color": [0, 220, 176], "isthing": 1, "id": 33, "name": "suitcase"},
+ {"color": [255, 99, 164], "isthing": 1, "id": 34, "name": "frisbee"},
+ {"color": [92, 0, 73], "isthing": 1, "id": 35, "name": "skis"},
+ {"color": [133, 129, 255], "isthing": 1, "id": 36, "name": "snowboard"},
+ {"color": [78, 180, 255], "isthing": 1, "id": 37, "name": "sports ball"},
+ {"color": [0, 228, 0], "isthing": 1, "id": 38, "name": "kite"},
+ {"color": [174, 255, 243], "isthing": 1, "id": 39, "name": "baseball bat"},
+ {"color": [45, 89, 255], "isthing": 1, "id": 40, "name": "baseball glove"},
+ {"color": [134, 134, 103], "isthing": 1, "id": 41, "name": "skateboard"},
+ {"color": [145, 148, 174], "isthing": 1, "id": 42, "name": "surfboard"},
+ {"color": [255, 208, 186], "isthing": 1, "id": 43, "name": "tennis racket"},
+ {"color": [197, 226, 255], "isthing": 1, "id": 44, "name": "bottle"},
+ {"color": [171, 134, 1], "isthing": 1, "id": 46, "name": "wine glass"},
+ {"color": [109, 63, 54], "isthing": 1, "id": 47, "name": "cup"},
+ {"color": [207, 138, 255], "isthing": 1, "id": 48, "name": "fork"},
+ {"color": [151, 0, 95], "isthing": 1, "id": 49, "name": "knife"},
+ {"color": [9, 80, 61], "isthing": 1, "id": 50, "name": "spoon"},
+ {"color": [84, 105, 51], "isthing": 1, "id": 51, "name": "bowl"},
+ {"color": [74, 65, 105], "isthing": 1, "id": 52, "name": "banana"},
+ {"color": [166, 196, 102], "isthing": 1, "id": 53, "name": "apple"},
+ {"color": [208, 195, 210], "isthing": 1, "id": 54, "name": "sandwich"},
+ {"color": [255, 109, 65], "isthing": 1, "id": 55, "name": "orange"},
+ {"color": [0, 143, 149], "isthing": 1, "id": 56, "name": "broccoli"},
+ {"color": [179, 0, 194], "isthing": 1, "id": 57, "name": "carrot"},
+ {"color": [209, 99, 106], "isthing": 1, "id": 58, "name": "hot dog"},
+ {"color": [5, 121, 0], "isthing": 1, "id": 59, "name": "pizza"},
+ {"color": [227, 255, 205], "isthing": 1, "id": 60, "name": "donut"},
+ {"color": [147, 186, 208], "isthing": 1, "id": 61, "name": "cake"},
+ {"color": [153, 69, 1], "isthing": 1, "id": 62, "name": "chair"},
+ {"color": [3, 95, 161], "isthing": 1, "id": 63, "name": "couch"},
+ {"color": [163, 255, 0], "isthing": 1, "id": 64, "name": "potted plant"},
+ {"color": [119, 0, 170], "isthing": 1, "id": 65, "name": "bed"},
+ {"color": [0, 182, 199], "isthing": 1, "id": 67, "name": "dining table"},
+ {"color": [0, 165, 120], "isthing": 1, "id": 70, "name": "toilet"},
+ {"color": [183, 130, 88], "isthing": 1, "id": 72, "name": "tv"},
+ {"color": [95, 32, 0], "isthing": 1, "id": 73, "name": "laptop"},
+ {"color": [130, 114, 135], "isthing": 1, "id": 74, "name": "mouse"},
+ {"color": [110, 129, 133], "isthing": 1, "id": 75, "name": "remote"},
+ {"color": [166, 74, 118], "isthing": 1, "id": 76, "name": "keyboard"},
+ {"color": [219, 142, 185], "isthing": 1, "id": 77, "name": "cell phone"},
+ {"color": [79, 210, 114], "isthing": 1, "id": 78, "name": "microwave"},
+ {"color": [178, 90, 62], "isthing": 1, "id": 79, "name": "oven"},
+ {"color": [65, 70, 15], "isthing": 1, "id": 80, "name": "toaster"},
+ {"color": [127, 167, 115], "isthing": 1, "id": 81, "name": "sink"},
+ {"color": [59, 105, 106], "isthing": 1, "id": 82, "name": "refrigerator"},
+ {"color": [142, 108, 45], "isthing": 1, "id": 84, "name": "book"},
+ {"color": [196, 172, 0], "isthing": 1, "id": 85, "name": "clock"},
+ {"color": [95, 54, 80], "isthing": 1, "id": 86, "name": "vase"},
+ {"color": [128, 76, 255], "isthing": 1, "id": 87, "name": "scissors"},
+ {"color": [201, 57, 1], "isthing": 1, "id": 88, "name": "teddy bear"},
+ {"color": [246, 0, 122], "isthing": 1, "id": 89, "name": "hair drier"},
+ {"color": [191, 162, 208], "isthing": 1, "id": 90, "name": "toothbrush"},
+ {"color": [255, 255, 128], "isthing": 0, "id": 92, "name": "banner"},
+ {"color": [147, 211, 203], "isthing": 0, "id": 93, "name": "blanket"},
+ {"color": [150, 100, 100], "isthing": 0, "id": 95, "name": "bridge"},
+ {"color": [168, 171, 172], "isthing": 0, "id": 100, "name": "cardboard"},
+ {"color": [146, 112, 198], "isthing": 0, "id": 107, "name": "counter"},
+ {"color": [210, 170, 100], "isthing": 0, "id": 109, "name": "curtain"},
+ {"color": [92, 136, 89], "isthing": 0, "id": 112, "name": "door-stuff"},
+ {"color": [218, 88, 184], "isthing": 0, "id": 118, "name": "floor-wood"},
+ {"color": [241, 129, 0], "isthing": 0, "id": 119, "name": "flower"},
+ {"color": [217, 17, 255], "isthing": 0, "id": 122, "name": "fruit"},
+ {"color": [124, 74, 181], "isthing": 0, "id": 125, "name": "gravel"},
+ {"color": [70, 70, 70], "isthing": 0, "id": 128, "name": "house"},
+ {"color": [255, 228, 255], "isthing": 0, "id": 130, "name": "light"},
+ {"color": [154, 208, 0], "isthing": 0, "id": 133, "name": "mirror-stuff"},
+ {"color": [193, 0, 92], "isthing": 0, "id": 138, "name": "net"},
+ {"color": [76, 91, 113], "isthing": 0, "id": 141, "name": "pillow"},
+ {"color": [255, 180, 195], "isthing": 0, "id": 144, "name": "platform"},
+ {"color": [106, 154, 176], "isthing": 0, "id": 145, "name": "playingfield"},
+ {"color": [230, 150, 140], "isthing": 0, "id": 147, "name": "railroad"},
+ {"color": [60, 143, 255], "isthing": 0, "id": 148, "name": "river"},
+ {"color": [128, 64, 128], "isthing": 0, "id": 149, "name": "road"},
+ {"color": [92, 82, 55], "isthing": 0, "id": 151, "name": "roof"},
+ {"color": [254, 212, 124], "isthing": 0, "id": 154, "name": "sand"},
+ {"color": [73, 77, 174], "isthing": 0, "id": 155, "name": "sea"},
+ {"color": [255, 160, 98], "isthing": 0, "id": 156, "name": "shelf"},
+ {"color": [255, 255, 255], "isthing": 0, "id": 159, "name": "snow"},
+ {"color": [104, 84, 109], "isthing": 0, "id": 161, "name": "stairs"},
+ {"color": [169, 164, 131], "isthing": 0, "id": 166, "name": "tent"},
+ {"color": [225, 199, 255], "isthing": 0, "id": 168, "name": "towel"},
+ {"color": [137, 54, 74], "isthing": 0, "id": 171, "name": "wall-brick"},
+ {"color": [135, 158, 223], "isthing": 0, "id": 175, "name": "wall-stone"},
+ {"color": [7, 246, 231], "isthing": 0, "id": 176, "name": "wall-tile"},
+ {"color": [107, 255, 200], "isthing": 0, "id": 177, "name": "wall-wood"},
+ {"color": [58, 41, 149], "isthing": 0, "id": 178, "name": "water-other"},
+ {"color": [183, 121, 142], "isthing": 0, "id": 180, "name": "window-blind"},
+ {"color": [255, 73, 97], "isthing": 0, "id": 181, "name": "window-other"},
+ {"color": [107, 142, 35], "isthing": 0, "id": 184, "name": "tree-merged"},
+ {"color": [190, 153, 153], "isthing": 0, "id": 185, "name": "fence-merged"},
+ {"color": [146, 139, 141], "isthing": 0, "id": 186, "name": "ceiling-merged"},
+ {"color": [70, 130, 180], "isthing": 0, "id": 187, "name": "sky-other-merged"},
+ {"color": [134, 199, 156], "isthing": 0, "id": 188, "name": "cabinet-merged"},
+ {"color": [209, 226, 140], "isthing": 0, "id": 189, "name": "table-merged"},
+ {"color": [96, 36, 108], "isthing": 0, "id": 190, "name": "floor-other-merged"},
+ {"color": [96, 96, 96], "isthing": 0, "id": 191, "name": "pavement-merged"},
+ {"color": [64, 170, 64], "isthing": 0, "id": 192, "name": "mountain-merged"},
+ {"color": [152, 251, 152], "isthing": 0, "id": 193, "name": "grass-merged"},
+ {"color": [208, 229, 228], "isthing": 0, "id": 194, "name": "dirt-merged"},
+ {"color": [206, 186, 171], "isthing": 0, "id": 195, "name": "paper-merged"},
+ {"color": [152, 161, 64], "isthing": 0, "id": 196, "name": "food-other-merged"},
+ {"color": [116, 112, 0], "isthing": 0, "id": 197, "name": "building-other-merged"},
+ {"color": [0, 114, 143], "isthing": 0, "id": 198, "name": "rock-merged"},
+ {"color": [102, 102, 156], "isthing": 0, "id": 199, "name": "wall-other-merged"},
+ {"color": [250, 141, 255], "isthing": 0, "id": 200, "name": "rug-merged"},
+]
+
+# fmt: off
+COCO_PERSON_KEYPOINT_NAMES = (
+ "nose",
+ "left_eye", "right_eye",
+ "left_ear", "right_ear",
+ "left_shoulder", "right_shoulder",
+ "left_elbow", "right_elbow",
+ "left_wrist", "right_wrist",
+ "left_hip", "right_hip",
+ "left_knee", "right_knee",
+ "left_ankle", "right_ankle",
+)
+# fmt: on
+
+# Pairs of keypoints that should be exchanged under horizontal flipping
+COCO_PERSON_KEYPOINT_FLIP_MAP = (
+ ("left_eye", "right_eye"),
+ ("left_ear", "right_ear"),
+ ("left_shoulder", "right_shoulder"),
+ ("left_elbow", "right_elbow"),
+ ("left_wrist", "right_wrist"),
+ ("left_hip", "right_hip"),
+ ("left_knee", "right_knee"),
+ ("left_ankle", "right_ankle"),
+)
+
+# rules for pairs of keypoints to draw a line between, and the line color to use.
+KEYPOINT_CONNECTION_RULES = [
+ # face
+ ("left_ear", "left_eye", (102, 204, 255)),
+ ("right_ear", "right_eye", (51, 153, 255)),
+ ("left_eye", "nose", (102, 0, 204)),
+ ("nose", "right_eye", (51, 102, 255)),
+ # upper-body
+ ("left_shoulder", "right_shoulder", (255, 128, 0)),
+ ("left_shoulder", "left_elbow", (153, 255, 204)),
+ ("right_shoulder", "right_elbow", (128, 229, 255)),
+ ("left_elbow", "left_wrist", (153, 255, 153)),
+ ("right_elbow", "right_wrist", (102, 255, 224)),
+ # lower-body
+ ("left_hip", "right_hip", (255, 102, 0)),
+ ("left_hip", "left_knee", (255, 255, 77)),
+ ("right_hip", "right_knee", (153, 255, 204)),
+ ("left_knee", "left_ankle", (191, 255, 128)),
+ ("right_knee", "right_ankle", (255, 195, 77)),
+]
+
+# All Cityscapes categories, together with their nice-looking visualization colors
+# It's from https://github.com/mcordts/cityscapesScripts/blob/master/cityscapesscripts/helpers/labels.py # noqa
+CITYSCAPES_CATEGORIES = [
+ {"color": (128, 64, 128), "isthing": 0, "id": 7, "trainId": 0, "name": "road"},
+ {"color": (244, 35, 232), "isthing": 0, "id": 8, "trainId": 1, "name": "sidewalk"},
+ {"color": (70, 70, 70), "isthing": 0, "id": 11, "trainId": 2, "name": "building"},
+ {"color": (102, 102, 156), "isthing": 0, "id": 12, "trainId": 3, "name": "wall"},
+ {"color": (190, 153, 153), "isthing": 0, "id": 13, "trainId": 4, "name": "fence"},
+ {"color": (153, 153, 153), "isthing": 0, "id": 17, "trainId": 5, "name": "pole"},
+ {"color": (250, 170, 30), "isthing": 0, "id": 19, "trainId": 6, "name": "traffic light"},
+ {"color": (220, 220, 0), "isthing": 0, "id": 20, "trainId": 7, "name": "traffic sign"},
+ {"color": (107, 142, 35), "isthing": 0, "id": 21, "trainId": 8, "name": "vegetation"},
+ {"color": (152, 251, 152), "isthing": 0, "id": 22, "trainId": 9, "name": "terrain"},
+ {"color": (70, 130, 180), "isthing": 0, "id": 23, "trainId": 10, "name": "sky"},
+ {"color": (220, 20, 60), "isthing": 1, "id": 24, "trainId": 11, "name": "person"},
+ {"color": (255, 0, 0), "isthing": 1, "id": 25, "trainId": 12, "name": "rider"},
+ {"color": (0, 0, 142), "isthing": 1, "id": 26, "trainId": 13, "name": "car"},
+ {"color": (0, 0, 70), "isthing": 1, "id": 27, "trainId": 14, "name": "truck"},
+ {"color": (0, 60, 100), "isthing": 1, "id": 28, "trainId": 15, "name": "bus"},
+ {"color": (0, 80, 100), "isthing": 1, "id": 31, "trainId": 16, "name": "train"},
+ {"color": (0, 0, 230), "isthing": 1, "id": 32, "trainId": 17, "name": "motorcycle"},
+ {"color": (119, 11, 32), "isthing": 1, "id": 33, "trainId": 18, "name": "bicycle"},
+]
+
+# fmt: off
+ADE20K_SEM_SEG_CATEGORIES = [
+ "wall", "building", "sky", "floor", "tree", "ceiling", "road, route", "bed", "window ", "grass", "cabinet", "sidewalk, pavement", "person", "earth, ground", "door", "table", "mountain, mount", "plant", "curtain", "chair", "car", "water", "painting, picture", "sofa", "shelf", "house", "sea", "mirror", "rug", "field", "armchair", "seat", "fence", "desk", "rock, stone", "wardrobe, closet, press", "lamp", "tub", "rail", "cushion", "base, pedestal, stand", "box", "column, pillar", "signboard, sign", "chest of drawers, chest, bureau, dresser", "counter", "sand", "sink", "skyscraper", "fireplace", "refrigerator, icebox", "grandstand, covered stand", "path", "stairs", "runway", "case, display case, showcase, vitrine", "pool table, billiard table, snooker table", "pillow", "screen door, screen", "stairway, staircase", "river", "bridge, span", "bookcase", "blind, screen", "coffee table", "toilet, can, commode, crapper, pot, potty, stool, throne", "flower", "book", "hill", "bench", "countertop", "stove", "palm, palm tree", "kitchen island", "computer", "swivel chair", "boat", "bar", "arcade machine", "hovel, hut, hutch, shack, shanty", "bus", "towel", "light", "truck", "tower", "chandelier", "awning, sunshade, sunblind", "street lamp", "booth", "tv", "plane", "dirt track", "clothes", "pole", "land, ground, soil", "bannister, banister, balustrade, balusters, handrail", "escalator, moving staircase, moving stairway", "ottoman, pouf, pouffe, puff, hassock", "bottle", "buffet, counter, sideboard", "poster, posting, placard, notice, bill, card", "stage", "van", "ship", "fountain", "conveyer belt, conveyor belt, conveyer, conveyor, transporter", "canopy", "washer, automatic washer, washing machine", "plaything, toy", "pool", "stool", "barrel, cask", "basket, handbasket", "falls", "tent", "bag", "minibike, motorbike", "cradle", "oven", "ball", "food, solid food", "step, stair", "tank, storage tank", "trade name", "microwave", "pot", "animal", "bicycle", "lake", "dishwasher", "screen", "blanket, cover", "sculpture", "hood, exhaust hood", "sconce", "vase", "traffic light", "tray", "trash can", "fan", "pier", "crt screen", "plate", "monitor", "bulletin board", "shower", "radiator", "glass, drinking glass", "clock", "flag", # noqa
+]
+# After processed by `prepare_ade20k_sem_seg.py`, id 255 means ignore
+# fmt: on
+
+
+def _get_coco_instances_meta():
+ thing_ids = [k["id"] for k in COCO_CATEGORIES if k["isthing"] == 1]
+ thing_colors = [k["color"] for k in COCO_CATEGORIES if k["isthing"] == 1]
+ assert len(thing_ids) == 80, len(thing_ids)
+ # Mapping from the incontiguous COCO category id to an id in [0, 79]
+ thing_dataset_id_to_contiguous_id = {k: i for i, k in enumerate(thing_ids)}
+ thing_classes = [k["name"] for k in COCO_CATEGORIES if k["isthing"] == 1]
+ ret = {
+ "thing_dataset_id_to_contiguous_id": thing_dataset_id_to_contiguous_id,
+ "thing_classes": thing_classes,
+ "thing_colors": thing_colors,
+ }
+ return ret
+
+
+def _get_coco_panoptic_separated_meta():
+ """
+ Returns metadata for "separated" version of the panoptic segmentation dataset.
+ """
+ stuff_ids = [k["id"] for k in COCO_CATEGORIES if k["isthing"] == 0]
+ assert len(stuff_ids) == 53, len(stuff_ids)
+
+ # For semantic segmentation, this mapping maps from contiguous stuff id
+ # (in [0, 53], used in models) to ids in the dataset (used for processing results)
+ # The id 0 is mapped to an extra category "thing".
+ stuff_dataset_id_to_contiguous_id = {k: i + 1 for i, k in enumerate(stuff_ids)}
+ # When converting COCO panoptic annotations to semantic annotations
+ # We label the "thing" category to 0
+ stuff_dataset_id_to_contiguous_id[0] = 0
+
+ # 54 names for COCO stuff categories (including "things")
+ stuff_classes = ["things"] + [
+ k["name"].replace("-other", "").replace("-merged", "")
+ for k in COCO_CATEGORIES
+ if k["isthing"] == 0
+ ]
+
+ # NOTE: I randomly picked a color for things
+ stuff_colors = [[82, 18, 128]] + [k["color"] for k in COCO_CATEGORIES if k["isthing"] == 0]
+ ret = {
+ "stuff_dataset_id_to_contiguous_id": stuff_dataset_id_to_contiguous_id,
+ "stuff_classes": stuff_classes,
+ "stuff_colors": stuff_colors,
+ }
+ ret.update(_get_coco_instances_meta())
+ return ret
+
+
+def _get_builtin_metadata(dataset_name):
+ if dataset_name == "coco":
+ return _get_coco_instances_meta()
+ if dataset_name == "coco_panoptic_separated":
+ return _get_coco_panoptic_separated_meta()
+ elif dataset_name == "coco_panoptic_standard":
+ meta = {}
+ # The following metadata maps contiguous id from [0, #thing categories +
+ # #stuff categories) to their names and colors. We have to replica of the
+ # same name and color under "thing_*" and "stuff_*" because the current
+ # visualization function in D2 handles thing and class classes differently
+ # due to some heuristic used in Panoptic FPN. We keep the same naming to
+ # enable reusing existing visualization functions.
+ thing_classes = [k["name"] for k in COCO_CATEGORIES]
+ thing_colors = [k["color"] for k in COCO_CATEGORIES]
+ stuff_classes = [k["name"] for k in COCO_CATEGORIES]
+ stuff_colors = [k["color"] for k in COCO_CATEGORIES]
+
+ meta["thing_classes"] = thing_classes
+ meta["thing_colors"] = thing_colors
+ meta["stuff_classes"] = stuff_classes
+ meta["stuff_colors"] = stuff_colors
+
+ # Convert category id for training:
+ # category id: like semantic segmentation, it is the class id for each
+ # pixel. Since there are some classes not used in evaluation, the category
+ # id is not always contiguous and thus we have two set of category ids:
+ # - original category id: category id in the original dataset, mainly
+ # used for evaluation.
+ # - contiguous category id: [0, #classes), in order to train the linear
+ # softmax classifier.
+ thing_dataset_id_to_contiguous_id = {}
+ stuff_dataset_id_to_contiguous_id = {}
+
+ for i, cat in enumerate(COCO_CATEGORIES):
+ if cat["isthing"]:
+ thing_dataset_id_to_contiguous_id[cat["id"]] = i
+ else:
+ stuff_dataset_id_to_contiguous_id[cat["id"]] = i
+
+ meta["thing_dataset_id_to_contiguous_id"] = thing_dataset_id_to_contiguous_id
+ meta["stuff_dataset_id_to_contiguous_id"] = stuff_dataset_id_to_contiguous_id
+
+ return meta
+ elif dataset_name == "coco_person":
+ return {
+ "thing_classes": ["person"],
+ "keypoint_names": COCO_PERSON_KEYPOINT_NAMES,
+ "keypoint_flip_map": COCO_PERSON_KEYPOINT_FLIP_MAP,
+ "keypoint_connection_rules": KEYPOINT_CONNECTION_RULES,
+ }
+ elif dataset_name == "cityscapes":
+ # fmt: off
+ CITYSCAPES_THING_CLASSES = [
+ "person", "rider", "car", "truck",
+ "bus", "train", "motorcycle", "bicycle",
+ ]
+ CITYSCAPES_STUFF_CLASSES = [
+ "road", "sidewalk", "building", "wall", "fence", "pole", "traffic light",
+ "traffic sign", "vegetation", "terrain", "sky", "person", "rider", "car",
+ "truck", "bus", "train", "motorcycle", "bicycle",
+ ]
+ # fmt: on
+ return {
+ "thing_classes": CITYSCAPES_THING_CLASSES,
+ "stuff_classes": CITYSCAPES_STUFF_CLASSES,
+ }
+ raise KeyError("No built-in metadata for dataset {}".format(dataset_name))
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/cityscapes.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/cityscapes.py
new file mode 100644
index 0000000000000000000000000000000000000000..18c3f3a8279e2511016fa61885d26dad726ffe5e
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/cityscapes.py
@@ -0,0 +1,329 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import functools
+import json
+import logging
+import multiprocessing as mp
+import numpy as np
+import os
+from itertools import chain
+import pycocotools.mask as mask_util
+from PIL import Image
+
+from annotator.oneformer.detectron2.structures import BoxMode
+from annotator.oneformer.detectron2.utils.comm import get_world_size
+from annotator.oneformer.detectron2.utils.file_io import PathManager
+from annotator.oneformer.detectron2.utils.logger import setup_logger
+
+try:
+ import cv2 # noqa
+except ImportError:
+ # OpenCV is an optional dependency at the moment
+ pass
+
+
+logger = logging.getLogger(__name__)
+
+
+def _get_cityscapes_files(image_dir, gt_dir):
+ files = []
+ # scan through the directory
+ cities = PathManager.ls(image_dir)
+ logger.info(f"{len(cities)} cities found in '{image_dir}'.")
+ for city in cities:
+ city_img_dir = os.path.join(image_dir, city)
+ city_gt_dir = os.path.join(gt_dir, city)
+ for basename in PathManager.ls(city_img_dir):
+ image_file = os.path.join(city_img_dir, basename)
+
+ suffix = "leftImg8bit.png"
+ assert basename.endswith(suffix), basename
+ basename = basename[: -len(suffix)]
+
+ instance_file = os.path.join(city_gt_dir, basename + "gtFine_instanceIds.png")
+ label_file = os.path.join(city_gt_dir, basename + "gtFine_labelIds.png")
+ json_file = os.path.join(city_gt_dir, basename + "gtFine_polygons.json")
+
+ files.append((image_file, instance_file, label_file, json_file))
+ assert len(files), "No images found in {}".format(image_dir)
+ for f in files[0]:
+ assert PathManager.isfile(f), f
+ return files
+
+
+def load_cityscapes_instances(image_dir, gt_dir, from_json=True, to_polygons=True):
+ """
+ Args:
+ image_dir (str): path to the raw dataset. e.g., "~/cityscapes/leftImg8bit/train".
+ gt_dir (str): path to the raw annotations. e.g., "~/cityscapes/gtFine/train".
+ from_json (bool): whether to read annotations from the raw json file or the png files.
+ to_polygons (bool): whether to represent the segmentation as polygons
+ (COCO's format) instead of masks (cityscapes's format).
+
+ Returns:
+ list[dict]: a list of dicts in Detectron2 standard format. (See
+ `Using Custom Datasets `_ )
+ """
+ if from_json:
+ assert to_polygons, (
+ "Cityscapes's json annotations are in polygon format. "
+ "Converting to mask format is not supported now."
+ )
+ files = _get_cityscapes_files(image_dir, gt_dir)
+
+ logger.info("Preprocessing cityscapes annotations ...")
+ # This is still not fast: all workers will execute duplicate works and will
+ # take up to 10m on a 8GPU server.
+ pool = mp.Pool(processes=max(mp.cpu_count() // get_world_size() // 2, 4))
+
+ ret = pool.map(
+ functools.partial(_cityscapes_files_to_dict, from_json=from_json, to_polygons=to_polygons),
+ files,
+ )
+ logger.info("Loaded {} images from {}".format(len(ret), image_dir))
+
+ # Map cityscape ids to contiguous ids
+ from cityscapesscripts.helpers.labels import labels
+
+ labels = [l for l in labels if l.hasInstances and not l.ignoreInEval]
+ dataset_id_to_contiguous_id = {l.id: idx for idx, l in enumerate(labels)}
+ for dict_per_image in ret:
+ for anno in dict_per_image["annotations"]:
+ anno["category_id"] = dataset_id_to_contiguous_id[anno["category_id"]]
+ return ret
+
+
+def load_cityscapes_semantic(image_dir, gt_dir):
+ """
+ Args:
+ image_dir (str): path to the raw dataset. e.g., "~/cityscapes/leftImg8bit/train".
+ gt_dir (str): path to the raw annotations. e.g., "~/cityscapes/gtFine/train".
+
+ Returns:
+ list[dict]: a list of dict, each has "file_name" and
+ "sem_seg_file_name".
+ """
+ ret = []
+ # gt_dir is small and contain many small files. make sense to fetch to local first
+ gt_dir = PathManager.get_local_path(gt_dir)
+ for image_file, _, label_file, json_file in _get_cityscapes_files(image_dir, gt_dir):
+ label_file = label_file.replace("labelIds", "labelTrainIds")
+
+ with PathManager.open(json_file, "r") as f:
+ jsonobj = json.load(f)
+ ret.append(
+ {
+ "file_name": image_file,
+ "sem_seg_file_name": label_file,
+ "height": jsonobj["imgHeight"],
+ "width": jsonobj["imgWidth"],
+ }
+ )
+ assert len(ret), f"No images found in {image_dir}!"
+ assert PathManager.isfile(
+ ret[0]["sem_seg_file_name"]
+ ), "Please generate labelTrainIds.png with cityscapesscripts/preparation/createTrainIdLabelImgs.py" # noqa
+ return ret
+
+
+def _cityscapes_files_to_dict(files, from_json, to_polygons):
+ """
+ Parse cityscapes annotation files to a instance segmentation dataset dict.
+
+ Args:
+ files (tuple): consists of (image_file, instance_id_file, label_id_file, json_file)
+ from_json (bool): whether to read annotations from the raw json file or the png files.
+ to_polygons (bool): whether to represent the segmentation as polygons
+ (COCO's format) instead of masks (cityscapes's format).
+
+ Returns:
+ A dict in Detectron2 Dataset format.
+ """
+ from cityscapesscripts.helpers.labels import id2label, name2label
+
+ image_file, instance_id_file, _, json_file = files
+
+ annos = []
+
+ if from_json:
+ from shapely.geometry import MultiPolygon, Polygon
+
+ with PathManager.open(json_file, "r") as f:
+ jsonobj = json.load(f)
+ ret = {
+ "file_name": image_file,
+ "image_id": os.path.basename(image_file),
+ "height": jsonobj["imgHeight"],
+ "width": jsonobj["imgWidth"],
+ }
+
+ # `polygons_union` contains the union of all valid polygons.
+ polygons_union = Polygon()
+
+ # CityscapesScripts draw the polygons in sequential order
+ # and each polygon *overwrites* existing ones. See
+ # (https://github.com/mcordts/cityscapesScripts/blob/master/cityscapesscripts/preparation/json2instanceImg.py) # noqa
+ # We use reverse order, and each polygon *avoids* early ones.
+ # This will resolve the ploygon overlaps in the same way as CityscapesScripts.
+ for obj in jsonobj["objects"][::-1]:
+ if "deleted" in obj: # cityscapes data format specific
+ continue
+ label_name = obj["label"]
+
+ try:
+ label = name2label[label_name]
+ except KeyError:
+ if label_name.endswith("group"): # crowd area
+ label = name2label[label_name[: -len("group")]]
+ else:
+ raise
+ if label.id < 0: # cityscapes data format
+ continue
+
+ # Cityscapes's raw annotations uses integer coordinates
+ # Therefore +0.5 here
+ poly_coord = np.asarray(obj["polygon"], dtype="f4") + 0.5
+ # CityscapesScript uses PIL.ImageDraw.polygon to rasterize
+ # polygons for evaluation. This function operates in integer space
+ # and draws each pixel whose center falls into the polygon.
+ # Therefore it draws a polygon which is 0.5 "fatter" in expectation.
+ # We therefore dilate the input polygon by 0.5 as our input.
+ poly = Polygon(poly_coord).buffer(0.5, resolution=4)
+
+ if not label.hasInstances or label.ignoreInEval:
+ # even if we won't store the polygon it still contributes to overlaps resolution
+ polygons_union = polygons_union.union(poly)
+ continue
+
+ # Take non-overlapping part of the polygon
+ poly_wo_overlaps = poly.difference(polygons_union)
+ if poly_wo_overlaps.is_empty:
+ continue
+ polygons_union = polygons_union.union(poly)
+
+ anno = {}
+ anno["iscrowd"] = label_name.endswith("group")
+ anno["category_id"] = label.id
+
+ if isinstance(poly_wo_overlaps, Polygon):
+ poly_list = [poly_wo_overlaps]
+ elif isinstance(poly_wo_overlaps, MultiPolygon):
+ poly_list = poly_wo_overlaps.geoms
+ else:
+ raise NotImplementedError("Unknown geometric structure {}".format(poly_wo_overlaps))
+
+ poly_coord = []
+ for poly_el in poly_list:
+ # COCO API can work only with exterior boundaries now, hence we store only them.
+ # TODO: store both exterior and interior boundaries once other parts of the
+ # codebase support holes in polygons.
+ poly_coord.append(list(chain(*poly_el.exterior.coords)))
+ anno["segmentation"] = poly_coord
+ (xmin, ymin, xmax, ymax) = poly_wo_overlaps.bounds
+
+ anno["bbox"] = (xmin, ymin, xmax, ymax)
+ anno["bbox_mode"] = BoxMode.XYXY_ABS
+
+ annos.append(anno)
+ else:
+ # See also the official annotation parsing scripts at
+ # https://github.com/mcordts/cityscapesScripts/blob/master/cityscapesscripts/evaluation/instances2dict.py # noqa
+ with PathManager.open(instance_id_file, "rb") as f:
+ inst_image = np.asarray(Image.open(f), order="F")
+ # ids < 24 are stuff labels (filtering them first is about 5% faster)
+ flattened_ids = np.unique(inst_image[inst_image >= 24])
+
+ ret = {
+ "file_name": image_file,
+ "image_id": os.path.basename(image_file),
+ "height": inst_image.shape[0],
+ "width": inst_image.shape[1],
+ }
+
+ for instance_id in flattened_ids:
+ # For non-crowd annotations, instance_id // 1000 is the label_id
+ # Crowd annotations have <1000 instance ids
+ label_id = instance_id // 1000 if instance_id >= 1000 else instance_id
+ label = id2label[label_id]
+ if not label.hasInstances or label.ignoreInEval:
+ continue
+
+ anno = {}
+ anno["iscrowd"] = instance_id < 1000
+ anno["category_id"] = label.id
+
+ mask = np.asarray(inst_image == instance_id, dtype=np.uint8, order="F")
+
+ inds = np.nonzero(mask)
+ ymin, ymax = inds[0].min(), inds[0].max()
+ xmin, xmax = inds[1].min(), inds[1].max()
+ anno["bbox"] = (xmin, ymin, xmax, ymax)
+ if xmax <= xmin or ymax <= ymin:
+ continue
+ anno["bbox_mode"] = BoxMode.XYXY_ABS
+ if to_polygons:
+ # This conversion comes from D4809743 and D5171122,
+ # when Mask-RCNN was first developed.
+ contours = cv2.findContours(mask.copy(), cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_NONE)[
+ -2
+ ]
+ polygons = [c.reshape(-1).tolist() for c in contours if len(c) >= 3]
+ # opencv's can produce invalid polygons
+ if len(polygons) == 0:
+ continue
+ anno["segmentation"] = polygons
+ else:
+ anno["segmentation"] = mask_util.encode(mask[:, :, None])[0]
+ annos.append(anno)
+ ret["annotations"] = annos
+ return ret
+
+
+if __name__ == "__main__":
+ """
+ Test the cityscapes dataset loader.
+
+ Usage:
+ python -m detectron2.data.datasets.cityscapes \
+ cityscapes/leftImg8bit/train cityscapes/gtFine/train
+ """
+ import argparse
+
+ parser = argparse.ArgumentParser()
+ parser.add_argument("image_dir")
+ parser.add_argument("gt_dir")
+ parser.add_argument("--type", choices=["instance", "semantic"], default="instance")
+ args = parser.parse_args()
+ from annotator.oneformer.detectron2.data.catalog import Metadata
+ from annotator.oneformer.detectron2.utils.visualizer import Visualizer
+ from cityscapesscripts.helpers.labels import labels
+
+ logger = setup_logger(name=__name__)
+
+ dirname = "cityscapes-data-vis"
+ os.makedirs(dirname, exist_ok=True)
+
+ if args.type == "instance":
+ dicts = load_cityscapes_instances(
+ args.image_dir, args.gt_dir, from_json=True, to_polygons=True
+ )
+ logger.info("Done loading {} samples.".format(len(dicts)))
+
+ thing_classes = [k.name for k in labels if k.hasInstances and not k.ignoreInEval]
+ meta = Metadata().set(thing_classes=thing_classes)
+
+ else:
+ dicts = load_cityscapes_semantic(args.image_dir, args.gt_dir)
+ logger.info("Done loading {} samples.".format(len(dicts)))
+
+ stuff_classes = [k.name for k in labels if k.trainId != 255]
+ stuff_colors = [k.color for k in labels if k.trainId != 255]
+ meta = Metadata().set(stuff_classes=stuff_classes, stuff_colors=stuff_colors)
+
+ for d in dicts:
+ img = np.array(Image.open(PathManager.open(d["file_name"], "rb")))
+ visualizer = Visualizer(img, metadata=meta)
+ vis = visualizer.draw_dataset_dict(d)
+ # cv2.imshow("a", vis.get_image()[:, :, ::-1])
+ # cv2.waitKey()
+ fpath = os.path.join(dirname, os.path.basename(d["file_name"]))
+ vis.save(fpath)
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/cityscapes_panoptic.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/cityscapes_panoptic.py
new file mode 100644
index 0000000000000000000000000000000000000000..7ce9ec48f673dadf3f5b4ae0592fc82415d9f925
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/cityscapes_panoptic.py
@@ -0,0 +1,187 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import json
+import logging
+import os
+
+from annotator.oneformer.detectron2.data import DatasetCatalog, MetadataCatalog
+from annotator.oneformer.detectron2.data.datasets.builtin_meta import CITYSCAPES_CATEGORIES
+from annotator.oneformer.detectron2.utils.file_io import PathManager
+
+"""
+This file contains functions to register the Cityscapes panoptic dataset to the DatasetCatalog.
+"""
+
+
+logger = logging.getLogger(__name__)
+
+
+def get_cityscapes_panoptic_files(image_dir, gt_dir, json_info):
+ files = []
+ # scan through the directory
+ cities = PathManager.ls(image_dir)
+ logger.info(f"{len(cities)} cities found in '{image_dir}'.")
+ image_dict = {}
+ for city in cities:
+ city_img_dir = os.path.join(image_dir, city)
+ for basename in PathManager.ls(city_img_dir):
+ image_file = os.path.join(city_img_dir, basename)
+
+ suffix = "_leftImg8bit.png"
+ assert basename.endswith(suffix), basename
+ basename = os.path.basename(basename)[: -len(suffix)]
+
+ image_dict[basename] = image_file
+
+ for ann in json_info["annotations"]:
+ image_file = image_dict.get(ann["image_id"], None)
+ assert image_file is not None, "No image {} found for annotation {}".format(
+ ann["image_id"], ann["file_name"]
+ )
+ label_file = os.path.join(gt_dir, ann["file_name"])
+ segments_info = ann["segments_info"]
+
+ files.append((image_file, label_file, segments_info))
+
+ assert len(files), "No images found in {}".format(image_dir)
+ assert PathManager.isfile(files[0][0]), files[0][0]
+ assert PathManager.isfile(files[0][1]), files[0][1]
+ return files
+
+
+def load_cityscapes_panoptic(image_dir, gt_dir, gt_json, meta):
+ """
+ Args:
+ image_dir (str): path to the raw dataset. e.g., "~/cityscapes/leftImg8bit/train".
+ gt_dir (str): path to the raw annotations. e.g.,
+ "~/cityscapes/gtFine/cityscapes_panoptic_train".
+ gt_json (str): path to the json file. e.g.,
+ "~/cityscapes/gtFine/cityscapes_panoptic_train.json".
+ meta (dict): dictionary containing "thing_dataset_id_to_contiguous_id"
+ and "stuff_dataset_id_to_contiguous_id" to map category ids to
+ contiguous ids for training.
+
+ Returns:
+ list[dict]: a list of dicts in Detectron2 standard format. (See
+ `Using Custom Datasets `_ )
+ """
+
+ def _convert_category_id(segment_info, meta):
+ if segment_info["category_id"] in meta["thing_dataset_id_to_contiguous_id"]:
+ segment_info["category_id"] = meta["thing_dataset_id_to_contiguous_id"][
+ segment_info["category_id"]
+ ]
+ else:
+ segment_info["category_id"] = meta["stuff_dataset_id_to_contiguous_id"][
+ segment_info["category_id"]
+ ]
+ return segment_info
+
+ assert os.path.exists(
+ gt_json
+ ), "Please run `python cityscapesscripts/preparation/createPanopticImgs.py` to generate label files." # noqa
+ with open(gt_json) as f:
+ json_info = json.load(f)
+ files = get_cityscapes_panoptic_files(image_dir, gt_dir, json_info)
+ ret = []
+ for image_file, label_file, segments_info in files:
+ sem_label_file = (
+ image_file.replace("leftImg8bit", "gtFine").split(".")[0] + "_labelTrainIds.png"
+ )
+ segments_info = [_convert_category_id(x, meta) for x in segments_info]
+ ret.append(
+ {
+ "file_name": image_file,
+ "image_id": "_".join(
+ os.path.splitext(os.path.basename(image_file))[0].split("_")[:3]
+ ),
+ "sem_seg_file_name": sem_label_file,
+ "pan_seg_file_name": label_file,
+ "segments_info": segments_info,
+ }
+ )
+ assert len(ret), f"No images found in {image_dir}!"
+ assert PathManager.isfile(
+ ret[0]["sem_seg_file_name"]
+ ), "Please generate labelTrainIds.png with cityscapesscripts/preparation/createTrainIdLabelImgs.py" # noqa
+ assert PathManager.isfile(
+ ret[0]["pan_seg_file_name"]
+ ), "Please generate panoptic annotation with python cityscapesscripts/preparation/createPanopticImgs.py" # noqa
+ return ret
+
+
+_RAW_CITYSCAPES_PANOPTIC_SPLITS = {
+ "cityscapes_fine_panoptic_train": (
+ "cityscapes/leftImg8bit/train",
+ "cityscapes/gtFine/cityscapes_panoptic_train",
+ "cityscapes/gtFine/cityscapes_panoptic_train.json",
+ ),
+ "cityscapes_fine_panoptic_val": (
+ "cityscapes/leftImg8bit/val",
+ "cityscapes/gtFine/cityscapes_panoptic_val",
+ "cityscapes/gtFine/cityscapes_panoptic_val.json",
+ ),
+ # "cityscapes_fine_panoptic_test": not supported yet
+}
+
+
+def register_all_cityscapes_panoptic(root):
+ meta = {}
+ # The following metadata maps contiguous id from [0, #thing categories +
+ # #stuff categories) to their names and colors. We have to replica of the
+ # same name and color under "thing_*" and "stuff_*" because the current
+ # visualization function in D2 handles thing and class classes differently
+ # due to some heuristic used in Panoptic FPN. We keep the same naming to
+ # enable reusing existing visualization functions.
+ thing_classes = [k["name"] for k in CITYSCAPES_CATEGORIES]
+ thing_colors = [k["color"] for k in CITYSCAPES_CATEGORIES]
+ stuff_classes = [k["name"] for k in CITYSCAPES_CATEGORIES]
+ stuff_colors = [k["color"] for k in CITYSCAPES_CATEGORIES]
+
+ meta["thing_classes"] = thing_classes
+ meta["thing_colors"] = thing_colors
+ meta["stuff_classes"] = stuff_classes
+ meta["stuff_colors"] = stuff_colors
+
+ # There are three types of ids in cityscapes panoptic segmentation:
+ # (1) category id: like semantic segmentation, it is the class id for each
+ # pixel. Since there are some classes not used in evaluation, the category
+ # id is not always contiguous and thus we have two set of category ids:
+ # - original category id: category id in the original dataset, mainly
+ # used for evaluation.
+ # - contiguous category id: [0, #classes), in order to train the classifier
+ # (2) instance id: this id is used to differentiate different instances from
+ # the same category. For "stuff" classes, the instance id is always 0; for
+ # "thing" classes, the instance id starts from 1 and 0 is reserved for
+ # ignored instances (e.g. crowd annotation).
+ # (3) panoptic id: this is the compact id that encode both category and
+ # instance id by: category_id * 1000 + instance_id.
+ thing_dataset_id_to_contiguous_id = {}
+ stuff_dataset_id_to_contiguous_id = {}
+
+ for k in CITYSCAPES_CATEGORIES:
+ if k["isthing"] == 1:
+ thing_dataset_id_to_contiguous_id[k["id"]] = k["trainId"]
+ else:
+ stuff_dataset_id_to_contiguous_id[k["id"]] = k["trainId"]
+
+ meta["thing_dataset_id_to_contiguous_id"] = thing_dataset_id_to_contiguous_id
+ meta["stuff_dataset_id_to_contiguous_id"] = stuff_dataset_id_to_contiguous_id
+
+ for key, (image_dir, gt_dir, gt_json) in _RAW_CITYSCAPES_PANOPTIC_SPLITS.items():
+ image_dir = os.path.join(root, image_dir)
+ gt_dir = os.path.join(root, gt_dir)
+ gt_json = os.path.join(root, gt_json)
+
+ DatasetCatalog.register(
+ key, lambda x=image_dir, y=gt_dir, z=gt_json: load_cityscapes_panoptic(x, y, z, meta)
+ )
+ MetadataCatalog.get(key).set(
+ panoptic_root=gt_dir,
+ image_root=image_dir,
+ panoptic_json=gt_json,
+ gt_dir=gt_dir.replace("cityscapes_panoptic_", ""),
+ evaluator_type="cityscapes_panoptic_seg",
+ ignore_label=255,
+ label_divisor=1000,
+ **meta,
+ )
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/coco.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/coco.py
new file mode 100644
index 0000000000000000000000000000000000000000..b0b2956b27568e5926a5d35adf0106fba1cd96b9
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/coco.py
@@ -0,0 +1,539 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import contextlib
+import datetime
+import io
+import json
+import logging
+import numpy as np
+import os
+import shutil
+import pycocotools.mask as mask_util
+from fvcore.common.timer import Timer
+from iopath.common.file_io import file_lock
+from PIL import Image
+
+from annotator.oneformer.detectron2.structures import Boxes, BoxMode, PolygonMasks, RotatedBoxes
+from annotator.oneformer.detectron2.utils.file_io import PathManager
+
+from .. import DatasetCatalog, MetadataCatalog
+
+"""
+This file contains functions to parse COCO-format annotations into dicts in "Detectron2 format".
+"""
+
+
+logger = logging.getLogger(__name__)
+
+__all__ = ["load_coco_json", "load_sem_seg", "convert_to_coco_json", "register_coco_instances"]
+
+
+def load_coco_json(json_file, image_root, dataset_name=None, extra_annotation_keys=None):
+ """
+ Load a json file with COCO's instances annotation format.
+ Currently supports instance detection, instance segmentation,
+ and person keypoints annotations.
+
+ Args:
+ json_file (str): full path to the json file in COCO instances annotation format.
+ image_root (str or path-like): the directory where the images in this json file exists.
+ dataset_name (str or None): the name of the dataset (e.g., coco_2017_train).
+ When provided, this function will also do the following:
+
+ * Put "thing_classes" into the metadata associated with this dataset.
+ * Map the category ids into a contiguous range (needed by standard dataset format),
+ and add "thing_dataset_id_to_contiguous_id" to the metadata associated
+ with this dataset.
+
+ This option should usually be provided, unless users need to load
+ the original json content and apply more processing manually.
+ extra_annotation_keys (list[str]): list of per-annotation keys that should also be
+ loaded into the dataset dict (besides "iscrowd", "bbox", "keypoints",
+ "category_id", "segmentation"). The values for these keys will be returned as-is.
+ For example, the densepose annotations are loaded in this way.
+
+ Returns:
+ list[dict]: a list of dicts in Detectron2 standard dataset dicts format (See
+ `Using Custom Datasets `_ ) when `dataset_name` is not None.
+ If `dataset_name` is None, the returned `category_ids` may be
+ incontiguous and may not conform to the Detectron2 standard format.
+
+ Notes:
+ 1. This function does not read the image files.
+ The results do not have the "image" field.
+ """
+ from pycocotools.coco import COCO
+
+ timer = Timer()
+ json_file = PathManager.get_local_path(json_file)
+ with contextlib.redirect_stdout(io.StringIO()):
+ coco_api = COCO(json_file)
+ if timer.seconds() > 1:
+ logger.info("Loading {} takes {:.2f} seconds.".format(json_file, timer.seconds()))
+
+ id_map = None
+ if dataset_name is not None:
+ meta = MetadataCatalog.get(dataset_name)
+ cat_ids = sorted(coco_api.getCatIds())
+ cats = coco_api.loadCats(cat_ids)
+ # The categories in a custom json file may not be sorted.
+ thing_classes = [c["name"] for c in sorted(cats, key=lambda x: x["id"])]
+ meta.thing_classes = thing_classes
+
+ # In COCO, certain category ids are artificially removed,
+ # and by convention they are always ignored.
+ # We deal with COCO's id issue and translate
+ # the category ids to contiguous ids in [0, 80).
+
+ # It works by looking at the "categories" field in the json, therefore
+ # if users' own json also have incontiguous ids, we'll
+ # apply this mapping as well but print a warning.
+ if not (min(cat_ids) == 1 and max(cat_ids) == len(cat_ids)):
+ if "coco" not in dataset_name:
+ logger.warning(
+ """
+Category ids in annotations are not in [1, #categories]! We'll apply a mapping for you.
+"""
+ )
+ id_map = {v: i for i, v in enumerate(cat_ids)}
+ meta.thing_dataset_id_to_contiguous_id = id_map
+
+ # sort indices for reproducible results
+ img_ids = sorted(coco_api.imgs.keys())
+ # imgs is a list of dicts, each looks something like:
+ # {'license': 4,
+ # 'url': 'http://farm6.staticflickr.com/5454/9413846304_881d5e5c3b_z.jpg',
+ # 'file_name': 'COCO_val2014_000000001268.jpg',
+ # 'height': 427,
+ # 'width': 640,
+ # 'date_captured': '2013-11-17 05:57:24',
+ # 'id': 1268}
+ imgs = coco_api.loadImgs(img_ids)
+ # anns is a list[list[dict]], where each dict is an annotation
+ # record for an object. The inner list enumerates the objects in an image
+ # and the outer list enumerates over images. Example of anns[0]:
+ # [{'segmentation': [[192.81,
+ # 247.09,
+ # ...
+ # 219.03,
+ # 249.06]],
+ # 'area': 1035.749,
+ # 'iscrowd': 0,
+ # 'image_id': 1268,
+ # 'bbox': [192.81, 224.8, 74.73, 33.43],
+ # 'category_id': 16,
+ # 'id': 42986},
+ # ...]
+ anns = [coco_api.imgToAnns[img_id] for img_id in img_ids]
+ total_num_valid_anns = sum([len(x) for x in anns])
+ total_num_anns = len(coco_api.anns)
+ if total_num_valid_anns < total_num_anns:
+ logger.warning(
+ f"{json_file} contains {total_num_anns} annotations, but only "
+ f"{total_num_valid_anns} of them match to images in the file."
+ )
+
+ if "minival" not in json_file:
+ # The popular valminusminival & minival annotations for COCO2014 contain this bug.
+ # However the ratio of buggy annotations there is tiny and does not affect accuracy.
+ # Therefore we explicitly white-list them.
+ ann_ids = [ann["id"] for anns_per_image in anns for ann in anns_per_image]
+ assert len(set(ann_ids)) == len(ann_ids), "Annotation ids in '{}' are not unique!".format(
+ json_file
+ )
+
+ imgs_anns = list(zip(imgs, anns))
+ logger.info("Loaded {} images in COCO format from {}".format(len(imgs_anns), json_file))
+
+ dataset_dicts = []
+
+ ann_keys = ["iscrowd", "bbox", "keypoints", "category_id"] + (extra_annotation_keys or [])
+
+ num_instances_without_valid_segmentation = 0
+
+ for (img_dict, anno_dict_list) in imgs_anns:
+ record = {}
+ record["file_name"] = os.path.join(image_root, img_dict["file_name"])
+ record["height"] = img_dict["height"]
+ record["width"] = img_dict["width"]
+ image_id = record["image_id"] = img_dict["id"]
+
+ objs = []
+ for anno in anno_dict_list:
+ # Check that the image_id in this annotation is the same as
+ # the image_id we're looking at.
+ # This fails only when the data parsing logic or the annotation file is buggy.
+
+ # The original COCO valminusminival2014 & minival2014 annotation files
+ # actually contains bugs that, together with certain ways of using COCO API,
+ # can trigger this assertion.
+ assert anno["image_id"] == image_id
+
+ assert anno.get("ignore", 0) == 0, '"ignore" in COCO json file is not supported.'
+
+ obj = {key: anno[key] for key in ann_keys if key in anno}
+ if "bbox" in obj and len(obj["bbox"]) == 0:
+ raise ValueError(
+ f"One annotation of image {image_id} contains empty 'bbox' value! "
+ "This json does not have valid COCO format."
+ )
+
+ segm = anno.get("segmentation", None)
+ if segm: # either list[list[float]] or dict(RLE)
+ if isinstance(segm, dict):
+ if isinstance(segm["counts"], list):
+ # convert to compressed RLE
+ segm = mask_util.frPyObjects(segm, *segm["size"])
+ else:
+ # filter out invalid polygons (< 3 points)
+ segm = [poly for poly in segm if len(poly) % 2 == 0 and len(poly) >= 6]
+ if len(segm) == 0:
+ num_instances_without_valid_segmentation += 1
+ continue # ignore this instance
+ obj["segmentation"] = segm
+
+ keypts = anno.get("keypoints", None)
+ if keypts: # list[int]
+ for idx, v in enumerate(keypts):
+ if idx % 3 != 2:
+ # COCO's segmentation coordinates are floating points in [0, H or W],
+ # but keypoint coordinates are integers in [0, H-1 or W-1]
+ # Therefore we assume the coordinates are "pixel indices" and
+ # add 0.5 to convert to floating point coordinates.
+ keypts[idx] = v + 0.5
+ obj["keypoints"] = keypts
+
+ obj["bbox_mode"] = BoxMode.XYWH_ABS
+ if id_map:
+ annotation_category_id = obj["category_id"]
+ try:
+ obj["category_id"] = id_map[annotation_category_id]
+ except KeyError as e:
+ raise KeyError(
+ f"Encountered category_id={annotation_category_id} "
+ "but this id does not exist in 'categories' of the json file."
+ ) from e
+ objs.append(obj)
+ record["annotations"] = objs
+ dataset_dicts.append(record)
+
+ if num_instances_without_valid_segmentation > 0:
+ logger.warning(
+ "Filtered out {} instances without valid segmentation. ".format(
+ num_instances_without_valid_segmentation
+ )
+ + "There might be issues in your dataset generation process. Please "
+ "check https://detectron2.readthedocs.io/en/latest/tutorials/datasets.html carefully"
+ )
+ return dataset_dicts
+
+
+def load_sem_seg(gt_root, image_root, gt_ext="png", image_ext="jpg"):
+ """
+ Load semantic segmentation datasets. All files under "gt_root" with "gt_ext" extension are
+ treated as ground truth annotations and all files under "image_root" with "image_ext" extension
+ as input images. Ground truth and input images are matched using file paths relative to
+ "gt_root" and "image_root" respectively without taking into account file extensions.
+ This works for COCO as well as some other datasets.
+
+ Args:
+ gt_root (str): full path to ground truth semantic segmentation files. Semantic segmentation
+ annotations are stored as images with integer values in pixels that represent
+ corresponding semantic labels.
+ image_root (str): the directory where the input images are.
+ gt_ext (str): file extension for ground truth annotations.
+ image_ext (str): file extension for input images.
+
+ Returns:
+ list[dict]:
+ a list of dicts in detectron2 standard format without instance-level
+ annotation.
+
+ Notes:
+ 1. This function does not read the image and ground truth files.
+ The results do not have the "image" and "sem_seg" fields.
+ """
+
+ # We match input images with ground truth based on their relative filepaths (without file
+ # extensions) starting from 'image_root' and 'gt_root' respectively.
+ def file2id(folder_path, file_path):
+ # extract relative path starting from `folder_path`
+ image_id = os.path.normpath(os.path.relpath(file_path, start=folder_path))
+ # remove file extension
+ image_id = os.path.splitext(image_id)[0]
+ return image_id
+
+ input_files = sorted(
+ (os.path.join(image_root, f) for f in PathManager.ls(image_root) if f.endswith(image_ext)),
+ key=lambda file_path: file2id(image_root, file_path),
+ )
+ gt_files = sorted(
+ (os.path.join(gt_root, f) for f in PathManager.ls(gt_root) if f.endswith(gt_ext)),
+ key=lambda file_path: file2id(gt_root, file_path),
+ )
+
+ assert len(gt_files) > 0, "No annotations found in {}.".format(gt_root)
+
+ # Use the intersection, so that val2017_100 annotations can run smoothly with val2017 images
+ if len(input_files) != len(gt_files):
+ logger.warn(
+ "Directory {} and {} has {} and {} files, respectively.".format(
+ image_root, gt_root, len(input_files), len(gt_files)
+ )
+ )
+ input_basenames = [os.path.basename(f)[: -len(image_ext)] for f in input_files]
+ gt_basenames = [os.path.basename(f)[: -len(gt_ext)] for f in gt_files]
+ intersect = list(set(input_basenames) & set(gt_basenames))
+ # sort, otherwise each worker may obtain a list[dict] in different order
+ intersect = sorted(intersect)
+ logger.warn("Will use their intersection of {} files.".format(len(intersect)))
+ input_files = [os.path.join(image_root, f + image_ext) for f in intersect]
+ gt_files = [os.path.join(gt_root, f + gt_ext) for f in intersect]
+
+ logger.info(
+ "Loaded {} images with semantic segmentation from {}".format(len(input_files), image_root)
+ )
+
+ dataset_dicts = []
+ for (img_path, gt_path) in zip(input_files, gt_files):
+ record = {}
+ record["file_name"] = img_path
+ record["sem_seg_file_name"] = gt_path
+ dataset_dicts.append(record)
+
+ return dataset_dicts
+
+
+def convert_to_coco_dict(dataset_name):
+ """
+ Convert an instance detection/segmentation or keypoint detection dataset
+ in detectron2's standard format into COCO json format.
+
+ Generic dataset description can be found here:
+ https://detectron2.readthedocs.io/tutorials/datasets.html#register-a-dataset
+
+ COCO data format description can be found here:
+ http://cocodataset.org/#format-data
+
+ Args:
+ dataset_name (str):
+ name of the source dataset
+ Must be registered in DatastCatalog and in detectron2's standard format.
+ Must have corresponding metadata "thing_classes"
+ Returns:
+ coco_dict: serializable dict in COCO json format
+ """
+
+ dataset_dicts = DatasetCatalog.get(dataset_name)
+ metadata = MetadataCatalog.get(dataset_name)
+
+ # unmap the category mapping ids for COCO
+ if hasattr(metadata, "thing_dataset_id_to_contiguous_id"):
+ reverse_id_mapping = {v: k for k, v in metadata.thing_dataset_id_to_contiguous_id.items()}
+ reverse_id_mapper = lambda contiguous_id: reverse_id_mapping[contiguous_id] # noqa
+ else:
+ reverse_id_mapper = lambda contiguous_id: contiguous_id # noqa
+
+ categories = [
+ {"id": reverse_id_mapper(id), "name": name}
+ for id, name in enumerate(metadata.thing_classes)
+ ]
+
+ logger.info("Converting dataset dicts into COCO format")
+ coco_images = []
+ coco_annotations = []
+
+ for image_id, image_dict in enumerate(dataset_dicts):
+ coco_image = {
+ "id": image_dict.get("image_id", image_id),
+ "width": int(image_dict["width"]),
+ "height": int(image_dict["height"]),
+ "file_name": str(image_dict["file_name"]),
+ }
+ coco_images.append(coco_image)
+
+ anns_per_image = image_dict.get("annotations", [])
+ for annotation in anns_per_image:
+ # create a new dict with only COCO fields
+ coco_annotation = {}
+
+ # COCO requirement: XYWH box format for axis-align and XYWHA for rotated
+ bbox = annotation["bbox"]
+ if isinstance(bbox, np.ndarray):
+ if bbox.ndim != 1:
+ raise ValueError(f"bbox has to be 1-dimensional. Got shape={bbox.shape}.")
+ bbox = bbox.tolist()
+ if len(bbox) not in [4, 5]:
+ raise ValueError(f"bbox has to has length 4 or 5. Got {bbox}.")
+ from_bbox_mode = annotation["bbox_mode"]
+ to_bbox_mode = BoxMode.XYWH_ABS if len(bbox) == 4 else BoxMode.XYWHA_ABS
+ bbox = BoxMode.convert(bbox, from_bbox_mode, to_bbox_mode)
+
+ # COCO requirement: instance area
+ if "segmentation" in annotation:
+ # Computing areas for instances by counting the pixels
+ segmentation = annotation["segmentation"]
+ # TODO: check segmentation type: RLE, BinaryMask or Polygon
+ if isinstance(segmentation, list):
+ polygons = PolygonMasks([segmentation])
+ area = polygons.area()[0].item()
+ elif isinstance(segmentation, dict): # RLE
+ area = mask_util.area(segmentation).item()
+ else:
+ raise TypeError(f"Unknown segmentation type {type(segmentation)}!")
+ else:
+ # Computing areas using bounding boxes
+ if to_bbox_mode == BoxMode.XYWH_ABS:
+ bbox_xy = BoxMode.convert(bbox, to_bbox_mode, BoxMode.XYXY_ABS)
+ area = Boxes([bbox_xy]).area()[0].item()
+ else:
+ area = RotatedBoxes([bbox]).area()[0].item()
+
+ if "keypoints" in annotation:
+ keypoints = annotation["keypoints"] # list[int]
+ for idx, v in enumerate(keypoints):
+ if idx % 3 != 2:
+ # COCO's segmentation coordinates are floating points in [0, H or W],
+ # but keypoint coordinates are integers in [0, H-1 or W-1]
+ # For COCO format consistency we substract 0.5
+ # https://github.com/facebookresearch/detectron2/pull/175#issuecomment-551202163
+ keypoints[idx] = v - 0.5
+ if "num_keypoints" in annotation:
+ num_keypoints = annotation["num_keypoints"]
+ else:
+ num_keypoints = sum(kp > 0 for kp in keypoints[2::3])
+
+ # COCO requirement:
+ # linking annotations to images
+ # "id" field must start with 1
+ coco_annotation["id"] = len(coco_annotations) + 1
+ coco_annotation["image_id"] = coco_image["id"]
+ coco_annotation["bbox"] = [round(float(x), 3) for x in bbox]
+ coco_annotation["area"] = float(area)
+ coco_annotation["iscrowd"] = int(annotation.get("iscrowd", 0))
+ coco_annotation["category_id"] = int(reverse_id_mapper(annotation["category_id"]))
+
+ # Add optional fields
+ if "keypoints" in annotation:
+ coco_annotation["keypoints"] = keypoints
+ coco_annotation["num_keypoints"] = num_keypoints
+
+ if "segmentation" in annotation:
+ seg = coco_annotation["segmentation"] = annotation["segmentation"]
+ if isinstance(seg, dict): # RLE
+ counts = seg["counts"]
+ if not isinstance(counts, str):
+ # make it json-serializable
+ seg["counts"] = counts.decode("ascii")
+
+ coco_annotations.append(coco_annotation)
+
+ logger.info(
+ "Conversion finished, "
+ f"#images: {len(coco_images)}, #annotations: {len(coco_annotations)}"
+ )
+
+ info = {
+ "date_created": str(datetime.datetime.now()),
+ "description": "Automatically generated COCO json file for Detectron2.",
+ }
+ coco_dict = {"info": info, "images": coco_images, "categories": categories, "licenses": None}
+ if len(coco_annotations) > 0:
+ coco_dict["annotations"] = coco_annotations
+ return coco_dict
+
+
+def convert_to_coco_json(dataset_name, output_file, allow_cached=True):
+ """
+ Converts dataset into COCO format and saves it to a json file.
+ dataset_name must be registered in DatasetCatalog and in detectron2's standard format.
+
+ Args:
+ dataset_name:
+ reference from the config file to the catalogs
+ must be registered in DatasetCatalog and in detectron2's standard format
+ output_file: path of json file that will be saved to
+ allow_cached: if json file is already present then skip conversion
+ """
+
+ # TODO: The dataset or the conversion script *may* change,
+ # a checksum would be useful for validating the cached data
+
+ PathManager.mkdirs(os.path.dirname(output_file))
+ with file_lock(output_file):
+ if PathManager.exists(output_file) and allow_cached:
+ logger.warning(
+ f"Using previously cached COCO format annotations at '{output_file}'. "
+ "You need to clear the cache file if your dataset has been modified."
+ )
+ else:
+ logger.info(f"Converting annotations of dataset '{dataset_name}' to COCO format ...)")
+ coco_dict = convert_to_coco_dict(dataset_name)
+
+ logger.info(f"Caching COCO format annotations at '{output_file}' ...")
+ tmp_file = output_file + ".tmp"
+ with PathManager.open(tmp_file, "w") as f:
+ json.dump(coco_dict, f)
+ shutil.move(tmp_file, output_file)
+
+
+def register_coco_instances(name, metadata, json_file, image_root):
+ """
+ Register a dataset in COCO's json annotation format for
+ instance detection, instance segmentation and keypoint detection.
+ (i.e., Type 1 and 2 in http://cocodataset.org/#format-data.
+ `instances*.json` and `person_keypoints*.json` in the dataset).
+
+ This is an example of how to register a new dataset.
+ You can do something similar to this function, to register new datasets.
+
+ Args:
+ name (str): the name that identifies a dataset, e.g. "coco_2014_train".
+ metadata (dict): extra metadata associated with this dataset. You can
+ leave it as an empty dict.
+ json_file (str): path to the json instance annotation file.
+ image_root (str or path-like): directory which contains all the images.
+ """
+ assert isinstance(name, str), name
+ assert isinstance(json_file, (str, os.PathLike)), json_file
+ assert isinstance(image_root, (str, os.PathLike)), image_root
+ # 1. register a function which returns dicts
+ DatasetCatalog.register(name, lambda: load_coco_json(json_file, image_root, name))
+
+ # 2. Optionally, add metadata about this dataset,
+ # since they might be useful in evaluation, visualization or logging
+ MetadataCatalog.get(name).set(
+ json_file=json_file, image_root=image_root, evaluator_type="coco", **metadata
+ )
+
+
+if __name__ == "__main__":
+ """
+ Test the COCO json dataset loader.
+
+ Usage:
+ python -m detectron2.data.datasets.coco \
+ path/to/json path/to/image_root dataset_name
+
+ "dataset_name" can be "coco_2014_minival_100", or other
+ pre-registered ones
+ """
+ from annotator.oneformer.detectron2.utils.logger import setup_logger
+ from annotator.oneformer.detectron2.utils.visualizer import Visualizer
+ import annotator.oneformer.detectron2.data.datasets # noqa # add pre-defined metadata
+ import sys
+
+ logger = setup_logger(name=__name__)
+ assert sys.argv[3] in DatasetCatalog.list()
+ meta = MetadataCatalog.get(sys.argv[3])
+
+ dicts = load_coco_json(sys.argv[1], sys.argv[2], sys.argv[3])
+ logger.info("Done loading {} samples.".format(len(dicts)))
+
+ dirname = "coco-data-vis"
+ os.makedirs(dirname, exist_ok=True)
+ for d in dicts:
+ img = np.array(Image.open(d["file_name"]))
+ visualizer = Visualizer(img, metadata=meta)
+ vis = visualizer.draw_dataset_dict(d)
+ fpath = os.path.join(dirname, os.path.basename(d["file_name"]))
+ vis.save(fpath)
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/coco_panoptic.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/coco_panoptic.py
new file mode 100644
index 0000000000000000000000000000000000000000..a7180df512c29665222b1a90323ccfa7e7623137
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/coco_panoptic.py
@@ -0,0 +1,228 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import copy
+import json
+import os
+
+from annotator.oneformer.detectron2.data import DatasetCatalog, MetadataCatalog
+from annotator.oneformer.detectron2.utils.file_io import PathManager
+
+from .coco import load_coco_json, load_sem_seg
+
+__all__ = ["register_coco_panoptic", "register_coco_panoptic_separated"]
+
+
+def load_coco_panoptic_json(json_file, image_dir, gt_dir, meta):
+ """
+ Args:
+ image_dir (str): path to the raw dataset. e.g., "~/coco/train2017".
+ gt_dir (str): path to the raw annotations. e.g., "~/coco/panoptic_train2017".
+ json_file (str): path to the json file. e.g., "~/coco/annotations/panoptic_train2017.json".
+
+ Returns:
+ list[dict]: a list of dicts in Detectron2 standard format. (See
+ `Using Custom Datasets `_ )
+ """
+
+ def _convert_category_id(segment_info, meta):
+ if segment_info["category_id"] in meta["thing_dataset_id_to_contiguous_id"]:
+ segment_info["category_id"] = meta["thing_dataset_id_to_contiguous_id"][
+ segment_info["category_id"]
+ ]
+ segment_info["isthing"] = True
+ else:
+ segment_info["category_id"] = meta["stuff_dataset_id_to_contiguous_id"][
+ segment_info["category_id"]
+ ]
+ segment_info["isthing"] = False
+ return segment_info
+
+ with PathManager.open(json_file) as f:
+ json_info = json.load(f)
+
+ ret = []
+ for ann in json_info["annotations"]:
+ image_id = int(ann["image_id"])
+ # TODO: currently we assume image and label has the same filename but
+ # different extension, and images have extension ".jpg" for COCO. Need
+ # to make image extension a user-provided argument if we extend this
+ # function to support other COCO-like datasets.
+ image_file = os.path.join(image_dir, os.path.splitext(ann["file_name"])[0] + ".jpg")
+ label_file = os.path.join(gt_dir, ann["file_name"])
+ segments_info = [_convert_category_id(x, meta) for x in ann["segments_info"]]
+ ret.append(
+ {
+ "file_name": image_file,
+ "image_id": image_id,
+ "pan_seg_file_name": label_file,
+ "segments_info": segments_info,
+ }
+ )
+ assert len(ret), f"No images found in {image_dir}!"
+ assert PathManager.isfile(ret[0]["file_name"]), ret[0]["file_name"]
+ assert PathManager.isfile(ret[0]["pan_seg_file_name"]), ret[0]["pan_seg_file_name"]
+ return ret
+
+
+def register_coco_panoptic(
+ name, metadata, image_root, panoptic_root, panoptic_json, instances_json=None
+):
+ """
+ Register a "standard" version of COCO panoptic segmentation dataset named `name`.
+ The dictionaries in this registered dataset follows detectron2's standard format.
+ Hence it's called "standard".
+
+ Args:
+ name (str): the name that identifies a dataset,
+ e.g. "coco_2017_train_panoptic"
+ metadata (dict): extra metadata associated with this dataset.
+ image_root (str): directory which contains all the images
+ panoptic_root (str): directory which contains panoptic annotation images in COCO format
+ panoptic_json (str): path to the json panoptic annotation file in COCO format
+ sem_seg_root (none): not used, to be consistent with
+ `register_coco_panoptic_separated`.
+ instances_json (str): path to the json instance annotation file
+ """
+ panoptic_name = name
+ DatasetCatalog.register(
+ panoptic_name,
+ lambda: load_coco_panoptic_json(panoptic_json, image_root, panoptic_root, metadata),
+ )
+ MetadataCatalog.get(panoptic_name).set(
+ panoptic_root=panoptic_root,
+ image_root=image_root,
+ panoptic_json=panoptic_json,
+ json_file=instances_json,
+ evaluator_type="coco_panoptic_seg",
+ ignore_label=255,
+ label_divisor=1000,
+ **metadata,
+ )
+
+
+def register_coco_panoptic_separated(
+ name, metadata, image_root, panoptic_root, panoptic_json, sem_seg_root, instances_json
+):
+ """
+ Register a "separated" version of COCO panoptic segmentation dataset named `name`.
+ The annotations in this registered dataset will contain both instance annotations and
+ semantic annotations, each with its own contiguous ids. Hence it's called "separated".
+
+ It follows the setting used by the PanopticFPN paper:
+
+ 1. The instance annotations directly come from polygons in the COCO
+ instances annotation task, rather than from the masks in the COCO panoptic annotations.
+
+ The two format have small differences:
+ Polygons in the instance annotations may have overlaps.
+ The mask annotations are produced by labeling the overlapped polygons
+ with depth ordering.
+
+ 2. The semantic annotations are converted from panoptic annotations, where
+ all "things" are assigned a semantic id of 0.
+ All semantic categories will therefore have ids in contiguous
+ range [1, #stuff_categories].
+
+ This function will also register a pure semantic segmentation dataset
+ named ``name + '_stuffonly'``.
+
+ Args:
+ name (str): the name that identifies a dataset,
+ e.g. "coco_2017_train_panoptic"
+ metadata (dict): extra metadata associated with this dataset.
+ image_root (str): directory which contains all the images
+ panoptic_root (str): directory which contains panoptic annotation images
+ panoptic_json (str): path to the json panoptic annotation file
+ sem_seg_root (str): directory which contains all the ground truth segmentation annotations.
+ instances_json (str): path to the json instance annotation file
+ """
+ panoptic_name = name + "_separated"
+ DatasetCatalog.register(
+ panoptic_name,
+ lambda: merge_to_panoptic(
+ load_coco_json(instances_json, image_root, panoptic_name),
+ load_sem_seg(sem_seg_root, image_root),
+ ),
+ )
+ MetadataCatalog.get(panoptic_name).set(
+ panoptic_root=panoptic_root,
+ image_root=image_root,
+ panoptic_json=panoptic_json,
+ sem_seg_root=sem_seg_root,
+ json_file=instances_json, # TODO rename
+ evaluator_type="coco_panoptic_seg",
+ ignore_label=255,
+ **metadata,
+ )
+
+ semantic_name = name + "_stuffonly"
+ DatasetCatalog.register(semantic_name, lambda: load_sem_seg(sem_seg_root, image_root))
+ MetadataCatalog.get(semantic_name).set(
+ sem_seg_root=sem_seg_root,
+ image_root=image_root,
+ evaluator_type="sem_seg",
+ ignore_label=255,
+ **metadata,
+ )
+
+
+def merge_to_panoptic(detection_dicts, sem_seg_dicts):
+ """
+ Create dataset dicts for panoptic segmentation, by
+ merging two dicts using "file_name" field to match their entries.
+
+ Args:
+ detection_dicts (list[dict]): lists of dicts for object detection or instance segmentation.
+ sem_seg_dicts (list[dict]): lists of dicts for semantic segmentation.
+
+ Returns:
+ list[dict] (one per input image): Each dict contains all (key, value) pairs from dicts in
+ both detection_dicts and sem_seg_dicts that correspond to the same image.
+ The function assumes that the same key in different dicts has the same value.
+ """
+ results = []
+ sem_seg_file_to_entry = {x["file_name"]: x for x in sem_seg_dicts}
+ assert len(sem_seg_file_to_entry) > 0
+
+ for det_dict in detection_dicts:
+ dic = copy.copy(det_dict)
+ dic.update(sem_seg_file_to_entry[dic["file_name"]])
+ results.append(dic)
+ return results
+
+
+if __name__ == "__main__":
+ """
+ Test the COCO panoptic dataset loader.
+
+ Usage:
+ python -m detectron2.data.datasets.coco_panoptic \
+ path/to/image_root path/to/panoptic_root path/to/panoptic_json dataset_name 10
+
+ "dataset_name" can be "coco_2017_train_panoptic", or other
+ pre-registered ones
+ """
+ from annotator.oneformer.detectron2.utils.logger import setup_logger
+ from annotator.oneformer.detectron2.utils.visualizer import Visualizer
+ import annotator.oneformer.detectron2.data.datasets # noqa # add pre-defined metadata
+ import sys
+ from PIL import Image
+ import numpy as np
+
+ logger = setup_logger(name=__name__)
+ assert sys.argv[4] in DatasetCatalog.list()
+ meta = MetadataCatalog.get(sys.argv[4])
+
+ dicts = load_coco_panoptic_json(sys.argv[3], sys.argv[1], sys.argv[2], meta.as_dict())
+ logger.info("Done loading {} samples.".format(len(dicts)))
+
+ dirname = "coco-data-vis"
+ os.makedirs(dirname, exist_ok=True)
+ num_imgs_to_vis = int(sys.argv[5])
+ for i, d in enumerate(dicts):
+ img = np.array(Image.open(d["file_name"]))
+ visualizer = Visualizer(img, metadata=meta)
+ vis = visualizer.draw_dataset_dict(d)
+ fpath = os.path.join(dirname, os.path.basename(d["file_name"]))
+ vis.save(fpath)
+ if i + 1 >= num_imgs_to_vis:
+ break
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/lvis.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/lvis.py
new file mode 100644
index 0000000000000000000000000000000000000000..6e1e6ecc657e83d6df57da342b0655177402c514
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/lvis.py
@@ -0,0 +1,241 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import logging
+import os
+from fvcore.common.timer import Timer
+
+from annotator.oneformer.detectron2.data import DatasetCatalog, MetadataCatalog
+from annotator.oneformer.detectron2.structures import BoxMode
+from annotator.oneformer.detectron2.utils.file_io import PathManager
+
+from .builtin_meta import _get_coco_instances_meta
+from .lvis_v0_5_categories import LVIS_CATEGORIES as LVIS_V0_5_CATEGORIES
+from .lvis_v1_categories import LVIS_CATEGORIES as LVIS_V1_CATEGORIES
+from .lvis_v1_category_image_count import LVIS_CATEGORY_IMAGE_COUNT as LVIS_V1_CATEGORY_IMAGE_COUNT
+
+"""
+This file contains functions to parse LVIS-format annotations into dicts in the
+"Detectron2 format".
+"""
+
+logger = logging.getLogger(__name__)
+
+__all__ = ["load_lvis_json", "register_lvis_instances", "get_lvis_instances_meta"]
+
+
+def register_lvis_instances(name, metadata, json_file, image_root):
+ """
+ Register a dataset in LVIS's json annotation format for instance detection and segmentation.
+
+ Args:
+ name (str): a name that identifies the dataset, e.g. "lvis_v0.5_train".
+ metadata (dict): extra metadata associated with this dataset. It can be an empty dict.
+ json_file (str): path to the json instance annotation file.
+ image_root (str or path-like): directory which contains all the images.
+ """
+ DatasetCatalog.register(name, lambda: load_lvis_json(json_file, image_root, name))
+ MetadataCatalog.get(name).set(
+ json_file=json_file, image_root=image_root, evaluator_type="lvis", **metadata
+ )
+
+
+def load_lvis_json(json_file, image_root, dataset_name=None, extra_annotation_keys=None):
+ """
+ Load a json file in LVIS's annotation format.
+
+ Args:
+ json_file (str): full path to the LVIS json annotation file.
+ image_root (str): the directory where the images in this json file exists.
+ dataset_name (str): the name of the dataset (e.g., "lvis_v0.5_train").
+ If provided, this function will put "thing_classes" into the metadata
+ associated with this dataset.
+ extra_annotation_keys (list[str]): list of per-annotation keys that should also be
+ loaded into the dataset dict (besides "bbox", "bbox_mode", "category_id",
+ "segmentation"). The values for these keys will be returned as-is.
+
+ Returns:
+ list[dict]: a list of dicts in Detectron2 standard format. (See
+ `Using Custom Datasets `_ )
+
+ Notes:
+ 1. This function does not read the image files.
+ The results do not have the "image" field.
+ """
+ from lvis import LVIS
+
+ json_file = PathManager.get_local_path(json_file)
+
+ timer = Timer()
+ lvis_api = LVIS(json_file)
+ if timer.seconds() > 1:
+ logger.info("Loading {} takes {:.2f} seconds.".format(json_file, timer.seconds()))
+
+ if dataset_name is not None:
+ meta = get_lvis_instances_meta(dataset_name)
+ MetadataCatalog.get(dataset_name).set(**meta)
+
+ # sort indices for reproducible results
+ img_ids = sorted(lvis_api.imgs.keys())
+ # imgs is a list of dicts, each looks something like:
+ # {'license': 4,
+ # 'url': 'http://farm6.staticflickr.com/5454/9413846304_881d5e5c3b_z.jpg',
+ # 'file_name': 'COCO_val2014_000000001268.jpg',
+ # 'height': 427,
+ # 'width': 640,
+ # 'date_captured': '2013-11-17 05:57:24',
+ # 'id': 1268}
+ imgs = lvis_api.load_imgs(img_ids)
+ # anns is a list[list[dict]], where each dict is an annotation
+ # record for an object. The inner list enumerates the objects in an image
+ # and the outer list enumerates over images. Example of anns[0]:
+ # [{'segmentation': [[192.81,
+ # 247.09,
+ # ...
+ # 219.03,
+ # 249.06]],
+ # 'area': 1035.749,
+ # 'image_id': 1268,
+ # 'bbox': [192.81, 224.8, 74.73, 33.43],
+ # 'category_id': 16,
+ # 'id': 42986},
+ # ...]
+ anns = [lvis_api.img_ann_map[img_id] for img_id in img_ids]
+
+ # Sanity check that each annotation has a unique id
+ ann_ids = [ann["id"] for anns_per_image in anns for ann in anns_per_image]
+ assert len(set(ann_ids)) == len(ann_ids), "Annotation ids in '{}' are not unique".format(
+ json_file
+ )
+
+ imgs_anns = list(zip(imgs, anns))
+
+ logger.info("Loaded {} images in the LVIS format from {}".format(len(imgs_anns), json_file))
+
+ if extra_annotation_keys:
+ logger.info(
+ "The following extra annotation keys will be loaded: {} ".format(extra_annotation_keys)
+ )
+ else:
+ extra_annotation_keys = []
+
+ def get_file_name(img_root, img_dict):
+ # Determine the path including the split folder ("train2017", "val2017", "test2017") from
+ # the coco_url field. Example:
+ # 'coco_url': 'http://images.cocodataset.org/train2017/000000155379.jpg'
+ split_folder, file_name = img_dict["coco_url"].split("/")[-2:]
+ return os.path.join(img_root + split_folder, file_name)
+
+ dataset_dicts = []
+
+ for (img_dict, anno_dict_list) in imgs_anns:
+ record = {}
+ record["file_name"] = get_file_name(image_root, img_dict)
+ record["height"] = img_dict["height"]
+ record["width"] = img_dict["width"]
+ record["not_exhaustive_category_ids"] = img_dict.get("not_exhaustive_category_ids", [])
+ record["neg_category_ids"] = img_dict.get("neg_category_ids", [])
+ image_id = record["image_id"] = img_dict["id"]
+
+ objs = []
+ for anno in anno_dict_list:
+ # Check that the image_id in this annotation is the same as
+ # the image_id we're looking at.
+ # This fails only when the data parsing logic or the annotation file is buggy.
+ assert anno["image_id"] == image_id
+ obj = {"bbox": anno["bbox"], "bbox_mode": BoxMode.XYWH_ABS}
+ # LVIS data loader can be used to load COCO dataset categories. In this case `meta`
+ # variable will have a field with COCO-specific category mapping.
+ if dataset_name is not None and "thing_dataset_id_to_contiguous_id" in meta:
+ obj["category_id"] = meta["thing_dataset_id_to_contiguous_id"][anno["category_id"]]
+ else:
+ obj["category_id"] = anno["category_id"] - 1 # Convert 1-indexed to 0-indexed
+ segm = anno["segmentation"] # list[list[float]]
+ # filter out invalid polygons (< 3 points)
+ valid_segm = [poly for poly in segm if len(poly) % 2 == 0 and len(poly) >= 6]
+ assert len(segm) == len(
+ valid_segm
+ ), "Annotation contains an invalid polygon with < 3 points"
+ assert len(segm) > 0
+ obj["segmentation"] = segm
+ for extra_ann_key in extra_annotation_keys:
+ obj[extra_ann_key] = anno[extra_ann_key]
+ objs.append(obj)
+ record["annotations"] = objs
+ dataset_dicts.append(record)
+
+ return dataset_dicts
+
+
+def get_lvis_instances_meta(dataset_name):
+ """
+ Load LVIS metadata.
+
+ Args:
+ dataset_name (str): LVIS dataset name without the split name (e.g., "lvis_v0.5").
+
+ Returns:
+ dict: LVIS metadata with keys: thing_classes
+ """
+ if "cocofied" in dataset_name:
+ return _get_coco_instances_meta()
+ if "v0.5" in dataset_name:
+ return _get_lvis_instances_meta_v0_5()
+ elif "v1" in dataset_name:
+ return _get_lvis_instances_meta_v1()
+ raise ValueError("No built-in metadata for dataset {}".format(dataset_name))
+
+
+def _get_lvis_instances_meta_v0_5():
+ assert len(LVIS_V0_5_CATEGORIES) == 1230
+ cat_ids = [k["id"] for k in LVIS_V0_5_CATEGORIES]
+ assert min(cat_ids) == 1 and max(cat_ids) == len(
+ cat_ids
+ ), "Category ids are not in [1, #categories], as expected"
+ # Ensure that the category list is sorted by id
+ lvis_categories = sorted(LVIS_V0_5_CATEGORIES, key=lambda x: x["id"])
+ thing_classes = [k["synonyms"][0] for k in lvis_categories]
+ meta = {"thing_classes": thing_classes}
+ return meta
+
+
+def _get_lvis_instances_meta_v1():
+ assert len(LVIS_V1_CATEGORIES) == 1203
+ cat_ids = [k["id"] for k in LVIS_V1_CATEGORIES]
+ assert min(cat_ids) == 1 and max(cat_ids) == len(
+ cat_ids
+ ), "Category ids are not in [1, #categories], as expected"
+ # Ensure that the category list is sorted by id
+ lvis_categories = sorted(LVIS_V1_CATEGORIES, key=lambda x: x["id"])
+ thing_classes = [k["synonyms"][0] for k in lvis_categories]
+ meta = {"thing_classes": thing_classes, "class_image_count": LVIS_V1_CATEGORY_IMAGE_COUNT}
+ return meta
+
+
+if __name__ == "__main__":
+ """
+ Test the LVIS json dataset loader.
+
+ Usage:
+ python -m detectron2.data.datasets.lvis \
+ path/to/json path/to/image_root dataset_name vis_limit
+ """
+ import sys
+ import numpy as np
+ from annotator.oneformer.detectron2.utils.logger import setup_logger
+ from PIL import Image
+ import annotator.oneformer.detectron2.data.datasets # noqa # add pre-defined metadata
+ from annotator.oneformer.detectron2.utils.visualizer import Visualizer
+
+ logger = setup_logger(name=__name__)
+ meta = MetadataCatalog.get(sys.argv[3])
+
+ dicts = load_lvis_json(sys.argv[1], sys.argv[2], sys.argv[3])
+ logger.info("Done loading {} samples.".format(len(dicts)))
+
+ dirname = "lvis-data-vis"
+ os.makedirs(dirname, exist_ok=True)
+ for d in dicts[: int(sys.argv[4])]:
+ img = np.array(Image.open(d["file_name"]))
+ visualizer = Visualizer(img, metadata=meta)
+ vis = visualizer.draw_dataset_dict(d)
+ fpath = os.path.join(dirname, os.path.basename(d["file_name"]))
+ vis.save(fpath)
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/lvis_v0_5_categories.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/lvis_v0_5_categories.py
new file mode 100644
index 0000000000000000000000000000000000000000..d3dab6198da614937b08682f4c9edf52bdf1d236
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/lvis_v0_5_categories.py
@@ -0,0 +1,13 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+# Autogen with
+# with open("lvis_v0.5_val.json", "r") as f:
+# a = json.load(f)
+# c = a["categories"]
+# for x in c:
+# del x["image_count"]
+# del x["instance_count"]
+# LVIS_CATEGORIES = repr(c) + " # noqa"
+
+# fmt: off
+LVIS_CATEGORIES = [{'frequency': 'r', 'id': 1, 'synset': 'acorn.n.01', 'synonyms': ['acorn'], 'def': 'nut from an oak tree', 'name': 'acorn'}, {'frequency': 'c', 'id': 2, 'synset': 'aerosol.n.02', 'synonyms': ['aerosol_can', 'spray_can'], 'def': 'a dispenser that holds a substance under pressure', 'name': 'aerosol_can'}, {'frequency': 'f', 'id': 3, 'synset': 'air_conditioner.n.01', 'synonyms': ['air_conditioner'], 'def': 'a machine that keeps air cool and dry', 'name': 'air_conditioner'}, {'frequency': 'f', 'id': 4, 'synset': 'airplane.n.01', 'synonyms': ['airplane', 'aeroplane'], 'def': 'an aircraft that has a fixed wing and is powered by propellers or jets', 'name': 'airplane'}, {'frequency': 'c', 'id': 5, 'synset': 'alarm_clock.n.01', 'synonyms': ['alarm_clock'], 'def': 'a clock that wakes a sleeper at some preset time', 'name': 'alarm_clock'}, {'frequency': 'c', 'id': 6, 'synset': 'alcohol.n.01', 'synonyms': ['alcohol', 'alcoholic_beverage'], 'def': 'a liquor or brew containing alcohol as the active agent', 'name': 'alcohol'}, {'frequency': 'r', 'id': 7, 'synset': 'alligator.n.02', 'synonyms': ['alligator', 'gator'], 'def': 'amphibious reptiles related to crocodiles but with shorter broader snouts', 'name': 'alligator'}, {'frequency': 'c', 'id': 8, 'synset': 'almond.n.02', 'synonyms': ['almond'], 'def': 'oval-shaped edible seed of the almond tree', 'name': 'almond'}, {'frequency': 'c', 'id': 9, 'synset': 'ambulance.n.01', 'synonyms': ['ambulance'], 'def': 'a vehicle that takes people to and from hospitals', 'name': 'ambulance'}, {'frequency': 'r', 'id': 10, 'synset': 'amplifier.n.01', 'synonyms': ['amplifier'], 'def': 'electronic equipment that increases strength of signals', 'name': 'amplifier'}, {'frequency': 'c', 'id': 11, 'synset': 'anklet.n.03', 'synonyms': ['anklet', 'ankle_bracelet'], 'def': 'an ornament worn around the ankle', 'name': 'anklet'}, {'frequency': 'f', 'id': 12, 'synset': 'antenna.n.01', 'synonyms': ['antenna', 'aerial', 'transmitting_aerial'], 'def': 'an electrical device that sends or receives radio or television signals', 'name': 'antenna'}, {'frequency': 'f', 'id': 13, 'synset': 'apple.n.01', 'synonyms': ['apple'], 'def': 'fruit with red or yellow or green skin and sweet to tart crisp whitish flesh', 'name': 'apple'}, {'frequency': 'r', 'id': 14, 'synset': 'apple_juice.n.01', 'synonyms': ['apple_juice'], 'def': 'the juice of apples', 'name': 'apple_juice'}, {'frequency': 'r', 'id': 15, 'synset': 'applesauce.n.01', 'synonyms': ['applesauce'], 'def': 'puree of stewed apples usually sweetened and spiced', 'name': 'applesauce'}, {'frequency': 'r', 'id': 16, 'synset': 'apricot.n.02', 'synonyms': ['apricot'], 'def': 'downy yellow to rosy-colored fruit resembling a small peach', 'name': 'apricot'}, {'frequency': 'f', 'id': 17, 'synset': 'apron.n.01', 'synonyms': ['apron'], 'def': 'a garment of cloth that is tied about the waist and worn to protect clothing', 'name': 'apron'}, {'frequency': 'c', 'id': 18, 'synset': 'aquarium.n.01', 'synonyms': ['aquarium', 'fish_tank'], 'def': 'a tank/pool/bowl filled with water for keeping live fish and underwater animals', 'name': 'aquarium'}, {'frequency': 'c', 'id': 19, 'synset': 'armband.n.02', 'synonyms': ['armband'], 'def': 'a band worn around the upper arm', 'name': 'armband'}, {'frequency': 'f', 'id': 20, 'synset': 'armchair.n.01', 'synonyms': ['armchair'], 'def': 'chair with a support on each side for arms', 'name': 'armchair'}, {'frequency': 'r', 'id': 21, 'synset': 'armoire.n.01', 'synonyms': ['armoire'], 'def': 'a large wardrobe or cabinet', 'name': 'armoire'}, {'frequency': 'r', 'id': 22, 'synset': 'armor.n.01', 'synonyms': ['armor', 'armour'], 'def': 'protective covering made of metal and used in combat', 'name': 'armor'}, {'frequency': 'c', 'id': 23, 'synset': 'artichoke.n.02', 'synonyms': ['artichoke'], 'def': 'a thistlelike flower head with edible fleshy leaves and heart', 'name': 'artichoke'}, {'frequency': 'f', 'id': 24, 'synset': 'ashcan.n.01', 'synonyms': ['trash_can', 'garbage_can', 'wastebin', 'dustbin', 'trash_barrel', 'trash_bin'], 'def': 'a bin that holds rubbish until it is collected', 'name': 'trash_can'}, {'frequency': 'c', 'id': 25, 'synset': 'ashtray.n.01', 'synonyms': ['ashtray'], 'def': "a receptacle for the ash from smokers' cigars or cigarettes", 'name': 'ashtray'}, {'frequency': 'c', 'id': 26, 'synset': 'asparagus.n.02', 'synonyms': ['asparagus'], 'def': 'edible young shoots of the asparagus plant', 'name': 'asparagus'}, {'frequency': 'c', 'id': 27, 'synset': 'atomizer.n.01', 'synonyms': ['atomizer', 'atomiser', 'spray', 'sprayer', 'nebulizer', 'nebuliser'], 'def': 'a dispenser that turns a liquid (such as perfume) into a fine mist', 'name': 'atomizer'}, {'frequency': 'c', 'id': 28, 'synset': 'avocado.n.01', 'synonyms': ['avocado'], 'def': 'a pear-shaped fruit with green or blackish skin and rich yellowish pulp enclosing a single large seed', 'name': 'avocado'}, {'frequency': 'c', 'id': 29, 'synset': 'award.n.02', 'synonyms': ['award', 'accolade'], 'def': 'a tangible symbol signifying approval or distinction', 'name': 'award'}, {'frequency': 'f', 'id': 30, 'synset': 'awning.n.01', 'synonyms': ['awning'], 'def': 'a canopy made of canvas to shelter people or things from rain or sun', 'name': 'awning'}, {'frequency': 'r', 'id': 31, 'synset': 'ax.n.01', 'synonyms': ['ax', 'axe'], 'def': 'an edge tool with a heavy bladed head mounted across a handle', 'name': 'ax'}, {'frequency': 'f', 'id': 32, 'synset': 'baby_buggy.n.01', 'synonyms': ['baby_buggy', 'baby_carriage', 'perambulator', 'pram', 'stroller'], 'def': 'a small vehicle with four wheels in which a baby or child is pushed around', 'name': 'baby_buggy'}, {'frequency': 'c', 'id': 33, 'synset': 'backboard.n.01', 'synonyms': ['basketball_backboard'], 'def': 'a raised vertical board with basket attached; used to play basketball', 'name': 'basketball_backboard'}, {'frequency': 'f', 'id': 34, 'synset': 'backpack.n.01', 'synonyms': ['backpack', 'knapsack', 'packsack', 'rucksack', 'haversack'], 'def': 'a bag carried by a strap on your back or shoulder', 'name': 'backpack'}, {'frequency': 'f', 'id': 35, 'synset': 'bag.n.04', 'synonyms': ['handbag', 'purse', 'pocketbook'], 'def': 'a container used for carrying money and small personal items or accessories', 'name': 'handbag'}, {'frequency': 'f', 'id': 36, 'synset': 'bag.n.06', 'synonyms': ['suitcase', 'baggage', 'luggage'], 'def': 'cases used to carry belongings when traveling', 'name': 'suitcase'}, {'frequency': 'c', 'id': 37, 'synset': 'bagel.n.01', 'synonyms': ['bagel', 'beigel'], 'def': 'glazed yeast-raised doughnut-shaped roll with hard crust', 'name': 'bagel'}, {'frequency': 'r', 'id': 38, 'synset': 'bagpipe.n.01', 'synonyms': ['bagpipe'], 'def': 'a tubular wind instrument; the player blows air into a bag and squeezes it out', 'name': 'bagpipe'}, {'frequency': 'r', 'id': 39, 'synset': 'baguet.n.01', 'synonyms': ['baguet', 'baguette'], 'def': 'narrow French stick loaf', 'name': 'baguet'}, {'frequency': 'r', 'id': 40, 'synset': 'bait.n.02', 'synonyms': ['bait', 'lure'], 'def': 'something used to lure fish or other animals into danger so they can be trapped or killed', 'name': 'bait'}, {'frequency': 'f', 'id': 41, 'synset': 'ball.n.06', 'synonyms': ['ball'], 'def': 'a spherical object used as a plaything', 'name': 'ball'}, {'frequency': 'r', 'id': 42, 'synset': 'ballet_skirt.n.01', 'synonyms': ['ballet_skirt', 'tutu'], 'def': 'very short skirt worn by ballerinas', 'name': 'ballet_skirt'}, {'frequency': 'f', 'id': 43, 'synset': 'balloon.n.01', 'synonyms': ['balloon'], 'def': 'large tough nonrigid bag filled with gas or heated air', 'name': 'balloon'}, {'frequency': 'c', 'id': 44, 'synset': 'bamboo.n.02', 'synonyms': ['bamboo'], 'def': 'woody tropical grass having hollow woody stems', 'name': 'bamboo'}, {'frequency': 'f', 'id': 45, 'synset': 'banana.n.02', 'synonyms': ['banana'], 'def': 'elongated crescent-shaped yellow fruit with soft sweet flesh', 'name': 'banana'}, {'frequency': 'r', 'id': 46, 'synset': 'band_aid.n.01', 'synonyms': ['Band_Aid'], 'def': 'trade name for an adhesive bandage to cover small cuts or blisters', 'name': 'Band_Aid'}, {'frequency': 'c', 'id': 47, 'synset': 'bandage.n.01', 'synonyms': ['bandage'], 'def': 'a piece of soft material that covers and protects an injured part of the body', 'name': 'bandage'}, {'frequency': 'c', 'id': 48, 'synset': 'bandanna.n.01', 'synonyms': ['bandanna', 'bandana'], 'def': 'large and brightly colored handkerchief; often used as a neckerchief', 'name': 'bandanna'}, {'frequency': 'r', 'id': 49, 'synset': 'banjo.n.01', 'synonyms': ['banjo'], 'def': 'a stringed instrument of the guitar family with a long neck and circular body', 'name': 'banjo'}, {'frequency': 'f', 'id': 50, 'synset': 'banner.n.01', 'synonyms': ['banner', 'streamer'], 'def': 'long strip of cloth or paper used for decoration or advertising', 'name': 'banner'}, {'frequency': 'r', 'id': 51, 'synset': 'barbell.n.01', 'synonyms': ['barbell'], 'def': 'a bar to which heavy discs are attached at each end; used in weightlifting', 'name': 'barbell'}, {'frequency': 'r', 'id': 52, 'synset': 'barge.n.01', 'synonyms': ['barge'], 'def': 'a flatbottom boat for carrying heavy loads (especially on canals)', 'name': 'barge'}, {'frequency': 'f', 'id': 53, 'synset': 'barrel.n.02', 'synonyms': ['barrel', 'cask'], 'def': 'a cylindrical container that holds liquids', 'name': 'barrel'}, {'frequency': 'c', 'id': 54, 'synset': 'barrette.n.01', 'synonyms': ['barrette'], 'def': "a pin for holding women's hair in place", 'name': 'barrette'}, {'frequency': 'c', 'id': 55, 'synset': 'barrow.n.03', 'synonyms': ['barrow', 'garden_cart', 'lawn_cart', 'wheelbarrow'], 'def': 'a cart for carrying small loads; has handles and one or more wheels', 'name': 'barrow'}, {'frequency': 'f', 'id': 56, 'synset': 'base.n.03', 'synonyms': ['baseball_base'], 'def': 'a place that the runner must touch before scoring', 'name': 'baseball_base'}, {'frequency': 'f', 'id': 57, 'synset': 'baseball.n.02', 'synonyms': ['baseball'], 'def': 'a ball used in playing baseball', 'name': 'baseball'}, {'frequency': 'f', 'id': 58, 'synset': 'baseball_bat.n.01', 'synonyms': ['baseball_bat'], 'def': 'an implement used in baseball by the batter', 'name': 'baseball_bat'}, {'frequency': 'f', 'id': 59, 'synset': 'baseball_cap.n.01', 'synonyms': ['baseball_cap', 'jockey_cap', 'golf_cap'], 'def': 'a cap with a bill', 'name': 'baseball_cap'}, {'frequency': 'f', 'id': 60, 'synset': 'baseball_glove.n.01', 'synonyms': ['baseball_glove', 'baseball_mitt'], 'def': 'the handwear used by fielders in playing baseball', 'name': 'baseball_glove'}, {'frequency': 'f', 'id': 61, 'synset': 'basket.n.01', 'synonyms': ['basket', 'handbasket'], 'def': 'a container that is usually woven and has handles', 'name': 'basket'}, {'frequency': 'c', 'id': 62, 'synset': 'basket.n.03', 'synonyms': ['basketball_hoop'], 'def': 'metal hoop supporting a net through which players try to throw the basketball', 'name': 'basketball_hoop'}, {'frequency': 'c', 'id': 63, 'synset': 'basketball.n.02', 'synonyms': ['basketball'], 'def': 'an inflated ball used in playing basketball', 'name': 'basketball'}, {'frequency': 'r', 'id': 64, 'synset': 'bass_horn.n.01', 'synonyms': ['bass_horn', 'sousaphone', 'tuba'], 'def': 'the lowest brass wind instrument', 'name': 'bass_horn'}, {'frequency': 'r', 'id': 65, 'synset': 'bat.n.01', 'synonyms': ['bat_(animal)'], 'def': 'nocturnal mouselike mammal with forelimbs modified to form membranous wings', 'name': 'bat_(animal)'}, {'frequency': 'f', 'id': 66, 'synset': 'bath_mat.n.01', 'synonyms': ['bath_mat'], 'def': 'a heavy towel or mat to stand on while drying yourself after a bath', 'name': 'bath_mat'}, {'frequency': 'f', 'id': 67, 'synset': 'bath_towel.n.01', 'synonyms': ['bath_towel'], 'def': 'a large towel; to dry yourself after a bath', 'name': 'bath_towel'}, {'frequency': 'c', 'id': 68, 'synset': 'bathrobe.n.01', 'synonyms': ['bathrobe'], 'def': 'a loose-fitting robe of towelling; worn after a bath or swim', 'name': 'bathrobe'}, {'frequency': 'f', 'id': 69, 'synset': 'bathtub.n.01', 'synonyms': ['bathtub', 'bathing_tub'], 'def': 'a large open container that you fill with water and use to wash the body', 'name': 'bathtub'}, {'frequency': 'r', 'id': 70, 'synset': 'batter.n.02', 'synonyms': ['batter_(food)'], 'def': 'a liquid or semiliquid mixture, as of flour, eggs, and milk, used in cooking', 'name': 'batter_(food)'}, {'frequency': 'c', 'id': 71, 'synset': 'battery.n.02', 'synonyms': ['battery'], 'def': 'a portable device that produces electricity', 'name': 'battery'}, {'frequency': 'r', 'id': 72, 'synset': 'beach_ball.n.01', 'synonyms': ['beachball'], 'def': 'large and light ball; for play at the seaside', 'name': 'beachball'}, {'frequency': 'c', 'id': 73, 'synset': 'bead.n.01', 'synonyms': ['bead'], 'def': 'a small ball with a hole through the middle used for ornamentation, jewellery, etc.', 'name': 'bead'}, {'frequency': 'r', 'id': 74, 'synset': 'beaker.n.01', 'synonyms': ['beaker'], 'def': 'a flatbottomed jar made of glass or plastic; used for chemistry', 'name': 'beaker'}, {'frequency': 'c', 'id': 75, 'synset': 'bean_curd.n.01', 'synonyms': ['bean_curd', 'tofu'], 'def': 'cheeselike food made of curdled soybean milk', 'name': 'bean_curd'}, {'frequency': 'c', 'id': 76, 'synset': 'beanbag.n.01', 'synonyms': ['beanbag'], 'def': 'a bag filled with dried beans or similar items; used in games or to sit on', 'name': 'beanbag'}, {'frequency': 'f', 'id': 77, 'synset': 'beanie.n.01', 'synonyms': ['beanie', 'beany'], 'def': 'a small skullcap; formerly worn by schoolboys and college freshmen', 'name': 'beanie'}, {'frequency': 'f', 'id': 78, 'synset': 'bear.n.01', 'synonyms': ['bear'], 'def': 'large carnivorous or omnivorous mammals with shaggy coats and claws', 'name': 'bear'}, {'frequency': 'f', 'id': 79, 'synset': 'bed.n.01', 'synonyms': ['bed'], 'def': 'a piece of furniture that provides a place to sleep', 'name': 'bed'}, {'frequency': 'c', 'id': 80, 'synset': 'bedspread.n.01', 'synonyms': ['bedspread', 'bedcover', 'bed_covering', 'counterpane', 'spread'], 'def': 'decorative cover for a bed', 'name': 'bedspread'}, {'frequency': 'f', 'id': 81, 'synset': 'beef.n.01', 'synonyms': ['cow'], 'def': 'cattle that are reared for their meat', 'name': 'cow'}, {'frequency': 'c', 'id': 82, 'synset': 'beef.n.02', 'synonyms': ['beef_(food)', 'boeuf_(food)'], 'def': 'meat from an adult domestic bovine', 'name': 'beef_(food)'}, {'frequency': 'r', 'id': 83, 'synset': 'beeper.n.01', 'synonyms': ['beeper', 'pager'], 'def': 'an device that beeps when the person carrying it is being paged', 'name': 'beeper'}, {'frequency': 'f', 'id': 84, 'synset': 'beer_bottle.n.01', 'synonyms': ['beer_bottle'], 'def': 'a bottle that holds beer', 'name': 'beer_bottle'}, {'frequency': 'c', 'id': 85, 'synset': 'beer_can.n.01', 'synonyms': ['beer_can'], 'def': 'a can that holds beer', 'name': 'beer_can'}, {'frequency': 'r', 'id': 86, 'synset': 'beetle.n.01', 'synonyms': ['beetle'], 'def': 'insect with hard wing covers', 'name': 'beetle'}, {'frequency': 'f', 'id': 87, 'synset': 'bell.n.01', 'synonyms': ['bell'], 'def': 'a hollow device made of metal that makes a ringing sound when struck', 'name': 'bell'}, {'frequency': 'f', 'id': 88, 'synset': 'bell_pepper.n.02', 'synonyms': ['bell_pepper', 'capsicum'], 'def': 'large bell-shaped sweet pepper in green or red or yellow or orange or black varieties', 'name': 'bell_pepper'}, {'frequency': 'f', 'id': 89, 'synset': 'belt.n.02', 'synonyms': ['belt'], 'def': 'a band to tie or buckle around the body (usually at the waist)', 'name': 'belt'}, {'frequency': 'f', 'id': 90, 'synset': 'belt_buckle.n.01', 'synonyms': ['belt_buckle'], 'def': 'the buckle used to fasten a belt', 'name': 'belt_buckle'}, {'frequency': 'f', 'id': 91, 'synset': 'bench.n.01', 'synonyms': ['bench'], 'def': 'a long seat for more than one person', 'name': 'bench'}, {'frequency': 'c', 'id': 92, 'synset': 'beret.n.01', 'synonyms': ['beret'], 'def': 'a cap with no brim or bill; made of soft cloth', 'name': 'beret'}, {'frequency': 'c', 'id': 93, 'synset': 'bib.n.02', 'synonyms': ['bib'], 'def': 'a napkin tied under the chin of a child while eating', 'name': 'bib'}, {'frequency': 'r', 'id': 94, 'synset': 'bible.n.01', 'synonyms': ['Bible'], 'def': 'the sacred writings of the Christian religions', 'name': 'Bible'}, {'frequency': 'f', 'id': 95, 'synset': 'bicycle.n.01', 'synonyms': ['bicycle', 'bike_(bicycle)'], 'def': 'a wheeled vehicle that has two wheels and is moved by foot pedals', 'name': 'bicycle'}, {'frequency': 'f', 'id': 96, 'synset': 'bill.n.09', 'synonyms': ['visor', 'vizor'], 'def': 'a brim that projects to the front to shade the eyes', 'name': 'visor'}, {'frequency': 'c', 'id': 97, 'synset': 'binder.n.03', 'synonyms': ['binder', 'ring-binder'], 'def': 'holds loose papers or magazines', 'name': 'binder'}, {'frequency': 'c', 'id': 98, 'synset': 'binoculars.n.01', 'synonyms': ['binoculars', 'field_glasses', 'opera_glasses'], 'def': 'an optical instrument designed for simultaneous use by both eyes', 'name': 'binoculars'}, {'frequency': 'f', 'id': 99, 'synset': 'bird.n.01', 'synonyms': ['bird'], 'def': 'animal characterized by feathers and wings', 'name': 'bird'}, {'frequency': 'r', 'id': 100, 'synset': 'bird_feeder.n.01', 'synonyms': ['birdfeeder'], 'def': 'an outdoor device that supplies food for wild birds', 'name': 'birdfeeder'}, {'frequency': 'r', 'id': 101, 'synset': 'birdbath.n.01', 'synonyms': ['birdbath'], 'def': 'an ornamental basin (usually in a garden) for birds to bathe in', 'name': 'birdbath'}, {'frequency': 'c', 'id': 102, 'synset': 'birdcage.n.01', 'synonyms': ['birdcage'], 'def': 'a cage in which a bird can be kept', 'name': 'birdcage'}, {'frequency': 'c', 'id': 103, 'synset': 'birdhouse.n.01', 'synonyms': ['birdhouse'], 'def': 'a shelter for birds', 'name': 'birdhouse'}, {'frequency': 'f', 'id': 104, 'synset': 'birthday_cake.n.01', 'synonyms': ['birthday_cake'], 'def': 'decorated cake served at a birthday party', 'name': 'birthday_cake'}, {'frequency': 'r', 'id': 105, 'synset': 'birthday_card.n.01', 'synonyms': ['birthday_card'], 'def': 'a card expressing a birthday greeting', 'name': 'birthday_card'}, {'frequency': 'r', 'id': 106, 'synset': 'biscuit.n.01', 'synonyms': ['biscuit_(bread)'], 'def': 'small round bread leavened with baking-powder or soda', 'name': 'biscuit_(bread)'}, {'frequency': 'r', 'id': 107, 'synset': 'black_flag.n.01', 'synonyms': ['pirate_flag'], 'def': 'a flag usually bearing a white skull and crossbones on a black background', 'name': 'pirate_flag'}, {'frequency': 'c', 'id': 108, 'synset': 'black_sheep.n.02', 'synonyms': ['black_sheep'], 'def': 'sheep with a black coat', 'name': 'black_sheep'}, {'frequency': 'c', 'id': 109, 'synset': 'blackboard.n.01', 'synonyms': ['blackboard', 'chalkboard'], 'def': 'sheet of slate; for writing with chalk', 'name': 'blackboard'}, {'frequency': 'f', 'id': 110, 'synset': 'blanket.n.01', 'synonyms': ['blanket'], 'def': 'bedding that keeps a person warm in bed', 'name': 'blanket'}, {'frequency': 'c', 'id': 111, 'synset': 'blazer.n.01', 'synonyms': ['blazer', 'sport_jacket', 'sport_coat', 'sports_jacket', 'sports_coat'], 'def': 'lightweight jacket; often striped in the colors of a club or school', 'name': 'blazer'}, {'frequency': 'f', 'id': 112, 'synset': 'blender.n.01', 'synonyms': ['blender', 'liquidizer', 'liquidiser'], 'def': 'an electrically powered mixer that mix or chop or liquefy foods', 'name': 'blender'}, {'frequency': 'r', 'id': 113, 'synset': 'blimp.n.02', 'synonyms': ['blimp'], 'def': 'a small nonrigid airship used for observation or as a barrage balloon', 'name': 'blimp'}, {'frequency': 'c', 'id': 114, 'synset': 'blinker.n.01', 'synonyms': ['blinker', 'flasher'], 'def': 'a light that flashes on and off; used as a signal or to send messages', 'name': 'blinker'}, {'frequency': 'c', 'id': 115, 'synset': 'blueberry.n.02', 'synonyms': ['blueberry'], 'def': 'sweet edible dark-blue berries of blueberry plants', 'name': 'blueberry'}, {'frequency': 'r', 'id': 116, 'synset': 'boar.n.02', 'synonyms': ['boar'], 'def': 'an uncastrated male hog', 'name': 'boar'}, {'frequency': 'r', 'id': 117, 'synset': 'board.n.09', 'synonyms': ['gameboard'], 'def': 'a flat portable surface (usually rectangular) designed for board games', 'name': 'gameboard'}, {'frequency': 'f', 'id': 118, 'synset': 'boat.n.01', 'synonyms': ['boat', 'ship_(boat)'], 'def': 'a vessel for travel on water', 'name': 'boat'}, {'frequency': 'c', 'id': 119, 'synset': 'bobbin.n.01', 'synonyms': ['bobbin', 'spool', 'reel'], 'def': 'a thing around which thread/tape/film or other flexible materials can be wound', 'name': 'bobbin'}, {'frequency': 'r', 'id': 120, 'synset': 'bobby_pin.n.01', 'synonyms': ['bobby_pin', 'hairgrip'], 'def': 'a flat wire hairpin used to hold bobbed hair in place', 'name': 'bobby_pin'}, {'frequency': 'c', 'id': 121, 'synset': 'boiled_egg.n.01', 'synonyms': ['boiled_egg', 'coddled_egg'], 'def': 'egg cooked briefly in the shell in gently boiling water', 'name': 'boiled_egg'}, {'frequency': 'r', 'id': 122, 'synset': 'bolo_tie.n.01', 'synonyms': ['bolo_tie', 'bolo', 'bola_tie', 'bola'], 'def': 'a cord fastened around the neck with an ornamental clasp and worn as a necktie', 'name': 'bolo_tie'}, {'frequency': 'c', 'id': 123, 'synset': 'bolt.n.03', 'synonyms': ['deadbolt'], 'def': 'the part of a lock that is engaged or withdrawn with a key', 'name': 'deadbolt'}, {'frequency': 'f', 'id': 124, 'synset': 'bolt.n.06', 'synonyms': ['bolt'], 'def': 'a screw that screws into a nut to form a fastener', 'name': 'bolt'}, {'frequency': 'r', 'id': 125, 'synset': 'bonnet.n.01', 'synonyms': ['bonnet'], 'def': 'a hat tied under the chin', 'name': 'bonnet'}, {'frequency': 'f', 'id': 126, 'synset': 'book.n.01', 'synonyms': ['book'], 'def': 'a written work or composition that has been published', 'name': 'book'}, {'frequency': 'r', 'id': 127, 'synset': 'book_bag.n.01', 'synonyms': ['book_bag'], 'def': 'a bag in which students carry their books', 'name': 'book_bag'}, {'frequency': 'c', 'id': 128, 'synset': 'bookcase.n.01', 'synonyms': ['bookcase'], 'def': 'a piece of furniture with shelves for storing books', 'name': 'bookcase'}, {'frequency': 'c', 'id': 129, 'synset': 'booklet.n.01', 'synonyms': ['booklet', 'brochure', 'leaflet', 'pamphlet'], 'def': 'a small book usually having a paper cover', 'name': 'booklet'}, {'frequency': 'r', 'id': 130, 'synset': 'bookmark.n.01', 'synonyms': ['bookmark', 'bookmarker'], 'def': 'a marker (a piece of paper or ribbon) placed between the pages of a book', 'name': 'bookmark'}, {'frequency': 'r', 'id': 131, 'synset': 'boom.n.04', 'synonyms': ['boom_microphone', 'microphone_boom'], 'def': 'a pole carrying an overhead microphone projected over a film or tv set', 'name': 'boom_microphone'}, {'frequency': 'f', 'id': 132, 'synset': 'boot.n.01', 'synonyms': ['boot'], 'def': 'footwear that covers the whole foot and lower leg', 'name': 'boot'}, {'frequency': 'f', 'id': 133, 'synset': 'bottle.n.01', 'synonyms': ['bottle'], 'def': 'a glass or plastic vessel used for storing drinks or other liquids', 'name': 'bottle'}, {'frequency': 'c', 'id': 134, 'synset': 'bottle_opener.n.01', 'synonyms': ['bottle_opener'], 'def': 'an opener for removing caps or corks from bottles', 'name': 'bottle_opener'}, {'frequency': 'c', 'id': 135, 'synset': 'bouquet.n.01', 'synonyms': ['bouquet'], 'def': 'an arrangement of flowers that is usually given as a present', 'name': 'bouquet'}, {'frequency': 'r', 'id': 136, 'synset': 'bow.n.04', 'synonyms': ['bow_(weapon)'], 'def': 'a weapon for shooting arrows', 'name': 'bow_(weapon)'}, {'frequency': 'f', 'id': 137, 'synset': 'bow.n.08', 'synonyms': ['bow_(decorative_ribbons)'], 'def': 'a decorative interlacing of ribbons', 'name': 'bow_(decorative_ribbons)'}, {'frequency': 'f', 'id': 138, 'synset': 'bow_tie.n.01', 'synonyms': ['bow-tie', 'bowtie'], 'def': "a man's tie that ties in a bow", 'name': 'bow-tie'}, {'frequency': 'f', 'id': 139, 'synset': 'bowl.n.03', 'synonyms': ['bowl'], 'def': 'a dish that is round and open at the top for serving foods', 'name': 'bowl'}, {'frequency': 'r', 'id': 140, 'synset': 'bowl.n.08', 'synonyms': ['pipe_bowl'], 'def': 'a small round container that is open at the top for holding tobacco', 'name': 'pipe_bowl'}, {'frequency': 'c', 'id': 141, 'synset': 'bowler_hat.n.01', 'synonyms': ['bowler_hat', 'bowler', 'derby_hat', 'derby', 'plug_hat'], 'def': 'a felt hat that is round and hard with a narrow brim', 'name': 'bowler_hat'}, {'frequency': 'r', 'id': 142, 'synset': 'bowling_ball.n.01', 'synonyms': ['bowling_ball'], 'def': 'a large ball with finger holes used in the sport of bowling', 'name': 'bowling_ball'}, {'frequency': 'r', 'id': 143, 'synset': 'bowling_pin.n.01', 'synonyms': ['bowling_pin'], 'def': 'a club-shaped wooden object used in bowling', 'name': 'bowling_pin'}, {'frequency': 'r', 'id': 144, 'synset': 'boxing_glove.n.01', 'synonyms': ['boxing_glove'], 'def': 'large glove coverings the fists of a fighter worn for the sport of boxing', 'name': 'boxing_glove'}, {'frequency': 'c', 'id': 145, 'synset': 'brace.n.06', 'synonyms': ['suspenders'], 'def': 'elastic straps that hold trousers up (usually used in the plural)', 'name': 'suspenders'}, {'frequency': 'f', 'id': 146, 'synset': 'bracelet.n.02', 'synonyms': ['bracelet', 'bangle'], 'def': 'jewelry worn around the wrist for decoration', 'name': 'bracelet'}, {'frequency': 'r', 'id': 147, 'synset': 'brass.n.07', 'synonyms': ['brass_plaque'], 'def': 'a memorial made of brass', 'name': 'brass_plaque'}, {'frequency': 'c', 'id': 148, 'synset': 'brassiere.n.01', 'synonyms': ['brassiere', 'bra', 'bandeau'], 'def': 'an undergarment worn by women to support their breasts', 'name': 'brassiere'}, {'frequency': 'c', 'id': 149, 'synset': 'bread-bin.n.01', 'synonyms': ['bread-bin', 'breadbox'], 'def': 'a container used to keep bread or cake in', 'name': 'bread-bin'}, {'frequency': 'r', 'id': 150, 'synset': 'breechcloth.n.01', 'synonyms': ['breechcloth', 'breechclout', 'loincloth'], 'def': 'a garment that provides covering for the loins', 'name': 'breechcloth'}, {'frequency': 'c', 'id': 151, 'synset': 'bridal_gown.n.01', 'synonyms': ['bridal_gown', 'wedding_gown', 'wedding_dress'], 'def': 'a gown worn by the bride at a wedding', 'name': 'bridal_gown'}, {'frequency': 'c', 'id': 152, 'synset': 'briefcase.n.01', 'synonyms': ['briefcase'], 'def': 'a case with a handle; for carrying papers or files or books', 'name': 'briefcase'}, {'frequency': 'c', 'id': 153, 'synset': 'bristle_brush.n.01', 'synonyms': ['bristle_brush'], 'def': 'a brush that is made with the short stiff hairs of an animal or plant', 'name': 'bristle_brush'}, {'frequency': 'f', 'id': 154, 'synset': 'broccoli.n.01', 'synonyms': ['broccoli'], 'def': 'plant with dense clusters of tight green flower buds', 'name': 'broccoli'}, {'frequency': 'r', 'id': 155, 'synset': 'brooch.n.01', 'synonyms': ['broach'], 'def': 'a decorative pin worn by women', 'name': 'broach'}, {'frequency': 'c', 'id': 156, 'synset': 'broom.n.01', 'synonyms': ['broom'], 'def': 'bundle of straws or twigs attached to a long handle; used for cleaning', 'name': 'broom'}, {'frequency': 'c', 'id': 157, 'synset': 'brownie.n.03', 'synonyms': ['brownie'], 'def': 'square or bar of very rich chocolate cake usually with nuts', 'name': 'brownie'}, {'frequency': 'c', 'id': 158, 'synset': 'brussels_sprouts.n.01', 'synonyms': ['brussels_sprouts'], 'def': 'the small edible cabbage-like buds growing along a stalk', 'name': 'brussels_sprouts'}, {'frequency': 'r', 'id': 159, 'synset': 'bubble_gum.n.01', 'synonyms': ['bubble_gum'], 'def': 'a kind of chewing gum that can be blown into bubbles', 'name': 'bubble_gum'}, {'frequency': 'f', 'id': 160, 'synset': 'bucket.n.01', 'synonyms': ['bucket', 'pail'], 'def': 'a roughly cylindrical vessel that is open at the top', 'name': 'bucket'}, {'frequency': 'r', 'id': 161, 'synset': 'buggy.n.01', 'synonyms': ['horse_buggy'], 'def': 'a small lightweight carriage; drawn by a single horse', 'name': 'horse_buggy'}, {'frequency': 'c', 'id': 162, 'synset': 'bull.n.11', 'synonyms': ['bull'], 'def': 'mature male cow', 'name': 'bull'}, {'frequency': 'r', 'id': 163, 'synset': 'bulldog.n.01', 'synonyms': ['bulldog'], 'def': 'a thickset short-haired dog with a large head and strong undershot lower jaw', 'name': 'bulldog'}, {'frequency': 'r', 'id': 164, 'synset': 'bulldozer.n.01', 'synonyms': ['bulldozer', 'dozer'], 'def': 'large powerful tractor; a large blade in front flattens areas of ground', 'name': 'bulldozer'}, {'frequency': 'c', 'id': 165, 'synset': 'bullet_train.n.01', 'synonyms': ['bullet_train'], 'def': 'a high-speed passenger train', 'name': 'bullet_train'}, {'frequency': 'c', 'id': 166, 'synset': 'bulletin_board.n.02', 'synonyms': ['bulletin_board', 'notice_board'], 'def': 'a board that hangs on a wall; displays announcements', 'name': 'bulletin_board'}, {'frequency': 'r', 'id': 167, 'synset': 'bulletproof_vest.n.01', 'synonyms': ['bulletproof_vest'], 'def': 'a vest capable of resisting the impact of a bullet', 'name': 'bulletproof_vest'}, {'frequency': 'c', 'id': 168, 'synset': 'bullhorn.n.01', 'synonyms': ['bullhorn', 'megaphone'], 'def': 'a portable loudspeaker with built-in microphone and amplifier', 'name': 'bullhorn'}, {'frequency': 'r', 'id': 169, 'synset': 'bully_beef.n.01', 'synonyms': ['corned_beef', 'corn_beef'], 'def': 'beef cured or pickled in brine', 'name': 'corned_beef'}, {'frequency': 'f', 'id': 170, 'synset': 'bun.n.01', 'synonyms': ['bun', 'roll'], 'def': 'small rounded bread either plain or sweet', 'name': 'bun'}, {'frequency': 'c', 'id': 171, 'synset': 'bunk_bed.n.01', 'synonyms': ['bunk_bed'], 'def': 'beds built one above the other', 'name': 'bunk_bed'}, {'frequency': 'f', 'id': 172, 'synset': 'buoy.n.01', 'synonyms': ['buoy'], 'def': 'a float attached by rope to the seabed to mark channels in a harbor or underwater hazards', 'name': 'buoy'}, {'frequency': 'r', 'id': 173, 'synset': 'burrito.n.01', 'synonyms': ['burrito'], 'def': 'a flour tortilla folded around a filling', 'name': 'burrito'}, {'frequency': 'f', 'id': 174, 'synset': 'bus.n.01', 'synonyms': ['bus_(vehicle)', 'autobus', 'charabanc', 'double-decker', 'motorbus', 'motorcoach'], 'def': 'a vehicle carrying many passengers; used for public transport', 'name': 'bus_(vehicle)'}, {'frequency': 'c', 'id': 175, 'synset': 'business_card.n.01', 'synonyms': ['business_card'], 'def': "a card on which are printed the person's name and business affiliation", 'name': 'business_card'}, {'frequency': 'c', 'id': 176, 'synset': 'butcher_knife.n.01', 'synonyms': ['butcher_knife'], 'def': 'a large sharp knife for cutting or trimming meat', 'name': 'butcher_knife'}, {'frequency': 'c', 'id': 177, 'synset': 'butter.n.01', 'synonyms': ['butter'], 'def': 'an edible emulsion of fat globules made by churning milk or cream; for cooking and table use', 'name': 'butter'}, {'frequency': 'c', 'id': 178, 'synset': 'butterfly.n.01', 'synonyms': ['butterfly'], 'def': 'insect typically having a slender body with knobbed antennae and broad colorful wings', 'name': 'butterfly'}, {'frequency': 'f', 'id': 179, 'synset': 'button.n.01', 'synonyms': ['button'], 'def': 'a round fastener sewn to shirts and coats etc to fit through buttonholes', 'name': 'button'}, {'frequency': 'f', 'id': 180, 'synset': 'cab.n.03', 'synonyms': ['cab_(taxi)', 'taxi', 'taxicab'], 'def': 'a car that takes passengers where they want to go in exchange for money', 'name': 'cab_(taxi)'}, {'frequency': 'r', 'id': 181, 'synset': 'cabana.n.01', 'synonyms': ['cabana'], 'def': 'a small tent used as a dressing room beside the sea or a swimming pool', 'name': 'cabana'}, {'frequency': 'r', 'id': 182, 'synset': 'cabin_car.n.01', 'synonyms': ['cabin_car', 'caboose'], 'def': 'a car on a freight train for use of the train crew; usually the last car on the train', 'name': 'cabin_car'}, {'frequency': 'f', 'id': 183, 'synset': 'cabinet.n.01', 'synonyms': ['cabinet'], 'def': 'a piece of furniture resembling a cupboard with doors and shelves and drawers', 'name': 'cabinet'}, {'frequency': 'r', 'id': 184, 'synset': 'cabinet.n.03', 'synonyms': ['locker', 'storage_locker'], 'def': 'a storage compartment for clothes and valuables; usually it has a lock', 'name': 'locker'}, {'frequency': 'f', 'id': 185, 'synset': 'cake.n.03', 'synonyms': ['cake'], 'def': 'baked goods made from or based on a mixture of flour, sugar, eggs, and fat', 'name': 'cake'}, {'frequency': 'c', 'id': 186, 'synset': 'calculator.n.02', 'synonyms': ['calculator'], 'def': 'a small machine that is used for mathematical calculations', 'name': 'calculator'}, {'frequency': 'f', 'id': 187, 'synset': 'calendar.n.02', 'synonyms': ['calendar'], 'def': 'a list or register of events (appointments/social events/court cases, etc)', 'name': 'calendar'}, {'frequency': 'c', 'id': 188, 'synset': 'calf.n.01', 'synonyms': ['calf'], 'def': 'young of domestic cattle', 'name': 'calf'}, {'frequency': 'c', 'id': 189, 'synset': 'camcorder.n.01', 'synonyms': ['camcorder'], 'def': 'a portable television camera and videocassette recorder', 'name': 'camcorder'}, {'frequency': 'c', 'id': 190, 'synset': 'camel.n.01', 'synonyms': ['camel'], 'def': 'cud-chewing mammal used as a draft or saddle animal in desert regions', 'name': 'camel'}, {'frequency': 'f', 'id': 191, 'synset': 'camera.n.01', 'synonyms': ['camera'], 'def': 'equipment for taking photographs', 'name': 'camera'}, {'frequency': 'c', 'id': 192, 'synset': 'camera_lens.n.01', 'synonyms': ['camera_lens'], 'def': 'a lens that focuses the image in a camera', 'name': 'camera_lens'}, {'frequency': 'c', 'id': 193, 'synset': 'camper.n.02', 'synonyms': ['camper_(vehicle)', 'camping_bus', 'motor_home'], 'def': 'a recreational vehicle equipped for camping out while traveling', 'name': 'camper_(vehicle)'}, {'frequency': 'f', 'id': 194, 'synset': 'can.n.01', 'synonyms': ['can', 'tin_can'], 'def': 'airtight sealed metal container for food or drink or paint etc.', 'name': 'can'}, {'frequency': 'c', 'id': 195, 'synset': 'can_opener.n.01', 'synonyms': ['can_opener', 'tin_opener'], 'def': 'a device for cutting cans open', 'name': 'can_opener'}, {'frequency': 'r', 'id': 196, 'synset': 'candelabrum.n.01', 'synonyms': ['candelabrum', 'candelabra'], 'def': 'branched candlestick; ornamental; has several lights', 'name': 'candelabrum'}, {'frequency': 'f', 'id': 197, 'synset': 'candle.n.01', 'synonyms': ['candle', 'candlestick'], 'def': 'stick of wax with a wick in the middle', 'name': 'candle'}, {'frequency': 'f', 'id': 198, 'synset': 'candlestick.n.01', 'synonyms': ['candle_holder'], 'def': 'a holder with sockets for candles', 'name': 'candle_holder'}, {'frequency': 'r', 'id': 199, 'synset': 'candy_bar.n.01', 'synonyms': ['candy_bar'], 'def': 'a candy shaped as a bar', 'name': 'candy_bar'}, {'frequency': 'c', 'id': 200, 'synset': 'candy_cane.n.01', 'synonyms': ['candy_cane'], 'def': 'a hard candy in the shape of a rod (usually with stripes)', 'name': 'candy_cane'}, {'frequency': 'c', 'id': 201, 'synset': 'cane.n.01', 'synonyms': ['walking_cane'], 'def': 'a stick that people can lean on to help them walk', 'name': 'walking_cane'}, {'frequency': 'c', 'id': 202, 'synset': 'canister.n.02', 'synonyms': ['canister', 'cannister'], 'def': 'metal container for storing dry foods such as tea or flour', 'name': 'canister'}, {'frequency': 'r', 'id': 203, 'synset': 'cannon.n.02', 'synonyms': ['cannon'], 'def': 'heavy gun fired from a tank', 'name': 'cannon'}, {'frequency': 'c', 'id': 204, 'synset': 'canoe.n.01', 'synonyms': ['canoe'], 'def': 'small and light boat; pointed at both ends; propelled with a paddle', 'name': 'canoe'}, {'frequency': 'r', 'id': 205, 'synset': 'cantaloup.n.02', 'synonyms': ['cantaloup', 'cantaloupe'], 'def': 'the fruit of a cantaloup vine; small to medium-sized melon with yellowish flesh', 'name': 'cantaloup'}, {'frequency': 'r', 'id': 206, 'synset': 'canteen.n.01', 'synonyms': ['canteen'], 'def': 'a flask for carrying water; used by soldiers or travelers', 'name': 'canteen'}, {'frequency': 'c', 'id': 207, 'synset': 'cap.n.01', 'synonyms': ['cap_(headwear)'], 'def': 'a tight-fitting headwear', 'name': 'cap_(headwear)'}, {'frequency': 'f', 'id': 208, 'synset': 'cap.n.02', 'synonyms': ['bottle_cap', 'cap_(container_lid)'], 'def': 'a top (as for a bottle)', 'name': 'bottle_cap'}, {'frequency': 'r', 'id': 209, 'synset': 'cape.n.02', 'synonyms': ['cape'], 'def': 'a sleeveless garment like a cloak but shorter', 'name': 'cape'}, {'frequency': 'c', 'id': 210, 'synset': 'cappuccino.n.01', 'synonyms': ['cappuccino', 'coffee_cappuccino'], 'def': 'equal parts of espresso and steamed milk', 'name': 'cappuccino'}, {'frequency': 'f', 'id': 211, 'synset': 'car.n.01', 'synonyms': ['car_(automobile)', 'auto_(automobile)', 'automobile'], 'def': 'a motor vehicle with four wheels', 'name': 'car_(automobile)'}, {'frequency': 'f', 'id': 212, 'synset': 'car.n.02', 'synonyms': ['railcar_(part_of_a_train)', 'railway_car_(part_of_a_train)', 'railroad_car_(part_of_a_train)'], 'def': 'a wheeled vehicle adapted to the rails of railroad', 'name': 'railcar_(part_of_a_train)'}, {'frequency': 'r', 'id': 213, 'synset': 'car.n.04', 'synonyms': ['elevator_car'], 'def': 'where passengers ride up and down', 'name': 'elevator_car'}, {'frequency': 'r', 'id': 214, 'synset': 'car_battery.n.01', 'synonyms': ['car_battery', 'automobile_battery'], 'def': 'a battery in a motor vehicle', 'name': 'car_battery'}, {'frequency': 'c', 'id': 215, 'synset': 'card.n.02', 'synonyms': ['identity_card'], 'def': 'a card certifying the identity of the bearer', 'name': 'identity_card'}, {'frequency': 'c', 'id': 216, 'synset': 'card.n.03', 'synonyms': ['card'], 'def': 'a rectangular piece of paper used to send messages (e.g. greetings or pictures)', 'name': 'card'}, {'frequency': 'r', 'id': 217, 'synset': 'cardigan.n.01', 'synonyms': ['cardigan'], 'def': 'knitted jacket that is fastened up the front with buttons or a zipper', 'name': 'cardigan'}, {'frequency': 'r', 'id': 218, 'synset': 'cargo_ship.n.01', 'synonyms': ['cargo_ship', 'cargo_vessel'], 'def': 'a ship designed to carry cargo', 'name': 'cargo_ship'}, {'frequency': 'r', 'id': 219, 'synset': 'carnation.n.01', 'synonyms': ['carnation'], 'def': 'plant with pink to purple-red spice-scented usually double flowers', 'name': 'carnation'}, {'frequency': 'c', 'id': 220, 'synset': 'carriage.n.02', 'synonyms': ['horse_carriage'], 'def': 'a vehicle with wheels drawn by one or more horses', 'name': 'horse_carriage'}, {'frequency': 'f', 'id': 221, 'synset': 'carrot.n.01', 'synonyms': ['carrot'], 'def': 'deep orange edible root of the cultivated carrot plant', 'name': 'carrot'}, {'frequency': 'c', 'id': 222, 'synset': 'carryall.n.01', 'synonyms': ['tote_bag'], 'def': 'a capacious bag or basket', 'name': 'tote_bag'}, {'frequency': 'c', 'id': 223, 'synset': 'cart.n.01', 'synonyms': ['cart'], 'def': 'a heavy open wagon usually having two wheels and drawn by an animal', 'name': 'cart'}, {'frequency': 'c', 'id': 224, 'synset': 'carton.n.02', 'synonyms': ['carton'], 'def': 'a box made of cardboard; opens by flaps on top', 'name': 'carton'}, {'frequency': 'c', 'id': 225, 'synset': 'cash_register.n.01', 'synonyms': ['cash_register', 'register_(for_cash_transactions)'], 'def': 'a cashbox with an adding machine to register transactions', 'name': 'cash_register'}, {'frequency': 'r', 'id': 226, 'synset': 'casserole.n.01', 'synonyms': ['casserole'], 'def': 'food cooked and served in a casserole', 'name': 'casserole'}, {'frequency': 'r', 'id': 227, 'synset': 'cassette.n.01', 'synonyms': ['cassette'], 'def': 'a container that holds a magnetic tape used for recording or playing sound or video', 'name': 'cassette'}, {'frequency': 'c', 'id': 228, 'synset': 'cast.n.05', 'synonyms': ['cast', 'plaster_cast', 'plaster_bandage'], 'def': 'bandage consisting of a firm covering that immobilizes broken bones while they heal', 'name': 'cast'}, {'frequency': 'f', 'id': 229, 'synset': 'cat.n.01', 'synonyms': ['cat'], 'def': 'a domestic house cat', 'name': 'cat'}, {'frequency': 'c', 'id': 230, 'synset': 'cauliflower.n.02', 'synonyms': ['cauliflower'], 'def': 'edible compact head of white undeveloped flowers', 'name': 'cauliflower'}, {'frequency': 'r', 'id': 231, 'synset': 'caviar.n.01', 'synonyms': ['caviar', 'caviare'], 'def': "salted roe of sturgeon or other large fish; usually served as an hors d'oeuvre", 'name': 'caviar'}, {'frequency': 'c', 'id': 232, 'synset': 'cayenne.n.02', 'synonyms': ['cayenne_(spice)', 'cayenne_pepper_(spice)', 'red_pepper_(spice)'], 'def': 'ground pods and seeds of pungent red peppers of the genus Capsicum', 'name': 'cayenne_(spice)'}, {'frequency': 'c', 'id': 233, 'synset': 'cd_player.n.01', 'synonyms': ['CD_player'], 'def': 'electronic equipment for playing compact discs (CDs)', 'name': 'CD_player'}, {'frequency': 'c', 'id': 234, 'synset': 'celery.n.01', 'synonyms': ['celery'], 'def': 'widely cultivated herb with aromatic leaf stalks that are eaten raw or cooked', 'name': 'celery'}, {'frequency': 'f', 'id': 235, 'synset': 'cellular_telephone.n.01', 'synonyms': ['cellular_telephone', 'cellular_phone', 'cellphone', 'mobile_phone', 'smart_phone'], 'def': 'a hand-held mobile telephone', 'name': 'cellular_telephone'}, {'frequency': 'r', 'id': 236, 'synset': 'chain_mail.n.01', 'synonyms': ['chain_mail', 'ring_mail', 'chain_armor', 'chain_armour', 'ring_armor', 'ring_armour'], 'def': '(Middle Ages) flexible armor made of interlinked metal rings', 'name': 'chain_mail'}, {'frequency': 'f', 'id': 237, 'synset': 'chair.n.01', 'synonyms': ['chair'], 'def': 'a seat for one person, with a support for the back', 'name': 'chair'}, {'frequency': 'r', 'id': 238, 'synset': 'chaise_longue.n.01', 'synonyms': ['chaise_longue', 'chaise', 'daybed'], 'def': 'a long chair; for reclining', 'name': 'chaise_longue'}, {'frequency': 'r', 'id': 239, 'synset': 'champagne.n.01', 'synonyms': ['champagne'], 'def': 'a white sparkling wine produced in Champagne or resembling that produced there', 'name': 'champagne'}, {'frequency': 'f', 'id': 240, 'synset': 'chandelier.n.01', 'synonyms': ['chandelier'], 'def': 'branched lighting fixture; often ornate; hangs from the ceiling', 'name': 'chandelier'}, {'frequency': 'r', 'id': 241, 'synset': 'chap.n.04', 'synonyms': ['chap'], 'def': 'leather leggings without a seat; worn over trousers by cowboys to protect their legs', 'name': 'chap'}, {'frequency': 'r', 'id': 242, 'synset': 'checkbook.n.01', 'synonyms': ['checkbook', 'chequebook'], 'def': 'a book issued to holders of checking accounts', 'name': 'checkbook'}, {'frequency': 'r', 'id': 243, 'synset': 'checkerboard.n.01', 'synonyms': ['checkerboard'], 'def': 'a board having 64 squares of two alternating colors', 'name': 'checkerboard'}, {'frequency': 'c', 'id': 244, 'synset': 'cherry.n.03', 'synonyms': ['cherry'], 'def': 'a red fruit with a single hard stone', 'name': 'cherry'}, {'frequency': 'r', 'id': 245, 'synset': 'chessboard.n.01', 'synonyms': ['chessboard'], 'def': 'a checkerboard used to play chess', 'name': 'chessboard'}, {'frequency': 'r', 'id': 246, 'synset': 'chest_of_drawers.n.01', 'synonyms': ['chest_of_drawers_(furniture)', 'bureau_(furniture)', 'chest_(furniture)'], 'def': 'furniture with drawers for keeping clothes', 'name': 'chest_of_drawers_(furniture)'}, {'frequency': 'c', 'id': 247, 'synset': 'chicken.n.02', 'synonyms': ['chicken_(animal)'], 'def': 'a domestic fowl bred for flesh or eggs', 'name': 'chicken_(animal)'}, {'frequency': 'c', 'id': 248, 'synset': 'chicken_wire.n.01', 'synonyms': ['chicken_wire'], 'def': 'a galvanized wire network with a hexagonal mesh; used to build fences', 'name': 'chicken_wire'}, {'frequency': 'r', 'id': 249, 'synset': 'chickpea.n.01', 'synonyms': ['chickpea', 'garbanzo'], 'def': 'the seed of the chickpea plant; usually dried', 'name': 'chickpea'}, {'frequency': 'r', 'id': 250, 'synset': 'chihuahua.n.03', 'synonyms': ['Chihuahua'], 'def': 'an old breed of tiny short-haired dog with protruding eyes from Mexico', 'name': 'Chihuahua'}, {'frequency': 'r', 'id': 251, 'synset': 'chili.n.02', 'synonyms': ['chili_(vegetable)', 'chili_pepper_(vegetable)', 'chilli_(vegetable)', 'chilly_(vegetable)', 'chile_(vegetable)'], 'def': 'very hot and finely tapering pepper of special pungency', 'name': 'chili_(vegetable)'}, {'frequency': 'r', 'id': 252, 'synset': 'chime.n.01', 'synonyms': ['chime', 'gong'], 'def': 'an instrument consisting of a set of bells that are struck with a hammer', 'name': 'chime'}, {'frequency': 'r', 'id': 253, 'synset': 'chinaware.n.01', 'synonyms': ['chinaware'], 'def': 'dishware made of high quality porcelain', 'name': 'chinaware'}, {'frequency': 'c', 'id': 254, 'synset': 'chip.n.04', 'synonyms': ['crisp_(potato_chip)', 'potato_chip'], 'def': 'a thin crisp slice of potato fried in deep fat', 'name': 'crisp_(potato_chip)'}, {'frequency': 'r', 'id': 255, 'synset': 'chip.n.06', 'synonyms': ['poker_chip'], 'def': 'a small disk-shaped counter used to represent money when gambling', 'name': 'poker_chip'}, {'frequency': 'c', 'id': 256, 'synset': 'chocolate_bar.n.01', 'synonyms': ['chocolate_bar'], 'def': 'a bar of chocolate candy', 'name': 'chocolate_bar'}, {'frequency': 'c', 'id': 257, 'synset': 'chocolate_cake.n.01', 'synonyms': ['chocolate_cake'], 'def': 'cake containing chocolate', 'name': 'chocolate_cake'}, {'frequency': 'r', 'id': 258, 'synset': 'chocolate_milk.n.01', 'synonyms': ['chocolate_milk'], 'def': 'milk flavored with chocolate syrup', 'name': 'chocolate_milk'}, {'frequency': 'r', 'id': 259, 'synset': 'chocolate_mousse.n.01', 'synonyms': ['chocolate_mousse'], 'def': 'dessert mousse made with chocolate', 'name': 'chocolate_mousse'}, {'frequency': 'f', 'id': 260, 'synset': 'choker.n.03', 'synonyms': ['choker', 'collar', 'neckband'], 'def': 'necklace that fits tightly around the neck', 'name': 'choker'}, {'frequency': 'f', 'id': 261, 'synset': 'chopping_board.n.01', 'synonyms': ['chopping_board', 'cutting_board', 'chopping_block'], 'def': 'a wooden board where meats or vegetables can be cut', 'name': 'chopping_board'}, {'frequency': 'c', 'id': 262, 'synset': 'chopstick.n.01', 'synonyms': ['chopstick'], 'def': 'one of a pair of slender sticks used as oriental tableware to eat food with', 'name': 'chopstick'}, {'frequency': 'f', 'id': 263, 'synset': 'christmas_tree.n.05', 'synonyms': ['Christmas_tree'], 'def': 'an ornamented evergreen used as a Christmas decoration', 'name': 'Christmas_tree'}, {'frequency': 'c', 'id': 264, 'synset': 'chute.n.02', 'synonyms': ['slide'], 'def': 'sloping channel through which things can descend', 'name': 'slide'}, {'frequency': 'r', 'id': 265, 'synset': 'cider.n.01', 'synonyms': ['cider', 'cyder'], 'def': 'a beverage made from juice pressed from apples', 'name': 'cider'}, {'frequency': 'r', 'id': 266, 'synset': 'cigar_box.n.01', 'synonyms': ['cigar_box'], 'def': 'a box for holding cigars', 'name': 'cigar_box'}, {'frequency': 'c', 'id': 267, 'synset': 'cigarette.n.01', 'synonyms': ['cigarette'], 'def': 'finely ground tobacco wrapped in paper; for smoking', 'name': 'cigarette'}, {'frequency': 'c', 'id': 268, 'synset': 'cigarette_case.n.01', 'synonyms': ['cigarette_case', 'cigarette_pack'], 'def': 'a small flat case for holding cigarettes', 'name': 'cigarette_case'}, {'frequency': 'f', 'id': 269, 'synset': 'cistern.n.02', 'synonyms': ['cistern', 'water_tank'], 'def': 'a tank that holds the water used to flush a toilet', 'name': 'cistern'}, {'frequency': 'r', 'id': 270, 'synset': 'clarinet.n.01', 'synonyms': ['clarinet'], 'def': 'a single-reed instrument with a straight tube', 'name': 'clarinet'}, {'frequency': 'r', 'id': 271, 'synset': 'clasp.n.01', 'synonyms': ['clasp'], 'def': 'a fastener (as a buckle or hook) that is used to hold two things together', 'name': 'clasp'}, {'frequency': 'c', 'id': 272, 'synset': 'cleansing_agent.n.01', 'synonyms': ['cleansing_agent', 'cleanser', 'cleaner'], 'def': 'a preparation used in cleaning something', 'name': 'cleansing_agent'}, {'frequency': 'r', 'id': 273, 'synset': 'clementine.n.01', 'synonyms': ['clementine'], 'def': 'a variety of mandarin orange', 'name': 'clementine'}, {'frequency': 'c', 'id': 274, 'synset': 'clip.n.03', 'synonyms': ['clip'], 'def': 'any of various small fasteners used to hold loose articles together', 'name': 'clip'}, {'frequency': 'c', 'id': 275, 'synset': 'clipboard.n.01', 'synonyms': ['clipboard'], 'def': 'a small writing board with a clip at the top for holding papers', 'name': 'clipboard'}, {'frequency': 'f', 'id': 276, 'synset': 'clock.n.01', 'synonyms': ['clock', 'timepiece', 'timekeeper'], 'def': 'a timepiece that shows the time of day', 'name': 'clock'}, {'frequency': 'f', 'id': 277, 'synset': 'clock_tower.n.01', 'synonyms': ['clock_tower'], 'def': 'a tower with a large clock visible high up on an outside face', 'name': 'clock_tower'}, {'frequency': 'c', 'id': 278, 'synset': 'clothes_hamper.n.01', 'synonyms': ['clothes_hamper', 'laundry_basket', 'clothes_basket'], 'def': 'a hamper that holds dirty clothes to be washed or wet clothes to be dried', 'name': 'clothes_hamper'}, {'frequency': 'c', 'id': 279, 'synset': 'clothespin.n.01', 'synonyms': ['clothespin', 'clothes_peg'], 'def': 'wood or plastic fastener; for holding clothes on a clothesline', 'name': 'clothespin'}, {'frequency': 'r', 'id': 280, 'synset': 'clutch_bag.n.01', 'synonyms': ['clutch_bag'], 'def': "a woman's strapless purse that is carried in the hand", 'name': 'clutch_bag'}, {'frequency': 'f', 'id': 281, 'synset': 'coaster.n.03', 'synonyms': ['coaster'], 'def': 'a covering (plate or mat) that protects the surface of a table', 'name': 'coaster'}, {'frequency': 'f', 'id': 282, 'synset': 'coat.n.01', 'synonyms': ['coat'], 'def': 'an outer garment that has sleeves and covers the body from shoulder down', 'name': 'coat'}, {'frequency': 'c', 'id': 283, 'synset': 'coat_hanger.n.01', 'synonyms': ['coat_hanger', 'clothes_hanger', 'dress_hanger'], 'def': "a hanger that is shaped like a person's shoulders", 'name': 'coat_hanger'}, {'frequency': 'r', 'id': 284, 'synset': 'coatrack.n.01', 'synonyms': ['coatrack', 'hatrack'], 'def': 'a rack with hooks for temporarily holding coats and hats', 'name': 'coatrack'}, {'frequency': 'c', 'id': 285, 'synset': 'cock.n.04', 'synonyms': ['cock', 'rooster'], 'def': 'adult male chicken', 'name': 'cock'}, {'frequency': 'c', 'id': 286, 'synset': 'coconut.n.02', 'synonyms': ['coconut', 'cocoanut'], 'def': 'large hard-shelled brown oval nut with a fibrous husk', 'name': 'coconut'}, {'frequency': 'r', 'id': 287, 'synset': 'coffee_filter.n.01', 'synonyms': ['coffee_filter'], 'def': 'filter (usually of paper) that passes the coffee and retains the coffee grounds', 'name': 'coffee_filter'}, {'frequency': 'f', 'id': 288, 'synset': 'coffee_maker.n.01', 'synonyms': ['coffee_maker', 'coffee_machine'], 'def': 'a kitchen appliance for brewing coffee automatically', 'name': 'coffee_maker'}, {'frequency': 'f', 'id': 289, 'synset': 'coffee_table.n.01', 'synonyms': ['coffee_table', 'cocktail_table'], 'def': 'low table where magazines can be placed and coffee or cocktails are served', 'name': 'coffee_table'}, {'frequency': 'c', 'id': 290, 'synset': 'coffeepot.n.01', 'synonyms': ['coffeepot'], 'def': 'tall pot in which coffee is brewed', 'name': 'coffeepot'}, {'frequency': 'r', 'id': 291, 'synset': 'coil.n.05', 'synonyms': ['coil'], 'def': 'tubing that is wound in a spiral', 'name': 'coil'}, {'frequency': 'c', 'id': 292, 'synset': 'coin.n.01', 'synonyms': ['coin'], 'def': 'a flat metal piece (usually a disc) used as money', 'name': 'coin'}, {'frequency': 'r', 'id': 293, 'synset': 'colander.n.01', 'synonyms': ['colander', 'cullender'], 'def': 'bowl-shaped strainer; used to wash or drain foods', 'name': 'colander'}, {'frequency': 'c', 'id': 294, 'synset': 'coleslaw.n.01', 'synonyms': ['coleslaw', 'slaw'], 'def': 'basically shredded cabbage', 'name': 'coleslaw'}, {'frequency': 'r', 'id': 295, 'synset': 'coloring_material.n.01', 'synonyms': ['coloring_material', 'colouring_material'], 'def': 'any material used for its color', 'name': 'coloring_material'}, {'frequency': 'r', 'id': 296, 'synset': 'combination_lock.n.01', 'synonyms': ['combination_lock'], 'def': 'lock that can be opened only by turning dials in a special sequence', 'name': 'combination_lock'}, {'frequency': 'c', 'id': 297, 'synset': 'comforter.n.04', 'synonyms': ['pacifier', 'teething_ring'], 'def': 'device used for an infant to suck or bite on', 'name': 'pacifier'}, {'frequency': 'r', 'id': 298, 'synset': 'comic_book.n.01', 'synonyms': ['comic_book'], 'def': 'a magazine devoted to comic strips', 'name': 'comic_book'}, {'frequency': 'f', 'id': 299, 'synset': 'computer_keyboard.n.01', 'synonyms': ['computer_keyboard', 'keyboard_(computer)'], 'def': 'a keyboard that is a data input device for computers', 'name': 'computer_keyboard'}, {'frequency': 'r', 'id': 300, 'synset': 'concrete_mixer.n.01', 'synonyms': ['concrete_mixer', 'cement_mixer'], 'def': 'a machine with a large revolving drum in which cement/concrete is mixed', 'name': 'concrete_mixer'}, {'frequency': 'f', 'id': 301, 'synset': 'cone.n.01', 'synonyms': ['cone', 'traffic_cone'], 'def': 'a cone-shaped object used to direct traffic', 'name': 'cone'}, {'frequency': 'f', 'id': 302, 'synset': 'control.n.09', 'synonyms': ['control', 'controller'], 'def': 'a mechanism that controls the operation of a machine', 'name': 'control'}, {'frequency': 'r', 'id': 303, 'synset': 'convertible.n.01', 'synonyms': ['convertible_(automobile)'], 'def': 'a car that has top that can be folded or removed', 'name': 'convertible_(automobile)'}, {'frequency': 'r', 'id': 304, 'synset': 'convertible.n.03', 'synonyms': ['sofa_bed'], 'def': 'a sofa that can be converted into a bed', 'name': 'sofa_bed'}, {'frequency': 'c', 'id': 305, 'synset': 'cookie.n.01', 'synonyms': ['cookie', 'cooky', 'biscuit_(cookie)'], 'def': "any of various small flat sweet cakes (`biscuit' is the British term)", 'name': 'cookie'}, {'frequency': 'r', 'id': 306, 'synset': 'cookie_jar.n.01', 'synonyms': ['cookie_jar', 'cooky_jar'], 'def': 'a jar in which cookies are kept (and sometimes money is hidden)', 'name': 'cookie_jar'}, {'frequency': 'r', 'id': 307, 'synset': 'cooking_utensil.n.01', 'synonyms': ['cooking_utensil'], 'def': 'a kitchen utensil made of material that does not melt easily; used for cooking', 'name': 'cooking_utensil'}, {'frequency': 'f', 'id': 308, 'synset': 'cooler.n.01', 'synonyms': ['cooler_(for_food)', 'ice_chest'], 'def': 'an insulated box for storing food often with ice', 'name': 'cooler_(for_food)'}, {'frequency': 'c', 'id': 309, 'synset': 'cork.n.04', 'synonyms': ['cork_(bottle_plug)', 'bottle_cork'], 'def': 'the plug in the mouth of a bottle (especially a wine bottle)', 'name': 'cork_(bottle_plug)'}, {'frequency': 'r', 'id': 310, 'synset': 'corkboard.n.01', 'synonyms': ['corkboard'], 'def': 'a sheet consisting of cork granules', 'name': 'corkboard'}, {'frequency': 'r', 'id': 311, 'synset': 'corkscrew.n.01', 'synonyms': ['corkscrew', 'bottle_screw'], 'def': 'a bottle opener that pulls corks', 'name': 'corkscrew'}, {'frequency': 'c', 'id': 312, 'synset': 'corn.n.03', 'synonyms': ['edible_corn', 'corn', 'maize'], 'def': 'ears of corn that can be prepared and served for human food', 'name': 'edible_corn'}, {'frequency': 'r', 'id': 313, 'synset': 'cornbread.n.01', 'synonyms': ['cornbread'], 'def': 'bread made primarily of cornmeal', 'name': 'cornbread'}, {'frequency': 'c', 'id': 314, 'synset': 'cornet.n.01', 'synonyms': ['cornet', 'horn', 'trumpet'], 'def': 'a brass musical instrument with a narrow tube and a flared bell and many valves', 'name': 'cornet'}, {'frequency': 'c', 'id': 315, 'synset': 'cornice.n.01', 'synonyms': ['cornice', 'valance', 'valance_board', 'pelmet'], 'def': 'a decorative framework to conceal curtain fixtures at the top of a window casing', 'name': 'cornice'}, {'frequency': 'r', 'id': 316, 'synset': 'cornmeal.n.01', 'synonyms': ['cornmeal'], 'def': 'coarsely ground corn', 'name': 'cornmeal'}, {'frequency': 'r', 'id': 317, 'synset': 'corset.n.01', 'synonyms': ['corset', 'girdle'], 'def': "a woman's close-fitting foundation garment", 'name': 'corset'}, {'frequency': 'r', 'id': 318, 'synset': 'cos.n.02', 'synonyms': ['romaine_lettuce'], 'def': 'lettuce with long dark-green leaves in a loosely packed elongated head', 'name': 'romaine_lettuce'}, {'frequency': 'c', 'id': 319, 'synset': 'costume.n.04', 'synonyms': ['costume'], 'def': 'the attire characteristic of a country or a time or a social class', 'name': 'costume'}, {'frequency': 'r', 'id': 320, 'synset': 'cougar.n.01', 'synonyms': ['cougar', 'puma', 'catamount', 'mountain_lion', 'panther'], 'def': 'large American feline resembling a lion', 'name': 'cougar'}, {'frequency': 'r', 'id': 321, 'synset': 'coverall.n.01', 'synonyms': ['coverall'], 'def': 'a loose-fitting protective garment that is worn over other clothing', 'name': 'coverall'}, {'frequency': 'r', 'id': 322, 'synset': 'cowbell.n.01', 'synonyms': ['cowbell'], 'def': 'a bell hung around the neck of cow so that the cow can be easily located', 'name': 'cowbell'}, {'frequency': 'f', 'id': 323, 'synset': 'cowboy_hat.n.01', 'synonyms': ['cowboy_hat', 'ten-gallon_hat'], 'def': 'a hat with a wide brim and a soft crown; worn by American ranch hands', 'name': 'cowboy_hat'}, {'frequency': 'r', 'id': 324, 'synset': 'crab.n.01', 'synonyms': ['crab_(animal)'], 'def': 'decapod having eyes on short stalks and a broad flattened shell and pincers', 'name': 'crab_(animal)'}, {'frequency': 'c', 'id': 325, 'synset': 'cracker.n.01', 'synonyms': ['cracker'], 'def': 'a thin crisp wafer', 'name': 'cracker'}, {'frequency': 'r', 'id': 326, 'synset': 'crape.n.01', 'synonyms': ['crape', 'crepe', 'French_pancake'], 'def': 'small very thin pancake', 'name': 'crape'}, {'frequency': 'f', 'id': 327, 'synset': 'crate.n.01', 'synonyms': ['crate'], 'def': 'a rugged box (usually made of wood); used for shipping', 'name': 'crate'}, {'frequency': 'r', 'id': 328, 'synset': 'crayon.n.01', 'synonyms': ['crayon', 'wax_crayon'], 'def': 'writing or drawing implement made of a colored stick of composition wax', 'name': 'crayon'}, {'frequency': 'r', 'id': 329, 'synset': 'cream_pitcher.n.01', 'synonyms': ['cream_pitcher'], 'def': 'a small pitcher for serving cream', 'name': 'cream_pitcher'}, {'frequency': 'r', 'id': 330, 'synset': 'credit_card.n.01', 'synonyms': ['credit_card', 'charge_card', 'debit_card'], 'def': 'a card, usually plastic, used to pay for goods and services', 'name': 'credit_card'}, {'frequency': 'c', 'id': 331, 'synset': 'crescent_roll.n.01', 'synonyms': ['crescent_roll', 'croissant'], 'def': 'very rich flaky crescent-shaped roll', 'name': 'crescent_roll'}, {'frequency': 'c', 'id': 332, 'synset': 'crib.n.01', 'synonyms': ['crib', 'cot'], 'def': 'baby bed with high sides made of slats', 'name': 'crib'}, {'frequency': 'c', 'id': 333, 'synset': 'crock.n.03', 'synonyms': ['crock_pot', 'earthenware_jar'], 'def': 'an earthen jar (made of baked clay)', 'name': 'crock_pot'}, {'frequency': 'f', 'id': 334, 'synset': 'crossbar.n.01', 'synonyms': ['crossbar'], 'def': 'a horizontal bar that goes across something', 'name': 'crossbar'}, {'frequency': 'r', 'id': 335, 'synset': 'crouton.n.01', 'synonyms': ['crouton'], 'def': 'a small piece of toasted or fried bread; served in soup or salads', 'name': 'crouton'}, {'frequency': 'r', 'id': 336, 'synset': 'crow.n.01', 'synonyms': ['crow'], 'def': 'black birds having a raucous call', 'name': 'crow'}, {'frequency': 'c', 'id': 337, 'synset': 'crown.n.04', 'synonyms': ['crown'], 'def': 'an ornamental jeweled headdress signifying sovereignty', 'name': 'crown'}, {'frequency': 'c', 'id': 338, 'synset': 'crucifix.n.01', 'synonyms': ['crucifix'], 'def': 'representation of the cross on which Jesus died', 'name': 'crucifix'}, {'frequency': 'c', 'id': 339, 'synset': 'cruise_ship.n.01', 'synonyms': ['cruise_ship', 'cruise_liner'], 'def': 'a passenger ship used commercially for pleasure cruises', 'name': 'cruise_ship'}, {'frequency': 'c', 'id': 340, 'synset': 'cruiser.n.01', 'synonyms': ['police_cruiser', 'patrol_car', 'police_car', 'squad_car'], 'def': 'a car in which policemen cruise the streets', 'name': 'police_cruiser'}, {'frequency': 'c', 'id': 341, 'synset': 'crumb.n.03', 'synonyms': ['crumb'], 'def': 'small piece of e.g. bread or cake', 'name': 'crumb'}, {'frequency': 'r', 'id': 342, 'synset': 'crutch.n.01', 'synonyms': ['crutch'], 'def': 'a wooden or metal staff that fits under the armpit and reaches to the ground', 'name': 'crutch'}, {'frequency': 'c', 'id': 343, 'synset': 'cub.n.03', 'synonyms': ['cub_(animal)'], 'def': 'the young of certain carnivorous mammals such as the bear or wolf or lion', 'name': 'cub_(animal)'}, {'frequency': 'r', 'id': 344, 'synset': 'cube.n.05', 'synonyms': ['cube', 'square_block'], 'def': 'a block in the (approximate) shape of a cube', 'name': 'cube'}, {'frequency': 'f', 'id': 345, 'synset': 'cucumber.n.02', 'synonyms': ['cucumber', 'cuke'], 'def': 'cylindrical green fruit with thin green rind and white flesh eaten as a vegetable', 'name': 'cucumber'}, {'frequency': 'c', 'id': 346, 'synset': 'cufflink.n.01', 'synonyms': ['cufflink'], 'def': 'jewelry consisting of linked buttons used to fasten the cuffs of a shirt', 'name': 'cufflink'}, {'frequency': 'f', 'id': 347, 'synset': 'cup.n.01', 'synonyms': ['cup'], 'def': 'a small open container usually used for drinking; usually has a handle', 'name': 'cup'}, {'frequency': 'c', 'id': 348, 'synset': 'cup.n.08', 'synonyms': ['trophy_cup'], 'def': 'a metal vessel with handles that is awarded as a trophy to a competition winner', 'name': 'trophy_cup'}, {'frequency': 'c', 'id': 349, 'synset': 'cupcake.n.01', 'synonyms': ['cupcake'], 'def': 'small cake baked in a muffin tin', 'name': 'cupcake'}, {'frequency': 'r', 'id': 350, 'synset': 'curler.n.01', 'synonyms': ['hair_curler', 'hair_roller', 'hair_crimper'], 'def': 'a cylindrical tube around which the hair is wound to curl it', 'name': 'hair_curler'}, {'frequency': 'r', 'id': 351, 'synset': 'curling_iron.n.01', 'synonyms': ['curling_iron'], 'def': 'a cylindrical home appliance that heats hair that has been curled around it', 'name': 'curling_iron'}, {'frequency': 'f', 'id': 352, 'synset': 'curtain.n.01', 'synonyms': ['curtain', 'drapery'], 'def': 'hanging cloth used as a blind (especially for a window)', 'name': 'curtain'}, {'frequency': 'f', 'id': 353, 'synset': 'cushion.n.03', 'synonyms': ['cushion'], 'def': 'a soft bag filled with air or padding such as feathers or foam rubber', 'name': 'cushion'}, {'frequency': 'r', 'id': 354, 'synset': 'custard.n.01', 'synonyms': ['custard'], 'def': 'sweetened mixture of milk and eggs baked or boiled or frozen', 'name': 'custard'}, {'frequency': 'c', 'id': 355, 'synset': 'cutter.n.06', 'synonyms': ['cutting_tool'], 'def': 'a cutting implement; a tool for cutting', 'name': 'cutting_tool'}, {'frequency': 'r', 'id': 356, 'synset': 'cylinder.n.04', 'synonyms': ['cylinder'], 'def': 'a cylindrical container', 'name': 'cylinder'}, {'frequency': 'r', 'id': 357, 'synset': 'cymbal.n.01', 'synonyms': ['cymbal'], 'def': 'a percussion instrument consisting of a concave brass disk', 'name': 'cymbal'}, {'frequency': 'r', 'id': 358, 'synset': 'dachshund.n.01', 'synonyms': ['dachshund', 'dachsie', 'badger_dog'], 'def': 'small long-bodied short-legged breed of dog having a short sleek coat and long drooping ears', 'name': 'dachshund'}, {'frequency': 'r', 'id': 359, 'synset': 'dagger.n.01', 'synonyms': ['dagger'], 'def': 'a short knife with a pointed blade used for piercing or stabbing', 'name': 'dagger'}, {'frequency': 'r', 'id': 360, 'synset': 'dartboard.n.01', 'synonyms': ['dartboard'], 'def': 'a circular board of wood or cork used as the target in the game of darts', 'name': 'dartboard'}, {'frequency': 'r', 'id': 361, 'synset': 'date.n.08', 'synonyms': ['date_(fruit)'], 'def': 'sweet edible fruit of the date palm with a single long woody seed', 'name': 'date_(fruit)'}, {'frequency': 'f', 'id': 362, 'synset': 'deck_chair.n.01', 'synonyms': ['deck_chair', 'beach_chair'], 'def': 'a folding chair for use outdoors; a wooden frame supports a length of canvas', 'name': 'deck_chair'}, {'frequency': 'c', 'id': 363, 'synset': 'deer.n.01', 'synonyms': ['deer', 'cervid'], 'def': "distinguished from Bovidae by the male's having solid deciduous antlers", 'name': 'deer'}, {'frequency': 'c', 'id': 364, 'synset': 'dental_floss.n.01', 'synonyms': ['dental_floss', 'floss'], 'def': 'a soft thread for cleaning the spaces between the teeth', 'name': 'dental_floss'}, {'frequency': 'f', 'id': 365, 'synset': 'desk.n.01', 'synonyms': ['desk'], 'def': 'a piece of furniture with a writing surface and usually drawers or other compartments', 'name': 'desk'}, {'frequency': 'r', 'id': 366, 'synset': 'detergent.n.01', 'synonyms': ['detergent'], 'def': 'a surface-active chemical widely used in industry and laundering', 'name': 'detergent'}, {'frequency': 'c', 'id': 367, 'synset': 'diaper.n.01', 'synonyms': ['diaper'], 'def': 'garment consisting of a folded cloth drawn up between the legs and fastened at the waist', 'name': 'diaper'}, {'frequency': 'r', 'id': 368, 'synset': 'diary.n.01', 'synonyms': ['diary', 'journal'], 'def': 'a daily written record of (usually personal) experiences and observations', 'name': 'diary'}, {'frequency': 'r', 'id': 369, 'synset': 'die.n.01', 'synonyms': ['die', 'dice'], 'def': 'a small cube with 1 to 6 spots on the six faces; used in gambling', 'name': 'die'}, {'frequency': 'r', 'id': 370, 'synset': 'dinghy.n.01', 'synonyms': ['dinghy', 'dory', 'rowboat'], 'def': 'a small boat of shallow draft with seats and oars with which it is propelled', 'name': 'dinghy'}, {'frequency': 'f', 'id': 371, 'synset': 'dining_table.n.01', 'synonyms': ['dining_table'], 'def': 'a table at which meals are served', 'name': 'dining_table'}, {'frequency': 'r', 'id': 372, 'synset': 'dinner_jacket.n.01', 'synonyms': ['tux', 'tuxedo'], 'def': 'semiformal evening dress for men', 'name': 'tux'}, {'frequency': 'c', 'id': 373, 'synset': 'dish.n.01', 'synonyms': ['dish'], 'def': 'a piece of dishware normally used as a container for holding or serving food', 'name': 'dish'}, {'frequency': 'c', 'id': 374, 'synset': 'dish.n.05', 'synonyms': ['dish_antenna'], 'def': 'directional antenna consisting of a parabolic reflector', 'name': 'dish_antenna'}, {'frequency': 'c', 'id': 375, 'synset': 'dishrag.n.01', 'synonyms': ['dishrag', 'dishcloth'], 'def': 'a cloth for washing dishes', 'name': 'dishrag'}, {'frequency': 'c', 'id': 376, 'synset': 'dishtowel.n.01', 'synonyms': ['dishtowel', 'tea_towel'], 'def': 'a towel for drying dishes', 'name': 'dishtowel'}, {'frequency': 'f', 'id': 377, 'synset': 'dishwasher.n.01', 'synonyms': ['dishwasher', 'dishwashing_machine'], 'def': 'a machine for washing dishes', 'name': 'dishwasher'}, {'frequency': 'r', 'id': 378, 'synset': 'dishwasher_detergent.n.01', 'synonyms': ['dishwasher_detergent', 'dishwashing_detergent', 'dishwashing_liquid'], 'def': 'a low-sudsing detergent designed for use in dishwashers', 'name': 'dishwasher_detergent'}, {'frequency': 'r', 'id': 379, 'synset': 'diskette.n.01', 'synonyms': ['diskette', 'floppy', 'floppy_disk'], 'def': 'a small plastic magnetic disk enclosed in a stiff envelope used to store data', 'name': 'diskette'}, {'frequency': 'c', 'id': 380, 'synset': 'dispenser.n.01', 'synonyms': ['dispenser'], 'def': 'a container so designed that the contents can be used in prescribed amounts', 'name': 'dispenser'}, {'frequency': 'c', 'id': 381, 'synset': 'dixie_cup.n.01', 'synonyms': ['Dixie_cup', 'paper_cup'], 'def': 'a disposable cup made of paper; for holding drinks', 'name': 'Dixie_cup'}, {'frequency': 'f', 'id': 382, 'synset': 'dog.n.01', 'synonyms': ['dog'], 'def': 'a common domesticated dog', 'name': 'dog'}, {'frequency': 'f', 'id': 383, 'synset': 'dog_collar.n.01', 'synonyms': ['dog_collar'], 'def': 'a collar for a dog', 'name': 'dog_collar'}, {'frequency': 'c', 'id': 384, 'synset': 'doll.n.01', 'synonyms': ['doll'], 'def': 'a toy replica of a HUMAN (NOT AN ANIMAL)', 'name': 'doll'}, {'frequency': 'r', 'id': 385, 'synset': 'dollar.n.02', 'synonyms': ['dollar', 'dollar_bill', 'one_dollar_bill'], 'def': 'a piece of paper money worth one dollar', 'name': 'dollar'}, {'frequency': 'r', 'id': 386, 'synset': 'dolphin.n.02', 'synonyms': ['dolphin'], 'def': 'any of various small toothed whales with a beaklike snout; larger than porpoises', 'name': 'dolphin'}, {'frequency': 'c', 'id': 387, 'synset': 'domestic_ass.n.01', 'synonyms': ['domestic_ass', 'donkey'], 'def': 'domestic beast of burden descended from the African wild ass; patient but stubborn', 'name': 'domestic_ass'}, {'frequency': 'r', 'id': 388, 'synset': 'domino.n.03', 'synonyms': ['eye_mask'], 'def': 'a mask covering the upper part of the face but with holes for the eyes', 'name': 'eye_mask'}, {'frequency': 'r', 'id': 389, 'synset': 'doorbell.n.01', 'synonyms': ['doorbell', 'buzzer'], 'def': 'a button at an outer door that gives a ringing or buzzing signal when pushed', 'name': 'doorbell'}, {'frequency': 'f', 'id': 390, 'synset': 'doorknob.n.01', 'synonyms': ['doorknob', 'doorhandle'], 'def': "a knob used to open a door (often called `doorhandle' in Great Britain)", 'name': 'doorknob'}, {'frequency': 'c', 'id': 391, 'synset': 'doormat.n.02', 'synonyms': ['doormat', 'welcome_mat'], 'def': 'a mat placed outside an exterior door for wiping the shoes before entering', 'name': 'doormat'}, {'frequency': 'f', 'id': 392, 'synset': 'doughnut.n.02', 'synonyms': ['doughnut', 'donut'], 'def': 'a small ring-shaped friedcake', 'name': 'doughnut'}, {'frequency': 'r', 'id': 393, 'synset': 'dove.n.01', 'synonyms': ['dove'], 'def': 'any of numerous small pigeons', 'name': 'dove'}, {'frequency': 'r', 'id': 394, 'synset': 'dragonfly.n.01', 'synonyms': ['dragonfly'], 'def': 'slender-bodied non-stinging insect having iridescent wings that are outspread at rest', 'name': 'dragonfly'}, {'frequency': 'f', 'id': 395, 'synset': 'drawer.n.01', 'synonyms': ['drawer'], 'def': 'a boxlike container in a piece of furniture; made so as to slide in and out', 'name': 'drawer'}, {'frequency': 'c', 'id': 396, 'synset': 'drawers.n.01', 'synonyms': ['underdrawers', 'boxers', 'boxershorts'], 'def': 'underpants worn by men', 'name': 'underdrawers'}, {'frequency': 'f', 'id': 397, 'synset': 'dress.n.01', 'synonyms': ['dress', 'frock'], 'def': 'a one-piece garment for a woman; has skirt and bodice', 'name': 'dress'}, {'frequency': 'c', 'id': 398, 'synset': 'dress_hat.n.01', 'synonyms': ['dress_hat', 'high_hat', 'opera_hat', 'silk_hat', 'top_hat'], 'def': "a man's hat with a tall crown; usually covered with silk or with beaver fur", 'name': 'dress_hat'}, {'frequency': 'c', 'id': 399, 'synset': 'dress_suit.n.01', 'synonyms': ['dress_suit'], 'def': 'formalwear consisting of full evening dress for men', 'name': 'dress_suit'}, {'frequency': 'c', 'id': 400, 'synset': 'dresser.n.05', 'synonyms': ['dresser'], 'def': 'a cabinet with shelves', 'name': 'dresser'}, {'frequency': 'c', 'id': 401, 'synset': 'drill.n.01', 'synonyms': ['drill'], 'def': 'a tool with a sharp rotating point for making holes in hard materials', 'name': 'drill'}, {'frequency': 'r', 'id': 402, 'synset': 'drinking_fountain.n.01', 'synonyms': ['drinking_fountain'], 'def': 'a public fountain to provide a jet of drinking water', 'name': 'drinking_fountain'}, {'frequency': 'r', 'id': 403, 'synset': 'drone.n.04', 'synonyms': ['drone'], 'def': 'an aircraft without a pilot that is operated by remote control', 'name': 'drone'}, {'frequency': 'r', 'id': 404, 'synset': 'dropper.n.01', 'synonyms': ['dropper', 'eye_dropper'], 'def': 'pipet consisting of a small tube with a vacuum bulb at one end for drawing liquid in and releasing it a drop at a time', 'name': 'dropper'}, {'frequency': 'c', 'id': 405, 'synset': 'drum.n.01', 'synonyms': ['drum_(musical_instrument)'], 'def': 'a musical percussion instrument; usually consists of a hollow cylinder with a membrane stretched across each end', 'name': 'drum_(musical_instrument)'}, {'frequency': 'r', 'id': 406, 'synset': 'drumstick.n.02', 'synonyms': ['drumstick'], 'def': 'a stick used for playing a drum', 'name': 'drumstick'}, {'frequency': 'f', 'id': 407, 'synset': 'duck.n.01', 'synonyms': ['duck'], 'def': 'small web-footed broad-billed swimming bird', 'name': 'duck'}, {'frequency': 'r', 'id': 408, 'synset': 'duckling.n.02', 'synonyms': ['duckling'], 'def': 'young duck', 'name': 'duckling'}, {'frequency': 'c', 'id': 409, 'synset': 'duct_tape.n.01', 'synonyms': ['duct_tape'], 'def': 'a wide silvery adhesive tape', 'name': 'duct_tape'}, {'frequency': 'f', 'id': 410, 'synset': 'duffel_bag.n.01', 'synonyms': ['duffel_bag', 'duffle_bag', 'duffel', 'duffle'], 'def': 'a large cylindrical bag of heavy cloth', 'name': 'duffel_bag'}, {'frequency': 'r', 'id': 411, 'synset': 'dumbbell.n.01', 'synonyms': ['dumbbell'], 'def': 'an exercising weight with two ball-like ends connected by a short handle', 'name': 'dumbbell'}, {'frequency': 'c', 'id': 412, 'synset': 'dumpster.n.01', 'synonyms': ['dumpster'], 'def': 'a container designed to receive and transport and dump waste', 'name': 'dumpster'}, {'frequency': 'r', 'id': 413, 'synset': 'dustpan.n.02', 'synonyms': ['dustpan'], 'def': 'a short-handled receptacle into which dust can be swept', 'name': 'dustpan'}, {'frequency': 'r', 'id': 414, 'synset': 'dutch_oven.n.02', 'synonyms': ['Dutch_oven'], 'def': 'iron or earthenware cooking pot; used for stews', 'name': 'Dutch_oven'}, {'frequency': 'c', 'id': 415, 'synset': 'eagle.n.01', 'synonyms': ['eagle'], 'def': 'large birds of prey noted for their broad wings and strong soaring flight', 'name': 'eagle'}, {'frequency': 'f', 'id': 416, 'synset': 'earphone.n.01', 'synonyms': ['earphone', 'earpiece', 'headphone'], 'def': 'device for listening to audio that is held over or inserted into the ear', 'name': 'earphone'}, {'frequency': 'r', 'id': 417, 'synset': 'earplug.n.01', 'synonyms': ['earplug'], 'def': 'a soft plug that is inserted into the ear canal to block sound', 'name': 'earplug'}, {'frequency': 'f', 'id': 418, 'synset': 'earring.n.01', 'synonyms': ['earring'], 'def': 'jewelry to ornament the ear', 'name': 'earring'}, {'frequency': 'c', 'id': 419, 'synset': 'easel.n.01', 'synonyms': ['easel'], 'def': "an upright tripod for displaying something (usually an artist's canvas)", 'name': 'easel'}, {'frequency': 'r', 'id': 420, 'synset': 'eclair.n.01', 'synonyms': ['eclair'], 'def': 'oblong cream puff', 'name': 'eclair'}, {'frequency': 'r', 'id': 421, 'synset': 'eel.n.01', 'synonyms': ['eel'], 'def': 'an elongate fish with fatty flesh', 'name': 'eel'}, {'frequency': 'f', 'id': 422, 'synset': 'egg.n.02', 'synonyms': ['egg', 'eggs'], 'def': 'oval reproductive body of a fowl (especially a hen) used as food', 'name': 'egg'}, {'frequency': 'r', 'id': 423, 'synset': 'egg_roll.n.01', 'synonyms': ['egg_roll', 'spring_roll'], 'def': 'minced vegetables and meat wrapped in a pancake and fried', 'name': 'egg_roll'}, {'frequency': 'c', 'id': 424, 'synset': 'egg_yolk.n.01', 'synonyms': ['egg_yolk', 'yolk_(egg)'], 'def': 'the yellow spherical part of an egg', 'name': 'egg_yolk'}, {'frequency': 'c', 'id': 425, 'synset': 'eggbeater.n.02', 'synonyms': ['eggbeater', 'eggwhisk'], 'def': 'a mixer for beating eggs or whipping cream', 'name': 'eggbeater'}, {'frequency': 'c', 'id': 426, 'synset': 'eggplant.n.01', 'synonyms': ['eggplant', 'aubergine'], 'def': 'egg-shaped vegetable having a shiny skin typically dark purple', 'name': 'eggplant'}, {'frequency': 'r', 'id': 427, 'synset': 'electric_chair.n.01', 'synonyms': ['electric_chair'], 'def': 'a chair-shaped instrument of execution by electrocution', 'name': 'electric_chair'}, {'frequency': 'f', 'id': 428, 'synset': 'electric_refrigerator.n.01', 'synonyms': ['refrigerator'], 'def': 'a refrigerator in which the coolant is pumped around by an electric motor', 'name': 'refrigerator'}, {'frequency': 'f', 'id': 429, 'synset': 'elephant.n.01', 'synonyms': ['elephant'], 'def': 'a common elephant', 'name': 'elephant'}, {'frequency': 'r', 'id': 430, 'synset': 'elk.n.01', 'synonyms': ['elk', 'moose'], 'def': 'large northern deer with enormous flattened antlers in the male', 'name': 'elk'}, {'frequency': 'c', 'id': 431, 'synset': 'envelope.n.01', 'synonyms': ['envelope'], 'def': 'a flat (usually rectangular) container for a letter, thin package, etc.', 'name': 'envelope'}, {'frequency': 'c', 'id': 432, 'synset': 'eraser.n.01', 'synonyms': ['eraser'], 'def': 'an implement used to erase something', 'name': 'eraser'}, {'frequency': 'r', 'id': 433, 'synset': 'escargot.n.01', 'synonyms': ['escargot'], 'def': 'edible snail usually served in the shell with a sauce of melted butter and garlic', 'name': 'escargot'}, {'frequency': 'r', 'id': 434, 'synset': 'eyepatch.n.01', 'synonyms': ['eyepatch'], 'def': 'a protective cloth covering for an injured eye', 'name': 'eyepatch'}, {'frequency': 'r', 'id': 435, 'synset': 'falcon.n.01', 'synonyms': ['falcon'], 'def': 'birds of prey having long pointed powerful wings adapted for swift flight', 'name': 'falcon'}, {'frequency': 'f', 'id': 436, 'synset': 'fan.n.01', 'synonyms': ['fan'], 'def': 'a device for creating a current of air by movement of a surface or surfaces', 'name': 'fan'}, {'frequency': 'f', 'id': 437, 'synset': 'faucet.n.01', 'synonyms': ['faucet', 'spigot', 'tap'], 'def': 'a regulator for controlling the flow of a liquid from a reservoir', 'name': 'faucet'}, {'frequency': 'r', 'id': 438, 'synset': 'fedora.n.01', 'synonyms': ['fedora'], 'def': 'a hat made of felt with a creased crown', 'name': 'fedora'}, {'frequency': 'r', 'id': 439, 'synset': 'ferret.n.02', 'synonyms': ['ferret'], 'def': 'domesticated albino variety of the European polecat bred for hunting rats and rabbits', 'name': 'ferret'}, {'frequency': 'c', 'id': 440, 'synset': 'ferris_wheel.n.01', 'synonyms': ['Ferris_wheel'], 'def': 'a large wheel with suspended seats that remain upright as the wheel rotates', 'name': 'Ferris_wheel'}, {'frequency': 'r', 'id': 441, 'synset': 'ferry.n.01', 'synonyms': ['ferry', 'ferryboat'], 'def': 'a boat that transports people or vehicles across a body of water and operates on a regular schedule', 'name': 'ferry'}, {'frequency': 'r', 'id': 442, 'synset': 'fig.n.04', 'synonyms': ['fig_(fruit)'], 'def': 'fleshy sweet pear-shaped yellowish or purple fruit eaten fresh or preserved or dried', 'name': 'fig_(fruit)'}, {'frequency': 'c', 'id': 443, 'synset': 'fighter.n.02', 'synonyms': ['fighter_jet', 'fighter_aircraft', 'attack_aircraft'], 'def': 'a high-speed military or naval airplane designed to destroy enemy targets', 'name': 'fighter_jet'}, {'frequency': 'f', 'id': 444, 'synset': 'figurine.n.01', 'synonyms': ['figurine'], 'def': 'a small carved or molded figure', 'name': 'figurine'}, {'frequency': 'c', 'id': 445, 'synset': 'file.n.03', 'synonyms': ['file_cabinet', 'filing_cabinet'], 'def': 'office furniture consisting of a container for keeping papers in order', 'name': 'file_cabinet'}, {'frequency': 'r', 'id': 446, 'synset': 'file.n.04', 'synonyms': ['file_(tool)'], 'def': 'a steel hand tool with small sharp teeth on some or all of its surfaces; used for smoothing wood or metal', 'name': 'file_(tool)'}, {'frequency': 'f', 'id': 447, 'synset': 'fire_alarm.n.02', 'synonyms': ['fire_alarm', 'smoke_alarm'], 'def': 'an alarm that is tripped off by fire or smoke', 'name': 'fire_alarm'}, {'frequency': 'c', 'id': 448, 'synset': 'fire_engine.n.01', 'synonyms': ['fire_engine', 'fire_truck'], 'def': 'large trucks that carry firefighters and equipment to the site of a fire', 'name': 'fire_engine'}, {'frequency': 'c', 'id': 449, 'synset': 'fire_extinguisher.n.01', 'synonyms': ['fire_extinguisher', 'extinguisher'], 'def': 'a manually operated device for extinguishing small fires', 'name': 'fire_extinguisher'}, {'frequency': 'c', 'id': 450, 'synset': 'fire_hose.n.01', 'synonyms': ['fire_hose'], 'def': 'a large hose that carries water from a fire hydrant to the site of the fire', 'name': 'fire_hose'}, {'frequency': 'f', 'id': 451, 'synset': 'fireplace.n.01', 'synonyms': ['fireplace'], 'def': 'an open recess in a wall at the base of a chimney where a fire can be built', 'name': 'fireplace'}, {'frequency': 'f', 'id': 452, 'synset': 'fireplug.n.01', 'synonyms': ['fireplug', 'fire_hydrant', 'hydrant'], 'def': 'an upright hydrant for drawing water to use in fighting a fire', 'name': 'fireplug'}, {'frequency': 'c', 'id': 453, 'synset': 'fish.n.01', 'synonyms': ['fish'], 'def': 'any of various mostly cold-blooded aquatic vertebrates usually having scales and breathing through gills', 'name': 'fish'}, {'frequency': 'r', 'id': 454, 'synset': 'fish.n.02', 'synonyms': ['fish_(food)'], 'def': 'the flesh of fish used as food', 'name': 'fish_(food)'}, {'frequency': 'r', 'id': 455, 'synset': 'fishbowl.n.02', 'synonyms': ['fishbowl', 'goldfish_bowl'], 'def': 'a transparent bowl in which small fish are kept', 'name': 'fishbowl'}, {'frequency': 'r', 'id': 456, 'synset': 'fishing_boat.n.01', 'synonyms': ['fishing_boat', 'fishing_vessel'], 'def': 'a vessel for fishing', 'name': 'fishing_boat'}, {'frequency': 'c', 'id': 457, 'synset': 'fishing_rod.n.01', 'synonyms': ['fishing_rod', 'fishing_pole'], 'def': 'a rod that is used in fishing to extend the fishing line', 'name': 'fishing_rod'}, {'frequency': 'f', 'id': 458, 'synset': 'flag.n.01', 'synonyms': ['flag'], 'def': 'emblem usually consisting of a rectangular piece of cloth of distinctive design (do not include pole)', 'name': 'flag'}, {'frequency': 'f', 'id': 459, 'synset': 'flagpole.n.02', 'synonyms': ['flagpole', 'flagstaff'], 'def': 'a tall staff or pole on which a flag is raised', 'name': 'flagpole'}, {'frequency': 'c', 'id': 460, 'synset': 'flamingo.n.01', 'synonyms': ['flamingo'], 'def': 'large pink web-footed bird with down-bent bill', 'name': 'flamingo'}, {'frequency': 'c', 'id': 461, 'synset': 'flannel.n.01', 'synonyms': ['flannel'], 'def': 'a soft light woolen fabric; used for clothing', 'name': 'flannel'}, {'frequency': 'r', 'id': 462, 'synset': 'flash.n.10', 'synonyms': ['flash', 'flashbulb'], 'def': 'a lamp for providing momentary light to take a photograph', 'name': 'flash'}, {'frequency': 'c', 'id': 463, 'synset': 'flashlight.n.01', 'synonyms': ['flashlight', 'torch'], 'def': 'a small portable battery-powered electric lamp', 'name': 'flashlight'}, {'frequency': 'r', 'id': 464, 'synset': 'fleece.n.03', 'synonyms': ['fleece'], 'def': 'a soft bulky fabric with deep pile; used chiefly for clothing', 'name': 'fleece'}, {'frequency': 'f', 'id': 465, 'synset': 'flip-flop.n.02', 'synonyms': ['flip-flop_(sandal)'], 'def': 'a backless sandal held to the foot by a thong between two toes', 'name': 'flip-flop_(sandal)'}, {'frequency': 'c', 'id': 466, 'synset': 'flipper.n.01', 'synonyms': ['flipper_(footwear)', 'fin_(footwear)'], 'def': 'a shoe to aid a person in swimming', 'name': 'flipper_(footwear)'}, {'frequency': 'f', 'id': 467, 'synset': 'flower_arrangement.n.01', 'synonyms': ['flower_arrangement', 'floral_arrangement'], 'def': 'a decorative arrangement of flowers', 'name': 'flower_arrangement'}, {'frequency': 'c', 'id': 468, 'synset': 'flute.n.02', 'synonyms': ['flute_glass', 'champagne_flute'], 'def': 'a tall narrow wineglass', 'name': 'flute_glass'}, {'frequency': 'r', 'id': 469, 'synset': 'foal.n.01', 'synonyms': ['foal'], 'def': 'a young horse', 'name': 'foal'}, {'frequency': 'c', 'id': 470, 'synset': 'folding_chair.n.01', 'synonyms': ['folding_chair'], 'def': 'a chair that can be folded flat for storage', 'name': 'folding_chair'}, {'frequency': 'c', 'id': 471, 'synset': 'food_processor.n.01', 'synonyms': ['food_processor'], 'def': 'a kitchen appliance for shredding, blending, chopping, or slicing food', 'name': 'food_processor'}, {'frequency': 'c', 'id': 472, 'synset': 'football.n.02', 'synonyms': ['football_(American)'], 'def': 'the inflated oblong ball used in playing American football', 'name': 'football_(American)'}, {'frequency': 'r', 'id': 473, 'synset': 'football_helmet.n.01', 'synonyms': ['football_helmet'], 'def': 'a padded helmet with a face mask to protect the head of football players', 'name': 'football_helmet'}, {'frequency': 'c', 'id': 474, 'synset': 'footstool.n.01', 'synonyms': ['footstool', 'footrest'], 'def': 'a low seat or a stool to rest the feet of a seated person', 'name': 'footstool'}, {'frequency': 'f', 'id': 475, 'synset': 'fork.n.01', 'synonyms': ['fork'], 'def': 'cutlery used for serving and eating food', 'name': 'fork'}, {'frequency': 'r', 'id': 476, 'synset': 'forklift.n.01', 'synonyms': ['forklift'], 'def': 'an industrial vehicle with a power operated fork in front that can be inserted under loads to lift and move them', 'name': 'forklift'}, {'frequency': 'r', 'id': 477, 'synset': 'freight_car.n.01', 'synonyms': ['freight_car'], 'def': 'a railway car that carries freight', 'name': 'freight_car'}, {'frequency': 'r', 'id': 478, 'synset': 'french_toast.n.01', 'synonyms': ['French_toast'], 'def': 'bread slice dipped in egg and milk and fried', 'name': 'French_toast'}, {'frequency': 'c', 'id': 479, 'synset': 'freshener.n.01', 'synonyms': ['freshener', 'air_freshener'], 'def': 'anything that freshens', 'name': 'freshener'}, {'frequency': 'f', 'id': 480, 'synset': 'frisbee.n.01', 'synonyms': ['frisbee'], 'def': 'a light, plastic disk propelled with a flip of the wrist for recreation or competition', 'name': 'frisbee'}, {'frequency': 'c', 'id': 481, 'synset': 'frog.n.01', 'synonyms': ['frog', 'toad', 'toad_frog'], 'def': 'a tailless stout-bodied amphibians with long hind limbs for leaping', 'name': 'frog'}, {'frequency': 'c', 'id': 482, 'synset': 'fruit_juice.n.01', 'synonyms': ['fruit_juice'], 'def': 'drink produced by squeezing or crushing fruit', 'name': 'fruit_juice'}, {'frequency': 'r', 'id': 483, 'synset': 'fruit_salad.n.01', 'synonyms': ['fruit_salad'], 'def': 'salad composed of fruits', 'name': 'fruit_salad'}, {'frequency': 'c', 'id': 484, 'synset': 'frying_pan.n.01', 'synonyms': ['frying_pan', 'frypan', 'skillet'], 'def': 'a pan used for frying foods', 'name': 'frying_pan'}, {'frequency': 'r', 'id': 485, 'synset': 'fudge.n.01', 'synonyms': ['fudge'], 'def': 'soft creamy candy', 'name': 'fudge'}, {'frequency': 'r', 'id': 486, 'synset': 'funnel.n.02', 'synonyms': ['funnel'], 'def': 'a cone-shaped utensil used to channel a substance into a container with a small mouth', 'name': 'funnel'}, {'frequency': 'c', 'id': 487, 'synset': 'futon.n.01', 'synonyms': ['futon'], 'def': 'a pad that is used for sleeping on the floor or on a raised frame', 'name': 'futon'}, {'frequency': 'r', 'id': 488, 'synset': 'gag.n.02', 'synonyms': ['gag', 'muzzle'], 'def': "restraint put into a person's mouth to prevent speaking or shouting", 'name': 'gag'}, {'frequency': 'r', 'id': 489, 'synset': 'garbage.n.03', 'synonyms': ['garbage'], 'def': 'a receptacle where waste can be discarded', 'name': 'garbage'}, {'frequency': 'c', 'id': 490, 'synset': 'garbage_truck.n.01', 'synonyms': ['garbage_truck'], 'def': 'a truck for collecting domestic refuse', 'name': 'garbage_truck'}, {'frequency': 'c', 'id': 491, 'synset': 'garden_hose.n.01', 'synonyms': ['garden_hose'], 'def': 'a hose used for watering a lawn or garden', 'name': 'garden_hose'}, {'frequency': 'c', 'id': 492, 'synset': 'gargle.n.01', 'synonyms': ['gargle', 'mouthwash'], 'def': 'a medicated solution used for gargling and rinsing the mouth', 'name': 'gargle'}, {'frequency': 'r', 'id': 493, 'synset': 'gargoyle.n.02', 'synonyms': ['gargoyle'], 'def': 'an ornament consisting of a grotesquely carved figure of a person or animal', 'name': 'gargoyle'}, {'frequency': 'c', 'id': 494, 'synset': 'garlic.n.02', 'synonyms': ['garlic', 'ail'], 'def': 'aromatic bulb used as seasoning', 'name': 'garlic'}, {'frequency': 'r', 'id': 495, 'synset': 'gasmask.n.01', 'synonyms': ['gasmask', 'respirator', 'gas_helmet'], 'def': 'a protective face mask with a filter', 'name': 'gasmask'}, {'frequency': 'r', 'id': 496, 'synset': 'gazelle.n.01', 'synonyms': ['gazelle'], 'def': 'small swift graceful antelope of Africa and Asia having lustrous eyes', 'name': 'gazelle'}, {'frequency': 'c', 'id': 497, 'synset': 'gelatin.n.02', 'synonyms': ['gelatin', 'jelly'], 'def': 'an edible jelly made with gelatin and used as a dessert or salad base or a coating for foods', 'name': 'gelatin'}, {'frequency': 'r', 'id': 498, 'synset': 'gem.n.02', 'synonyms': ['gemstone'], 'def': 'a crystalline rock that can be cut and polished for jewelry', 'name': 'gemstone'}, {'frequency': 'c', 'id': 499, 'synset': 'giant_panda.n.01', 'synonyms': ['giant_panda', 'panda', 'panda_bear'], 'def': 'large black-and-white herbivorous mammal of bamboo forests of China and Tibet', 'name': 'giant_panda'}, {'frequency': 'c', 'id': 500, 'synset': 'gift_wrap.n.01', 'synonyms': ['gift_wrap'], 'def': 'attractive wrapping paper suitable for wrapping gifts', 'name': 'gift_wrap'}, {'frequency': 'c', 'id': 501, 'synset': 'ginger.n.03', 'synonyms': ['ginger', 'gingerroot'], 'def': 'the root of the common ginger plant; used fresh as a seasoning', 'name': 'ginger'}, {'frequency': 'f', 'id': 502, 'synset': 'giraffe.n.01', 'synonyms': ['giraffe'], 'def': 'tall animal having a spotted coat and small horns and very long neck and legs', 'name': 'giraffe'}, {'frequency': 'c', 'id': 503, 'synset': 'girdle.n.02', 'synonyms': ['cincture', 'sash', 'waistband', 'waistcloth'], 'def': 'a band of material around the waist that strengthens a skirt or trousers', 'name': 'cincture'}, {'frequency': 'f', 'id': 504, 'synset': 'glass.n.02', 'synonyms': ['glass_(drink_container)', 'drinking_glass'], 'def': 'a container for holding liquids while drinking', 'name': 'glass_(drink_container)'}, {'frequency': 'c', 'id': 505, 'synset': 'globe.n.03', 'synonyms': ['globe'], 'def': 'a sphere on which a map (especially of the earth) is represented', 'name': 'globe'}, {'frequency': 'f', 'id': 506, 'synset': 'glove.n.02', 'synonyms': ['glove'], 'def': 'handwear covering the hand', 'name': 'glove'}, {'frequency': 'c', 'id': 507, 'synset': 'goat.n.01', 'synonyms': ['goat'], 'def': 'a common goat', 'name': 'goat'}, {'frequency': 'f', 'id': 508, 'synset': 'goggles.n.01', 'synonyms': ['goggles'], 'def': 'tight-fitting spectacles worn to protect the eyes', 'name': 'goggles'}, {'frequency': 'r', 'id': 509, 'synset': 'goldfish.n.01', 'synonyms': ['goldfish'], 'def': 'small golden or orange-red freshwater fishes used as pond or aquarium pets', 'name': 'goldfish'}, {'frequency': 'r', 'id': 510, 'synset': 'golf_club.n.02', 'synonyms': ['golf_club', 'golf-club'], 'def': 'golf equipment used by a golfer to hit a golf ball', 'name': 'golf_club'}, {'frequency': 'c', 'id': 511, 'synset': 'golfcart.n.01', 'synonyms': ['golfcart'], 'def': 'a small motor vehicle in which golfers can ride between shots', 'name': 'golfcart'}, {'frequency': 'r', 'id': 512, 'synset': 'gondola.n.02', 'synonyms': ['gondola_(boat)'], 'def': 'long narrow flat-bottomed boat propelled by sculling; traditionally used on canals of Venice', 'name': 'gondola_(boat)'}, {'frequency': 'c', 'id': 513, 'synset': 'goose.n.01', 'synonyms': ['goose'], 'def': 'loud, web-footed long-necked aquatic birds usually larger than ducks', 'name': 'goose'}, {'frequency': 'r', 'id': 514, 'synset': 'gorilla.n.01', 'synonyms': ['gorilla'], 'def': 'largest ape', 'name': 'gorilla'}, {'frequency': 'r', 'id': 515, 'synset': 'gourd.n.02', 'synonyms': ['gourd'], 'def': 'any of numerous inedible fruits with hard rinds', 'name': 'gourd'}, {'frequency': 'r', 'id': 516, 'synset': 'gown.n.04', 'synonyms': ['surgical_gown', 'scrubs_(surgical_clothing)'], 'def': 'protective garment worn by surgeons during operations', 'name': 'surgical_gown'}, {'frequency': 'f', 'id': 517, 'synset': 'grape.n.01', 'synonyms': ['grape'], 'def': 'any of various juicy fruit with green or purple skins; grow in clusters', 'name': 'grape'}, {'frequency': 'r', 'id': 518, 'synset': 'grasshopper.n.01', 'synonyms': ['grasshopper'], 'def': 'plant-eating insect with hind legs adapted for leaping', 'name': 'grasshopper'}, {'frequency': 'c', 'id': 519, 'synset': 'grater.n.01', 'synonyms': ['grater'], 'def': 'utensil with sharp perforations for shredding foods (as vegetables or cheese)', 'name': 'grater'}, {'frequency': 'c', 'id': 520, 'synset': 'gravestone.n.01', 'synonyms': ['gravestone', 'headstone', 'tombstone'], 'def': 'a stone that is used to mark a grave', 'name': 'gravestone'}, {'frequency': 'r', 'id': 521, 'synset': 'gravy_boat.n.01', 'synonyms': ['gravy_boat', 'gravy_holder'], 'def': 'a dish (often boat-shaped) for serving gravy or sauce', 'name': 'gravy_boat'}, {'frequency': 'c', 'id': 522, 'synset': 'green_bean.n.02', 'synonyms': ['green_bean'], 'def': 'a common bean plant cultivated for its slender green edible pods', 'name': 'green_bean'}, {'frequency': 'c', 'id': 523, 'synset': 'green_onion.n.01', 'synonyms': ['green_onion', 'spring_onion', 'scallion'], 'def': 'a young onion before the bulb has enlarged', 'name': 'green_onion'}, {'frequency': 'r', 'id': 524, 'synset': 'griddle.n.01', 'synonyms': ['griddle'], 'def': 'cooking utensil consisting of a flat heated surface on which food is cooked', 'name': 'griddle'}, {'frequency': 'r', 'id': 525, 'synset': 'grillroom.n.01', 'synonyms': ['grillroom', 'grill_(restaurant)'], 'def': 'a restaurant where food is cooked on a grill', 'name': 'grillroom'}, {'frequency': 'r', 'id': 526, 'synset': 'grinder.n.04', 'synonyms': ['grinder_(tool)'], 'def': 'a machine tool that polishes metal', 'name': 'grinder_(tool)'}, {'frequency': 'r', 'id': 527, 'synset': 'grits.n.01', 'synonyms': ['grits', 'hominy_grits'], 'def': 'coarsely ground corn boiled as a breakfast dish', 'name': 'grits'}, {'frequency': 'c', 'id': 528, 'synset': 'grizzly.n.01', 'synonyms': ['grizzly', 'grizzly_bear'], 'def': 'powerful brownish-yellow bear of the uplands of western North America', 'name': 'grizzly'}, {'frequency': 'c', 'id': 529, 'synset': 'grocery_bag.n.01', 'synonyms': ['grocery_bag'], 'def': "a sack for holding customer's groceries", 'name': 'grocery_bag'}, {'frequency': 'r', 'id': 530, 'synset': 'guacamole.n.01', 'synonyms': ['guacamole'], 'def': 'a dip made of mashed avocado mixed with chopped onions and other seasonings', 'name': 'guacamole'}, {'frequency': 'f', 'id': 531, 'synset': 'guitar.n.01', 'synonyms': ['guitar'], 'def': 'a stringed instrument usually having six strings; played by strumming or plucking', 'name': 'guitar'}, {'frequency': 'c', 'id': 532, 'synset': 'gull.n.02', 'synonyms': ['gull', 'seagull'], 'def': 'mostly white aquatic bird having long pointed wings and short legs', 'name': 'gull'}, {'frequency': 'c', 'id': 533, 'synset': 'gun.n.01', 'synonyms': ['gun'], 'def': 'a weapon that discharges a bullet at high velocity from a metal tube', 'name': 'gun'}, {'frequency': 'r', 'id': 534, 'synset': 'hair_spray.n.01', 'synonyms': ['hair_spray'], 'def': 'substance sprayed on the hair to hold it in place', 'name': 'hair_spray'}, {'frequency': 'c', 'id': 535, 'synset': 'hairbrush.n.01', 'synonyms': ['hairbrush'], 'def': "a brush used to groom a person's hair", 'name': 'hairbrush'}, {'frequency': 'c', 'id': 536, 'synset': 'hairnet.n.01', 'synonyms': ['hairnet'], 'def': 'a small net that someone wears over their hair to keep it in place', 'name': 'hairnet'}, {'frequency': 'c', 'id': 537, 'synset': 'hairpin.n.01', 'synonyms': ['hairpin'], 'def': "a double pronged pin used to hold women's hair in place", 'name': 'hairpin'}, {'frequency': 'f', 'id': 538, 'synset': 'ham.n.01', 'synonyms': ['ham', 'jambon', 'gammon'], 'def': 'meat cut from the thigh of a hog (usually smoked)', 'name': 'ham'}, {'frequency': 'c', 'id': 539, 'synset': 'hamburger.n.01', 'synonyms': ['hamburger', 'beefburger', 'burger'], 'def': 'a sandwich consisting of a patty of minced beef served on a bun', 'name': 'hamburger'}, {'frequency': 'c', 'id': 540, 'synset': 'hammer.n.02', 'synonyms': ['hammer'], 'def': 'a hand tool with a heavy head and a handle; used to deliver an impulsive force by striking', 'name': 'hammer'}, {'frequency': 'r', 'id': 541, 'synset': 'hammock.n.02', 'synonyms': ['hammock'], 'def': 'a hanging bed of canvas or rope netting (usually suspended between two trees)', 'name': 'hammock'}, {'frequency': 'r', 'id': 542, 'synset': 'hamper.n.02', 'synonyms': ['hamper'], 'def': 'a basket usually with a cover', 'name': 'hamper'}, {'frequency': 'r', 'id': 543, 'synset': 'hamster.n.01', 'synonyms': ['hamster'], 'def': 'short-tailed burrowing rodent with large cheek pouches', 'name': 'hamster'}, {'frequency': 'c', 'id': 544, 'synset': 'hand_blower.n.01', 'synonyms': ['hair_dryer'], 'def': 'a hand-held electric blower that can blow warm air onto the hair', 'name': 'hair_dryer'}, {'frequency': 'r', 'id': 545, 'synset': 'hand_glass.n.01', 'synonyms': ['hand_glass', 'hand_mirror'], 'def': 'a mirror intended to be held in the hand', 'name': 'hand_glass'}, {'frequency': 'f', 'id': 546, 'synset': 'hand_towel.n.01', 'synonyms': ['hand_towel', 'face_towel'], 'def': 'a small towel used to dry the hands or face', 'name': 'hand_towel'}, {'frequency': 'c', 'id': 547, 'synset': 'handcart.n.01', 'synonyms': ['handcart', 'pushcart', 'hand_truck'], 'def': 'wheeled vehicle that can be pushed by a person', 'name': 'handcart'}, {'frequency': 'r', 'id': 548, 'synset': 'handcuff.n.01', 'synonyms': ['handcuff'], 'def': 'shackle that consists of a metal loop that can be locked around the wrist', 'name': 'handcuff'}, {'frequency': 'c', 'id': 549, 'synset': 'handkerchief.n.01', 'synonyms': ['handkerchief'], 'def': 'a square piece of cloth used for wiping the eyes or nose or as a costume accessory', 'name': 'handkerchief'}, {'frequency': 'f', 'id': 550, 'synset': 'handle.n.01', 'synonyms': ['handle', 'grip', 'handgrip'], 'def': 'the appendage to an object that is designed to be held in order to use or move it', 'name': 'handle'}, {'frequency': 'r', 'id': 551, 'synset': 'handsaw.n.01', 'synonyms': ['handsaw', "carpenter's_saw"], 'def': 'a saw used with one hand for cutting wood', 'name': 'handsaw'}, {'frequency': 'r', 'id': 552, 'synset': 'hardback.n.01', 'synonyms': ['hardback_book', 'hardcover_book'], 'def': 'a book with cardboard or cloth or leather covers', 'name': 'hardback_book'}, {'frequency': 'r', 'id': 553, 'synset': 'harmonium.n.01', 'synonyms': ['harmonium', 'organ_(musical_instrument)', 'reed_organ_(musical_instrument)'], 'def': 'a free-reed instrument in which air is forced through the reeds by bellows', 'name': 'harmonium'}, {'frequency': 'f', 'id': 554, 'synset': 'hat.n.01', 'synonyms': ['hat'], 'def': 'headwear that protects the head from bad weather, sun, or worn for fashion', 'name': 'hat'}, {'frequency': 'r', 'id': 555, 'synset': 'hatbox.n.01', 'synonyms': ['hatbox'], 'def': 'a round piece of luggage for carrying hats', 'name': 'hatbox'}, {'frequency': 'r', 'id': 556, 'synset': 'hatch.n.03', 'synonyms': ['hatch'], 'def': 'a movable barrier covering a hatchway', 'name': 'hatch'}, {'frequency': 'c', 'id': 557, 'synset': 'head_covering.n.01', 'synonyms': ['veil'], 'def': 'a garment that covers the head and face', 'name': 'veil'}, {'frequency': 'f', 'id': 558, 'synset': 'headband.n.01', 'synonyms': ['headband'], 'def': 'a band worn around or over the head', 'name': 'headband'}, {'frequency': 'f', 'id': 559, 'synset': 'headboard.n.01', 'synonyms': ['headboard'], 'def': 'a vertical board or panel forming the head of a bedstead', 'name': 'headboard'}, {'frequency': 'f', 'id': 560, 'synset': 'headlight.n.01', 'synonyms': ['headlight', 'headlamp'], 'def': 'a powerful light with reflector; attached to the front of an automobile or locomotive', 'name': 'headlight'}, {'frequency': 'c', 'id': 561, 'synset': 'headscarf.n.01', 'synonyms': ['headscarf'], 'def': 'a kerchief worn over the head and tied under the chin', 'name': 'headscarf'}, {'frequency': 'r', 'id': 562, 'synset': 'headset.n.01', 'synonyms': ['headset'], 'def': 'receiver consisting of a pair of headphones', 'name': 'headset'}, {'frequency': 'c', 'id': 563, 'synset': 'headstall.n.01', 'synonyms': ['headstall_(for_horses)', 'headpiece_(for_horses)'], 'def': "the band that is the part of a bridle that fits around a horse's head", 'name': 'headstall_(for_horses)'}, {'frequency': 'r', 'id': 564, 'synset': 'hearing_aid.n.02', 'synonyms': ['hearing_aid'], 'def': 'an acoustic device used to direct sound to the ear of a hearing-impaired person', 'name': 'hearing_aid'}, {'frequency': 'c', 'id': 565, 'synset': 'heart.n.02', 'synonyms': ['heart'], 'def': 'a muscular organ; its contractions move the blood through the body', 'name': 'heart'}, {'frequency': 'c', 'id': 566, 'synset': 'heater.n.01', 'synonyms': ['heater', 'warmer'], 'def': 'device that heats water or supplies warmth to a room', 'name': 'heater'}, {'frequency': 'c', 'id': 567, 'synset': 'helicopter.n.01', 'synonyms': ['helicopter'], 'def': 'an aircraft without wings that obtains its lift from the rotation of overhead blades', 'name': 'helicopter'}, {'frequency': 'f', 'id': 568, 'synset': 'helmet.n.02', 'synonyms': ['helmet'], 'def': 'a protective headgear made of hard material to resist blows', 'name': 'helmet'}, {'frequency': 'r', 'id': 569, 'synset': 'heron.n.02', 'synonyms': ['heron'], 'def': 'grey or white wading bird with long neck and long legs and (usually) long bill', 'name': 'heron'}, {'frequency': 'c', 'id': 570, 'synset': 'highchair.n.01', 'synonyms': ['highchair', 'feeding_chair'], 'def': 'a chair for feeding a very young child', 'name': 'highchair'}, {'frequency': 'f', 'id': 571, 'synset': 'hinge.n.01', 'synonyms': ['hinge'], 'def': 'a joint that holds two parts together so that one can swing relative to the other', 'name': 'hinge'}, {'frequency': 'r', 'id': 572, 'synset': 'hippopotamus.n.01', 'synonyms': ['hippopotamus'], 'def': 'massive thick-skinned animal living in or around rivers of tropical Africa', 'name': 'hippopotamus'}, {'frequency': 'r', 'id': 573, 'synset': 'hockey_stick.n.01', 'synonyms': ['hockey_stick'], 'def': 'sports implement consisting of a stick used by hockey players to move the puck', 'name': 'hockey_stick'}, {'frequency': 'c', 'id': 574, 'synset': 'hog.n.03', 'synonyms': ['hog', 'pig'], 'def': 'domestic swine', 'name': 'hog'}, {'frequency': 'f', 'id': 575, 'synset': 'home_plate.n.01', 'synonyms': ['home_plate_(baseball)', 'home_base_(baseball)'], 'def': '(baseball) a rubber slab where the batter stands; it must be touched by a base runner in order to score', 'name': 'home_plate_(baseball)'}, {'frequency': 'c', 'id': 576, 'synset': 'honey.n.01', 'synonyms': ['honey'], 'def': 'a sweet yellow liquid produced by bees', 'name': 'honey'}, {'frequency': 'f', 'id': 577, 'synset': 'hood.n.06', 'synonyms': ['fume_hood', 'exhaust_hood'], 'def': 'metal covering leading to a vent that exhausts smoke or fumes', 'name': 'fume_hood'}, {'frequency': 'f', 'id': 578, 'synset': 'hook.n.05', 'synonyms': ['hook'], 'def': 'a curved or bent implement for suspending or pulling something', 'name': 'hook'}, {'frequency': 'f', 'id': 579, 'synset': 'horse.n.01', 'synonyms': ['horse'], 'def': 'a common horse', 'name': 'horse'}, {'frequency': 'f', 'id': 580, 'synset': 'hose.n.03', 'synonyms': ['hose', 'hosepipe'], 'def': 'a flexible pipe for conveying a liquid or gas', 'name': 'hose'}, {'frequency': 'r', 'id': 581, 'synset': 'hot-air_balloon.n.01', 'synonyms': ['hot-air_balloon'], 'def': 'balloon for travel through the air in a basket suspended below a large bag of heated air', 'name': 'hot-air_balloon'}, {'frequency': 'r', 'id': 582, 'synset': 'hot_plate.n.01', 'synonyms': ['hotplate'], 'def': 'a portable electric appliance for heating or cooking or keeping food warm', 'name': 'hotplate'}, {'frequency': 'c', 'id': 583, 'synset': 'hot_sauce.n.01', 'synonyms': ['hot_sauce'], 'def': 'a pungent peppery sauce', 'name': 'hot_sauce'}, {'frequency': 'r', 'id': 584, 'synset': 'hourglass.n.01', 'synonyms': ['hourglass'], 'def': 'a sandglass timer that runs for sixty minutes', 'name': 'hourglass'}, {'frequency': 'r', 'id': 585, 'synset': 'houseboat.n.01', 'synonyms': ['houseboat'], 'def': 'a barge that is designed and equipped for use as a dwelling', 'name': 'houseboat'}, {'frequency': 'r', 'id': 586, 'synset': 'hummingbird.n.01', 'synonyms': ['hummingbird'], 'def': 'tiny American bird having brilliant iridescent plumage and long slender bills', 'name': 'hummingbird'}, {'frequency': 'r', 'id': 587, 'synset': 'hummus.n.01', 'synonyms': ['hummus', 'humus', 'hommos', 'hoummos', 'humous'], 'def': 'a thick spread made from mashed chickpeas', 'name': 'hummus'}, {'frequency': 'c', 'id': 588, 'synset': 'ice_bear.n.01', 'synonyms': ['polar_bear'], 'def': 'white bear of Arctic regions', 'name': 'polar_bear'}, {'frequency': 'c', 'id': 589, 'synset': 'ice_cream.n.01', 'synonyms': ['icecream'], 'def': 'frozen dessert containing cream and sugar and flavoring', 'name': 'icecream'}, {'frequency': 'r', 'id': 590, 'synset': 'ice_lolly.n.01', 'synonyms': ['popsicle'], 'def': 'ice cream or water ice on a small wooden stick', 'name': 'popsicle'}, {'frequency': 'c', 'id': 591, 'synset': 'ice_maker.n.01', 'synonyms': ['ice_maker'], 'def': 'an appliance included in some electric refrigerators for making ice cubes', 'name': 'ice_maker'}, {'frequency': 'r', 'id': 592, 'synset': 'ice_pack.n.01', 'synonyms': ['ice_pack', 'ice_bag'], 'def': 'a waterproof bag filled with ice: applied to the body (especially the head) to cool or reduce swelling', 'name': 'ice_pack'}, {'frequency': 'r', 'id': 593, 'synset': 'ice_skate.n.01', 'synonyms': ['ice_skate'], 'def': 'skate consisting of a boot with a steel blade fitted to the sole', 'name': 'ice_skate'}, {'frequency': 'r', 'id': 594, 'synset': 'ice_tea.n.01', 'synonyms': ['ice_tea', 'iced_tea'], 'def': 'strong tea served over ice', 'name': 'ice_tea'}, {'frequency': 'c', 'id': 595, 'synset': 'igniter.n.01', 'synonyms': ['igniter', 'ignitor', 'lighter'], 'def': 'a substance or device used to start a fire', 'name': 'igniter'}, {'frequency': 'r', 'id': 596, 'synset': 'incense.n.01', 'synonyms': ['incense'], 'def': 'a substance that produces a fragrant odor when burned', 'name': 'incense'}, {'frequency': 'r', 'id': 597, 'synset': 'inhaler.n.01', 'synonyms': ['inhaler', 'inhalator'], 'def': 'a dispenser that produces a chemical vapor to be inhaled through mouth or nose', 'name': 'inhaler'}, {'frequency': 'c', 'id': 598, 'synset': 'ipod.n.01', 'synonyms': ['iPod'], 'def': 'a pocket-sized device used to play music files', 'name': 'iPod'}, {'frequency': 'c', 'id': 599, 'synset': 'iron.n.04', 'synonyms': ['iron_(for_clothing)', 'smoothing_iron_(for_clothing)'], 'def': 'home appliance consisting of a flat metal base that is heated and used to smooth cloth', 'name': 'iron_(for_clothing)'}, {'frequency': 'r', 'id': 600, 'synset': 'ironing_board.n.01', 'synonyms': ['ironing_board'], 'def': 'narrow padded board on collapsible supports; used for ironing clothes', 'name': 'ironing_board'}, {'frequency': 'f', 'id': 601, 'synset': 'jacket.n.01', 'synonyms': ['jacket'], 'def': 'a waist-length coat', 'name': 'jacket'}, {'frequency': 'r', 'id': 602, 'synset': 'jam.n.01', 'synonyms': ['jam'], 'def': 'preserve of crushed fruit', 'name': 'jam'}, {'frequency': 'f', 'id': 603, 'synset': 'jean.n.01', 'synonyms': ['jean', 'blue_jean', 'denim'], 'def': '(usually plural) close-fitting trousers of heavy denim for manual work or casual wear', 'name': 'jean'}, {'frequency': 'c', 'id': 604, 'synset': 'jeep.n.01', 'synonyms': ['jeep', 'landrover'], 'def': 'a car suitable for traveling over rough terrain', 'name': 'jeep'}, {'frequency': 'r', 'id': 605, 'synset': 'jelly_bean.n.01', 'synonyms': ['jelly_bean', 'jelly_egg'], 'def': 'sugar-glazed jellied candy', 'name': 'jelly_bean'}, {'frequency': 'f', 'id': 606, 'synset': 'jersey.n.03', 'synonyms': ['jersey', 'T-shirt', 'tee_shirt'], 'def': 'a close-fitting pullover shirt', 'name': 'jersey'}, {'frequency': 'c', 'id': 607, 'synset': 'jet.n.01', 'synonyms': ['jet_plane', 'jet-propelled_plane'], 'def': 'an airplane powered by one or more jet engines', 'name': 'jet_plane'}, {'frequency': 'c', 'id': 608, 'synset': 'jewelry.n.01', 'synonyms': ['jewelry', 'jewellery'], 'def': 'an adornment (as a bracelet or ring or necklace) made of precious metals and set with gems (or imitation gems)', 'name': 'jewelry'}, {'frequency': 'r', 'id': 609, 'synset': 'joystick.n.02', 'synonyms': ['joystick'], 'def': 'a control device for computers consisting of a vertical handle that can move freely in two directions', 'name': 'joystick'}, {'frequency': 'r', 'id': 610, 'synset': 'jump_suit.n.01', 'synonyms': ['jumpsuit'], 'def': "one-piece garment fashioned after a parachutist's uniform", 'name': 'jumpsuit'}, {'frequency': 'c', 'id': 611, 'synset': 'kayak.n.01', 'synonyms': ['kayak'], 'def': 'a small canoe consisting of a light frame made watertight with animal skins', 'name': 'kayak'}, {'frequency': 'r', 'id': 612, 'synset': 'keg.n.02', 'synonyms': ['keg'], 'def': 'small cask or barrel', 'name': 'keg'}, {'frequency': 'r', 'id': 613, 'synset': 'kennel.n.01', 'synonyms': ['kennel', 'doghouse'], 'def': 'outbuilding that serves as a shelter for a dog', 'name': 'kennel'}, {'frequency': 'c', 'id': 614, 'synset': 'kettle.n.01', 'synonyms': ['kettle', 'boiler'], 'def': 'a metal pot for stewing or boiling; usually has a lid', 'name': 'kettle'}, {'frequency': 'f', 'id': 615, 'synset': 'key.n.01', 'synonyms': ['key'], 'def': 'metal instrument used to unlock a lock', 'name': 'key'}, {'frequency': 'r', 'id': 616, 'synset': 'keycard.n.01', 'synonyms': ['keycard'], 'def': 'a plastic card used to gain access typically to a door', 'name': 'keycard'}, {'frequency': 'r', 'id': 617, 'synset': 'kilt.n.01', 'synonyms': ['kilt'], 'def': 'a knee-length pleated tartan skirt worn by men as part of the traditional dress in the Highlands of northern Scotland', 'name': 'kilt'}, {'frequency': 'c', 'id': 618, 'synset': 'kimono.n.01', 'synonyms': ['kimono'], 'def': 'a loose robe; imitated from robes originally worn by Japanese', 'name': 'kimono'}, {'frequency': 'f', 'id': 619, 'synset': 'kitchen_sink.n.01', 'synonyms': ['kitchen_sink'], 'def': 'a sink in a kitchen', 'name': 'kitchen_sink'}, {'frequency': 'c', 'id': 620, 'synset': 'kitchen_table.n.01', 'synonyms': ['kitchen_table'], 'def': 'a table in the kitchen', 'name': 'kitchen_table'}, {'frequency': 'f', 'id': 621, 'synset': 'kite.n.03', 'synonyms': ['kite'], 'def': 'plaything consisting of a light frame covered with tissue paper; flown in wind at end of a string', 'name': 'kite'}, {'frequency': 'c', 'id': 622, 'synset': 'kitten.n.01', 'synonyms': ['kitten', 'kitty'], 'def': 'young domestic cat', 'name': 'kitten'}, {'frequency': 'c', 'id': 623, 'synset': 'kiwi.n.03', 'synonyms': ['kiwi_fruit'], 'def': 'fuzzy brown egg-shaped fruit with slightly tart green flesh', 'name': 'kiwi_fruit'}, {'frequency': 'f', 'id': 624, 'synset': 'knee_pad.n.01', 'synonyms': ['knee_pad'], 'def': 'protective garment consisting of a pad worn by football or baseball or hockey players', 'name': 'knee_pad'}, {'frequency': 'f', 'id': 625, 'synset': 'knife.n.01', 'synonyms': ['knife'], 'def': 'tool with a blade and point used as a cutting instrument', 'name': 'knife'}, {'frequency': 'r', 'id': 626, 'synset': 'knight.n.02', 'synonyms': ['knight_(chess_piece)', 'horse_(chess_piece)'], 'def': 'a chess game piece shaped to resemble the head of a horse', 'name': 'knight_(chess_piece)'}, {'frequency': 'r', 'id': 627, 'synset': 'knitting_needle.n.01', 'synonyms': ['knitting_needle'], 'def': 'needle consisting of a slender rod with pointed ends; usually used in pairs', 'name': 'knitting_needle'}, {'frequency': 'f', 'id': 628, 'synset': 'knob.n.02', 'synonyms': ['knob'], 'def': 'a round handle often found on a door', 'name': 'knob'}, {'frequency': 'r', 'id': 629, 'synset': 'knocker.n.05', 'synonyms': ['knocker_(on_a_door)', 'doorknocker'], 'def': 'a device (usually metal and ornamental) attached by a hinge to a door', 'name': 'knocker_(on_a_door)'}, {'frequency': 'r', 'id': 630, 'synset': 'koala.n.01', 'synonyms': ['koala', 'koala_bear'], 'def': 'sluggish tailless Australian marsupial with grey furry ears and coat', 'name': 'koala'}, {'frequency': 'r', 'id': 631, 'synset': 'lab_coat.n.01', 'synonyms': ['lab_coat', 'laboratory_coat'], 'def': 'a light coat worn to protect clothing from substances used while working in a laboratory', 'name': 'lab_coat'}, {'frequency': 'f', 'id': 632, 'synset': 'ladder.n.01', 'synonyms': ['ladder'], 'def': 'steps consisting of two parallel members connected by rungs', 'name': 'ladder'}, {'frequency': 'c', 'id': 633, 'synset': 'ladle.n.01', 'synonyms': ['ladle'], 'def': 'a spoon-shaped vessel with a long handle frequently used to transfer liquids', 'name': 'ladle'}, {'frequency': 'r', 'id': 634, 'synset': 'ladybug.n.01', 'synonyms': ['ladybug', 'ladybeetle', 'ladybird_beetle'], 'def': 'small round bright-colored and spotted beetle, typically red and black', 'name': 'ladybug'}, {'frequency': 'c', 'id': 635, 'synset': 'lamb.n.01', 'synonyms': ['lamb_(animal)'], 'def': 'young sheep', 'name': 'lamb_(animal)'}, {'frequency': 'r', 'id': 636, 'synset': 'lamb_chop.n.01', 'synonyms': ['lamb-chop', 'lambchop'], 'def': 'chop cut from a lamb', 'name': 'lamb-chop'}, {'frequency': 'f', 'id': 637, 'synset': 'lamp.n.02', 'synonyms': ['lamp'], 'def': 'a piece of furniture holding one or more electric light bulbs', 'name': 'lamp'}, {'frequency': 'f', 'id': 638, 'synset': 'lamppost.n.01', 'synonyms': ['lamppost'], 'def': 'a metal post supporting an outdoor lamp (such as a streetlight)', 'name': 'lamppost'}, {'frequency': 'f', 'id': 639, 'synset': 'lampshade.n.01', 'synonyms': ['lampshade'], 'def': 'a protective ornamental shade used to screen a light bulb from direct view', 'name': 'lampshade'}, {'frequency': 'c', 'id': 640, 'synset': 'lantern.n.01', 'synonyms': ['lantern'], 'def': 'light in a transparent protective case', 'name': 'lantern'}, {'frequency': 'f', 'id': 641, 'synset': 'lanyard.n.02', 'synonyms': ['lanyard', 'laniard'], 'def': 'a cord worn around the neck to hold a knife or whistle, etc.', 'name': 'lanyard'}, {'frequency': 'f', 'id': 642, 'synset': 'laptop.n.01', 'synonyms': ['laptop_computer', 'notebook_computer'], 'def': 'a portable computer small enough to use in your lap', 'name': 'laptop_computer'}, {'frequency': 'r', 'id': 643, 'synset': 'lasagna.n.01', 'synonyms': ['lasagna', 'lasagne'], 'def': 'baked dish of layers of lasagna pasta with sauce and cheese and meat or vegetables', 'name': 'lasagna'}, {'frequency': 'c', 'id': 644, 'synset': 'latch.n.02', 'synonyms': ['latch'], 'def': 'a bar that can be lowered or slid into a groove to fasten a door or gate', 'name': 'latch'}, {'frequency': 'r', 'id': 645, 'synset': 'lawn_mower.n.01', 'synonyms': ['lawn_mower'], 'def': 'garden tool for mowing grass on lawns', 'name': 'lawn_mower'}, {'frequency': 'r', 'id': 646, 'synset': 'leather.n.01', 'synonyms': ['leather'], 'def': 'an animal skin made smooth and flexible by removing the hair and then tanning', 'name': 'leather'}, {'frequency': 'c', 'id': 647, 'synset': 'legging.n.01', 'synonyms': ['legging_(clothing)', 'leging_(clothing)', 'leg_covering'], 'def': 'a garment covering the leg (usually extending from the knee to the ankle)', 'name': 'legging_(clothing)'}, {'frequency': 'c', 'id': 648, 'synset': 'lego.n.01', 'synonyms': ['Lego', 'Lego_set'], 'def': "a child's plastic construction set for making models from blocks", 'name': 'Lego'}, {'frequency': 'f', 'id': 649, 'synset': 'lemon.n.01', 'synonyms': ['lemon'], 'def': 'yellow oval fruit with juicy acidic flesh', 'name': 'lemon'}, {'frequency': 'r', 'id': 650, 'synset': 'lemonade.n.01', 'synonyms': ['lemonade'], 'def': 'sweetened beverage of diluted lemon juice', 'name': 'lemonade'}, {'frequency': 'f', 'id': 651, 'synset': 'lettuce.n.02', 'synonyms': ['lettuce'], 'def': 'leafy plant commonly eaten in salad or on sandwiches', 'name': 'lettuce'}, {'frequency': 'f', 'id': 652, 'synset': 'license_plate.n.01', 'synonyms': ['license_plate', 'numberplate'], 'def': "a plate mounted on the front and back of car and bearing the car's registration number", 'name': 'license_plate'}, {'frequency': 'f', 'id': 653, 'synset': 'life_buoy.n.01', 'synonyms': ['life_buoy', 'lifesaver', 'life_belt', 'life_ring'], 'def': 'a ring-shaped life preserver used to prevent drowning (NOT a life-jacket or vest)', 'name': 'life_buoy'}, {'frequency': 'f', 'id': 654, 'synset': 'life_jacket.n.01', 'synonyms': ['life_jacket', 'life_vest'], 'def': 'life preserver consisting of a sleeveless jacket of buoyant or inflatable design', 'name': 'life_jacket'}, {'frequency': 'f', 'id': 655, 'synset': 'light_bulb.n.01', 'synonyms': ['lightbulb'], 'def': 'glass bulb or tube shaped electric device that emits light (DO NOT MARK LAMPS AS A WHOLE)', 'name': 'lightbulb'}, {'frequency': 'r', 'id': 656, 'synset': 'lightning_rod.n.02', 'synonyms': ['lightning_rod', 'lightning_conductor'], 'def': 'a metallic conductor that is attached to a high point and leads to the ground', 'name': 'lightning_rod'}, {'frequency': 'c', 'id': 657, 'synset': 'lime.n.06', 'synonyms': ['lime'], 'def': 'the green acidic fruit of any of various lime trees', 'name': 'lime'}, {'frequency': 'r', 'id': 658, 'synset': 'limousine.n.01', 'synonyms': ['limousine'], 'def': 'long luxurious car; usually driven by a chauffeur', 'name': 'limousine'}, {'frequency': 'r', 'id': 659, 'synset': 'linen.n.02', 'synonyms': ['linen_paper'], 'def': 'a high-quality paper made of linen fibers or with a linen finish', 'name': 'linen_paper'}, {'frequency': 'c', 'id': 660, 'synset': 'lion.n.01', 'synonyms': ['lion'], 'def': 'large gregarious predatory cat of Africa and India', 'name': 'lion'}, {'frequency': 'c', 'id': 661, 'synset': 'lip_balm.n.01', 'synonyms': ['lip_balm'], 'def': 'a balm applied to the lips', 'name': 'lip_balm'}, {'frequency': 'c', 'id': 662, 'synset': 'lipstick.n.01', 'synonyms': ['lipstick', 'lip_rouge'], 'def': 'makeup that is used to color the lips', 'name': 'lipstick'}, {'frequency': 'r', 'id': 663, 'synset': 'liquor.n.01', 'synonyms': ['liquor', 'spirits', 'hard_liquor', 'liqueur', 'cordial'], 'def': 'an alcoholic beverage that is distilled rather than fermented', 'name': 'liquor'}, {'frequency': 'r', 'id': 664, 'synset': 'lizard.n.01', 'synonyms': ['lizard'], 'def': 'a reptile with usually two pairs of legs and a tapering tail', 'name': 'lizard'}, {'frequency': 'r', 'id': 665, 'synset': 'loafer.n.02', 'synonyms': ['Loafer_(type_of_shoe)'], 'def': 'a low leather step-in shoe', 'name': 'Loafer_(type_of_shoe)'}, {'frequency': 'f', 'id': 666, 'synset': 'log.n.01', 'synonyms': ['log'], 'def': 'a segment of the trunk of a tree when stripped of branches', 'name': 'log'}, {'frequency': 'c', 'id': 667, 'synset': 'lollipop.n.02', 'synonyms': ['lollipop'], 'def': 'hard candy on a stick', 'name': 'lollipop'}, {'frequency': 'c', 'id': 668, 'synset': 'lotion.n.01', 'synonyms': ['lotion'], 'def': 'any of various cosmetic preparations that are applied to the skin', 'name': 'lotion'}, {'frequency': 'f', 'id': 669, 'synset': 'loudspeaker.n.01', 'synonyms': ['speaker_(stero_equipment)'], 'def': 'electronic device that produces sound often as part of a stereo system', 'name': 'speaker_(stero_equipment)'}, {'frequency': 'c', 'id': 670, 'synset': 'love_seat.n.01', 'synonyms': ['loveseat'], 'def': 'small sofa that seats two people', 'name': 'loveseat'}, {'frequency': 'r', 'id': 671, 'synset': 'machine_gun.n.01', 'synonyms': ['machine_gun'], 'def': 'a rapidly firing automatic gun', 'name': 'machine_gun'}, {'frequency': 'f', 'id': 672, 'synset': 'magazine.n.02', 'synonyms': ['magazine'], 'def': 'a paperback periodic publication', 'name': 'magazine'}, {'frequency': 'f', 'id': 673, 'synset': 'magnet.n.01', 'synonyms': ['magnet'], 'def': 'a device that attracts iron and produces a magnetic field', 'name': 'magnet'}, {'frequency': 'r', 'id': 674, 'synset': 'mail_slot.n.01', 'synonyms': ['mail_slot'], 'def': 'a slot (usually in a door) through which mail can be delivered', 'name': 'mail_slot'}, {'frequency': 'c', 'id': 675, 'synset': 'mailbox.n.01', 'synonyms': ['mailbox_(at_home)', 'letter_box_(at_home)'], 'def': 'a private box for delivery of mail', 'name': 'mailbox_(at_home)'}, {'frequency': 'r', 'id': 676, 'synset': 'mallet.n.01', 'synonyms': ['mallet'], 'def': 'a sports implement with a long handle and a hammer-like head used to hit a ball', 'name': 'mallet'}, {'frequency': 'r', 'id': 677, 'synset': 'mammoth.n.01', 'synonyms': ['mammoth'], 'def': 'any of numerous extinct elephants widely distributed in the Pleistocene', 'name': 'mammoth'}, {'frequency': 'c', 'id': 678, 'synset': 'mandarin.n.05', 'synonyms': ['mandarin_orange'], 'def': 'a somewhat flat reddish-orange loose skinned citrus of China', 'name': 'mandarin_orange'}, {'frequency': 'c', 'id': 679, 'synset': 'manger.n.01', 'synonyms': ['manger', 'trough'], 'def': 'a container (usually in a barn or stable) from which cattle or horses feed', 'name': 'manger'}, {'frequency': 'f', 'id': 680, 'synset': 'manhole.n.01', 'synonyms': ['manhole'], 'def': 'a hole (usually with a flush cover) through which a person can gain access to an underground structure', 'name': 'manhole'}, {'frequency': 'c', 'id': 681, 'synset': 'map.n.01', 'synonyms': ['map'], 'def': "a diagrammatic representation of the earth's surface (or part of it)", 'name': 'map'}, {'frequency': 'c', 'id': 682, 'synset': 'marker.n.03', 'synonyms': ['marker'], 'def': 'a writing implement for making a mark', 'name': 'marker'}, {'frequency': 'r', 'id': 683, 'synset': 'martini.n.01', 'synonyms': ['martini'], 'def': 'a cocktail made of gin (or vodka) with dry vermouth', 'name': 'martini'}, {'frequency': 'r', 'id': 684, 'synset': 'mascot.n.01', 'synonyms': ['mascot'], 'def': 'a person or animal that is adopted by a team or other group as a symbolic figure', 'name': 'mascot'}, {'frequency': 'c', 'id': 685, 'synset': 'mashed_potato.n.01', 'synonyms': ['mashed_potato'], 'def': 'potato that has been peeled and boiled and then mashed', 'name': 'mashed_potato'}, {'frequency': 'r', 'id': 686, 'synset': 'masher.n.02', 'synonyms': ['masher'], 'def': 'a kitchen utensil used for mashing (e.g. potatoes)', 'name': 'masher'}, {'frequency': 'f', 'id': 687, 'synset': 'mask.n.04', 'synonyms': ['mask', 'facemask'], 'def': 'a protective covering worn over the face', 'name': 'mask'}, {'frequency': 'f', 'id': 688, 'synset': 'mast.n.01', 'synonyms': ['mast'], 'def': 'a vertical spar for supporting sails', 'name': 'mast'}, {'frequency': 'c', 'id': 689, 'synset': 'mat.n.03', 'synonyms': ['mat_(gym_equipment)', 'gym_mat'], 'def': 'sports equipment consisting of a piece of thick padding on the floor for gymnastics', 'name': 'mat_(gym_equipment)'}, {'frequency': 'r', 'id': 690, 'synset': 'matchbox.n.01', 'synonyms': ['matchbox'], 'def': 'a box for holding matches', 'name': 'matchbox'}, {'frequency': 'f', 'id': 691, 'synset': 'mattress.n.01', 'synonyms': ['mattress'], 'def': 'a thick pad filled with resilient material used as a bed or part of a bed', 'name': 'mattress'}, {'frequency': 'c', 'id': 692, 'synset': 'measuring_cup.n.01', 'synonyms': ['measuring_cup'], 'def': 'graduated cup used to measure liquid or granular ingredients', 'name': 'measuring_cup'}, {'frequency': 'c', 'id': 693, 'synset': 'measuring_stick.n.01', 'synonyms': ['measuring_stick', 'ruler_(measuring_stick)', 'measuring_rod'], 'def': 'measuring instrument having a sequence of marks at regular intervals', 'name': 'measuring_stick'}, {'frequency': 'c', 'id': 694, 'synset': 'meatball.n.01', 'synonyms': ['meatball'], 'def': 'ground meat formed into a ball and fried or simmered in broth', 'name': 'meatball'}, {'frequency': 'c', 'id': 695, 'synset': 'medicine.n.02', 'synonyms': ['medicine'], 'def': 'something that treats or prevents or alleviates the symptoms of disease', 'name': 'medicine'}, {'frequency': 'r', 'id': 696, 'synset': 'melon.n.01', 'synonyms': ['melon'], 'def': 'fruit of the gourd family having a hard rind and sweet juicy flesh', 'name': 'melon'}, {'frequency': 'f', 'id': 697, 'synset': 'microphone.n.01', 'synonyms': ['microphone'], 'def': 'device for converting sound waves into electrical energy', 'name': 'microphone'}, {'frequency': 'r', 'id': 698, 'synset': 'microscope.n.01', 'synonyms': ['microscope'], 'def': 'magnifier of the image of small objects', 'name': 'microscope'}, {'frequency': 'f', 'id': 699, 'synset': 'microwave.n.02', 'synonyms': ['microwave_oven'], 'def': 'kitchen appliance that cooks food by passing an electromagnetic wave through it', 'name': 'microwave_oven'}, {'frequency': 'r', 'id': 700, 'synset': 'milestone.n.01', 'synonyms': ['milestone', 'milepost'], 'def': 'stone post at side of a road to show distances', 'name': 'milestone'}, {'frequency': 'c', 'id': 701, 'synset': 'milk.n.01', 'synonyms': ['milk'], 'def': 'a white nutritious liquid secreted by mammals and used as food by human beings', 'name': 'milk'}, {'frequency': 'f', 'id': 702, 'synset': 'minivan.n.01', 'synonyms': ['minivan'], 'def': 'a small box-shaped passenger van', 'name': 'minivan'}, {'frequency': 'r', 'id': 703, 'synset': 'mint.n.05', 'synonyms': ['mint_candy'], 'def': 'a candy that is flavored with a mint oil', 'name': 'mint_candy'}, {'frequency': 'f', 'id': 704, 'synset': 'mirror.n.01', 'synonyms': ['mirror'], 'def': 'polished surface that forms images by reflecting light', 'name': 'mirror'}, {'frequency': 'c', 'id': 705, 'synset': 'mitten.n.01', 'synonyms': ['mitten'], 'def': 'glove that encases the thumb separately and the other four fingers together', 'name': 'mitten'}, {'frequency': 'c', 'id': 706, 'synset': 'mixer.n.04', 'synonyms': ['mixer_(kitchen_tool)', 'stand_mixer'], 'def': 'a kitchen utensil that is used for mixing foods', 'name': 'mixer_(kitchen_tool)'}, {'frequency': 'c', 'id': 707, 'synset': 'money.n.03', 'synonyms': ['money'], 'def': 'the official currency issued by a government or national bank', 'name': 'money'}, {'frequency': 'f', 'id': 708, 'synset': 'monitor.n.04', 'synonyms': ['monitor_(computer_equipment) computer_monitor'], 'def': 'a computer monitor', 'name': 'monitor_(computer_equipment) computer_monitor'}, {'frequency': 'c', 'id': 709, 'synset': 'monkey.n.01', 'synonyms': ['monkey'], 'def': 'any of various long-tailed primates', 'name': 'monkey'}, {'frequency': 'f', 'id': 710, 'synset': 'motor.n.01', 'synonyms': ['motor'], 'def': 'machine that converts other forms of energy into mechanical energy and so imparts motion', 'name': 'motor'}, {'frequency': 'f', 'id': 711, 'synset': 'motor_scooter.n.01', 'synonyms': ['motor_scooter', 'scooter'], 'def': 'a wheeled vehicle with small wheels and a low-powered engine', 'name': 'motor_scooter'}, {'frequency': 'r', 'id': 712, 'synset': 'motor_vehicle.n.01', 'synonyms': ['motor_vehicle', 'automotive_vehicle'], 'def': 'a self-propelled wheeled vehicle that does not run on rails', 'name': 'motor_vehicle'}, {'frequency': 'r', 'id': 713, 'synset': 'motorboat.n.01', 'synonyms': ['motorboat', 'powerboat'], 'def': 'a boat propelled by an internal-combustion engine', 'name': 'motorboat'}, {'frequency': 'f', 'id': 714, 'synset': 'motorcycle.n.01', 'synonyms': ['motorcycle'], 'def': 'a motor vehicle with two wheels and a strong frame', 'name': 'motorcycle'}, {'frequency': 'f', 'id': 715, 'synset': 'mound.n.01', 'synonyms': ['mound_(baseball)', "pitcher's_mound"], 'def': '(baseball) the slight elevation on which the pitcher stands', 'name': 'mound_(baseball)'}, {'frequency': 'r', 'id': 716, 'synset': 'mouse.n.01', 'synonyms': ['mouse_(animal_rodent)'], 'def': 'a small rodent with pointed snouts and small ears on elongated bodies with slender usually hairless tails', 'name': 'mouse_(animal_rodent)'}, {'frequency': 'f', 'id': 717, 'synset': 'mouse.n.04', 'synonyms': ['mouse_(computer_equipment)', 'computer_mouse'], 'def': 'a computer input device that controls an on-screen pointer', 'name': 'mouse_(computer_equipment)'}, {'frequency': 'f', 'id': 718, 'synset': 'mousepad.n.01', 'synonyms': ['mousepad'], 'def': 'a small portable pad that provides an operating surface for a computer mouse', 'name': 'mousepad'}, {'frequency': 'c', 'id': 719, 'synset': 'muffin.n.01', 'synonyms': ['muffin'], 'def': 'a sweet quick bread baked in a cup-shaped pan', 'name': 'muffin'}, {'frequency': 'f', 'id': 720, 'synset': 'mug.n.04', 'synonyms': ['mug'], 'def': 'with handle and usually cylindrical', 'name': 'mug'}, {'frequency': 'f', 'id': 721, 'synset': 'mushroom.n.02', 'synonyms': ['mushroom'], 'def': 'a common mushroom', 'name': 'mushroom'}, {'frequency': 'r', 'id': 722, 'synset': 'music_stool.n.01', 'synonyms': ['music_stool', 'piano_stool'], 'def': 'a stool for piano players; usually adjustable in height', 'name': 'music_stool'}, {'frequency': 'r', 'id': 723, 'synset': 'musical_instrument.n.01', 'synonyms': ['musical_instrument', 'instrument_(musical)'], 'def': 'any of various devices or contrivances that can be used to produce musical tones or sounds', 'name': 'musical_instrument'}, {'frequency': 'r', 'id': 724, 'synset': 'nailfile.n.01', 'synonyms': ['nailfile'], 'def': 'a small flat file for shaping the nails', 'name': 'nailfile'}, {'frequency': 'r', 'id': 725, 'synset': 'nameplate.n.01', 'synonyms': ['nameplate'], 'def': 'a plate bearing a name', 'name': 'nameplate'}, {'frequency': 'f', 'id': 726, 'synset': 'napkin.n.01', 'synonyms': ['napkin', 'table_napkin', 'serviette'], 'def': 'a small piece of table linen or paper that is used to wipe the mouth and to cover the lap in order to protect clothing', 'name': 'napkin'}, {'frequency': 'r', 'id': 727, 'synset': 'neckerchief.n.01', 'synonyms': ['neckerchief'], 'def': 'a kerchief worn around the neck', 'name': 'neckerchief'}, {'frequency': 'f', 'id': 728, 'synset': 'necklace.n.01', 'synonyms': ['necklace'], 'def': 'jewelry consisting of a cord or chain (often bearing gems) worn about the neck as an ornament', 'name': 'necklace'}, {'frequency': 'f', 'id': 729, 'synset': 'necktie.n.01', 'synonyms': ['necktie', 'tie_(necktie)'], 'def': 'neckwear consisting of a long narrow piece of material worn under a collar and tied in knot at the front', 'name': 'necktie'}, {'frequency': 'r', 'id': 730, 'synset': 'needle.n.03', 'synonyms': ['needle'], 'def': 'a sharp pointed implement (usually metal)', 'name': 'needle'}, {'frequency': 'c', 'id': 731, 'synset': 'nest.n.01', 'synonyms': ['nest'], 'def': 'a structure in which animals lay eggs or give birth to their young', 'name': 'nest'}, {'frequency': 'r', 'id': 732, 'synset': 'newsstand.n.01', 'synonyms': ['newsstand'], 'def': 'a stall where newspapers and other periodicals are sold', 'name': 'newsstand'}, {'frequency': 'c', 'id': 733, 'synset': 'nightwear.n.01', 'synonyms': ['nightshirt', 'nightwear', 'sleepwear', 'nightclothes'], 'def': 'garments designed to be worn in bed', 'name': 'nightshirt'}, {'frequency': 'r', 'id': 734, 'synset': 'nosebag.n.01', 'synonyms': ['nosebag_(for_animals)', 'feedbag'], 'def': 'a canvas bag that is used to feed an animal (such as a horse); covers the muzzle and fastens at the top of the head', 'name': 'nosebag_(for_animals)'}, {'frequency': 'r', 'id': 735, 'synset': 'noseband.n.01', 'synonyms': ['noseband_(for_animals)', 'nosepiece_(for_animals)'], 'def': "a strap that is the part of a bridle that goes over the animal's nose", 'name': 'noseband_(for_animals)'}, {'frequency': 'f', 'id': 736, 'synset': 'notebook.n.01', 'synonyms': ['notebook'], 'def': 'a book with blank pages for recording notes or memoranda', 'name': 'notebook'}, {'frequency': 'c', 'id': 737, 'synset': 'notepad.n.01', 'synonyms': ['notepad'], 'def': 'a pad of paper for keeping notes', 'name': 'notepad'}, {'frequency': 'c', 'id': 738, 'synset': 'nut.n.03', 'synonyms': ['nut'], 'def': 'a small metal block (usually square or hexagonal) with internal screw thread to be fitted onto a bolt', 'name': 'nut'}, {'frequency': 'r', 'id': 739, 'synset': 'nutcracker.n.01', 'synonyms': ['nutcracker'], 'def': 'a hand tool used to crack nuts open', 'name': 'nutcracker'}, {'frequency': 'c', 'id': 740, 'synset': 'oar.n.01', 'synonyms': ['oar'], 'def': 'an implement used to propel or steer a boat', 'name': 'oar'}, {'frequency': 'r', 'id': 741, 'synset': 'octopus.n.01', 'synonyms': ['octopus_(food)'], 'def': 'tentacles of octopus prepared as food', 'name': 'octopus_(food)'}, {'frequency': 'r', 'id': 742, 'synset': 'octopus.n.02', 'synonyms': ['octopus_(animal)'], 'def': 'bottom-living cephalopod having a soft oval body with eight long tentacles', 'name': 'octopus_(animal)'}, {'frequency': 'c', 'id': 743, 'synset': 'oil_lamp.n.01', 'synonyms': ['oil_lamp', 'kerosene_lamp', 'kerosine_lamp'], 'def': 'a lamp that burns oil (as kerosine) for light', 'name': 'oil_lamp'}, {'frequency': 'c', 'id': 744, 'synset': 'olive_oil.n.01', 'synonyms': ['olive_oil'], 'def': 'oil from olives', 'name': 'olive_oil'}, {'frequency': 'r', 'id': 745, 'synset': 'omelet.n.01', 'synonyms': ['omelet', 'omelette'], 'def': 'beaten eggs cooked until just set; may be folded around e.g. ham or cheese or jelly', 'name': 'omelet'}, {'frequency': 'f', 'id': 746, 'synset': 'onion.n.01', 'synonyms': ['onion'], 'def': 'the bulb of an onion plant', 'name': 'onion'}, {'frequency': 'f', 'id': 747, 'synset': 'orange.n.01', 'synonyms': ['orange_(fruit)'], 'def': 'orange (FRUIT of an orange tree)', 'name': 'orange_(fruit)'}, {'frequency': 'c', 'id': 748, 'synset': 'orange_juice.n.01', 'synonyms': ['orange_juice'], 'def': 'bottled or freshly squeezed juice of oranges', 'name': 'orange_juice'}, {'frequency': 'r', 'id': 749, 'synset': 'oregano.n.01', 'synonyms': ['oregano', 'marjoram'], 'def': 'aromatic Eurasian perennial herb used in cooking and baking', 'name': 'oregano'}, {'frequency': 'c', 'id': 750, 'synset': 'ostrich.n.02', 'synonyms': ['ostrich'], 'def': 'fast-running African flightless bird with two-toed feet; largest living bird', 'name': 'ostrich'}, {'frequency': 'c', 'id': 751, 'synset': 'ottoman.n.03', 'synonyms': ['ottoman', 'pouf', 'pouffe', 'hassock'], 'def': 'thick cushion used as a seat', 'name': 'ottoman'}, {'frequency': 'c', 'id': 752, 'synset': 'overall.n.01', 'synonyms': ['overalls_(clothing)'], 'def': 'work clothing consisting of denim trousers usually with a bib and shoulder straps', 'name': 'overalls_(clothing)'}, {'frequency': 'c', 'id': 753, 'synset': 'owl.n.01', 'synonyms': ['owl'], 'def': 'nocturnal bird of prey with hawk-like beak and claws and large head with front-facing eyes', 'name': 'owl'}, {'frequency': 'c', 'id': 754, 'synset': 'packet.n.03', 'synonyms': ['packet'], 'def': 'a small package or bundle', 'name': 'packet'}, {'frequency': 'r', 'id': 755, 'synset': 'pad.n.03', 'synonyms': ['inkpad', 'inking_pad', 'stamp_pad'], 'def': 'absorbent material saturated with ink used to transfer ink evenly to a rubber stamp', 'name': 'inkpad'}, {'frequency': 'c', 'id': 756, 'synset': 'pad.n.04', 'synonyms': ['pad'], 'def': 'a flat mass of soft material used for protection, stuffing, or comfort', 'name': 'pad'}, {'frequency': 'c', 'id': 757, 'synset': 'paddle.n.04', 'synonyms': ['paddle', 'boat_paddle'], 'def': 'a short light oar used without an oarlock to propel a canoe or small boat', 'name': 'paddle'}, {'frequency': 'c', 'id': 758, 'synset': 'padlock.n.01', 'synonyms': ['padlock'], 'def': 'a detachable, portable lock', 'name': 'padlock'}, {'frequency': 'r', 'id': 759, 'synset': 'paintbox.n.01', 'synonyms': ['paintbox'], 'def': "a box containing a collection of cubes or tubes of artists' paint", 'name': 'paintbox'}, {'frequency': 'c', 'id': 760, 'synset': 'paintbrush.n.01', 'synonyms': ['paintbrush'], 'def': 'a brush used as an applicator to apply paint', 'name': 'paintbrush'}, {'frequency': 'f', 'id': 761, 'synset': 'painting.n.01', 'synonyms': ['painting'], 'def': 'graphic art consisting of an artistic composition made by applying paints to a surface', 'name': 'painting'}, {'frequency': 'c', 'id': 762, 'synset': 'pajama.n.02', 'synonyms': ['pajamas', 'pyjamas'], 'def': 'loose-fitting nightclothes worn for sleeping or lounging', 'name': 'pajamas'}, {'frequency': 'c', 'id': 763, 'synset': 'palette.n.02', 'synonyms': ['palette', 'pallet'], 'def': 'board that provides a flat surface on which artists mix paints and the range of colors used', 'name': 'palette'}, {'frequency': 'f', 'id': 764, 'synset': 'pan.n.01', 'synonyms': ['pan_(for_cooking)', 'cooking_pan'], 'def': 'cooking utensil consisting of a wide metal vessel', 'name': 'pan_(for_cooking)'}, {'frequency': 'r', 'id': 765, 'synset': 'pan.n.03', 'synonyms': ['pan_(metal_container)'], 'def': 'shallow container made of metal', 'name': 'pan_(metal_container)'}, {'frequency': 'c', 'id': 766, 'synset': 'pancake.n.01', 'synonyms': ['pancake'], 'def': 'a flat cake of thin batter fried on both sides on a griddle', 'name': 'pancake'}, {'frequency': 'r', 'id': 767, 'synset': 'pantyhose.n.01', 'synonyms': ['pantyhose'], 'def': "a woman's tights consisting of underpants and stockings", 'name': 'pantyhose'}, {'frequency': 'r', 'id': 768, 'synset': 'papaya.n.02', 'synonyms': ['papaya'], 'def': 'large oval melon-like tropical fruit with yellowish flesh', 'name': 'papaya'}, {'frequency': 'r', 'id': 769, 'synset': 'paper_clip.n.01', 'synonyms': ['paperclip'], 'def': 'a wire or plastic clip for holding sheets of paper together', 'name': 'paperclip'}, {'frequency': 'f', 'id': 770, 'synset': 'paper_plate.n.01', 'synonyms': ['paper_plate'], 'def': 'a disposable plate made of cardboard', 'name': 'paper_plate'}, {'frequency': 'f', 'id': 771, 'synset': 'paper_towel.n.01', 'synonyms': ['paper_towel'], 'def': 'a disposable towel made of absorbent paper', 'name': 'paper_towel'}, {'frequency': 'r', 'id': 772, 'synset': 'paperback_book.n.01', 'synonyms': ['paperback_book', 'paper-back_book', 'softback_book', 'soft-cover_book'], 'def': 'a book with paper covers', 'name': 'paperback_book'}, {'frequency': 'r', 'id': 773, 'synset': 'paperweight.n.01', 'synonyms': ['paperweight'], 'def': 'a weight used to hold down a stack of papers', 'name': 'paperweight'}, {'frequency': 'c', 'id': 774, 'synset': 'parachute.n.01', 'synonyms': ['parachute'], 'def': 'rescue equipment consisting of a device that fills with air and retards your fall', 'name': 'parachute'}, {'frequency': 'r', 'id': 775, 'synset': 'parakeet.n.01', 'synonyms': ['parakeet', 'parrakeet', 'parroket', 'paraquet', 'paroquet', 'parroquet'], 'def': 'any of numerous small slender long-tailed parrots', 'name': 'parakeet'}, {'frequency': 'c', 'id': 776, 'synset': 'parasail.n.01', 'synonyms': ['parasail_(sports)'], 'def': 'parachute that will lift a person up into the air when it is towed by a motorboat or a car', 'name': 'parasail_(sports)'}, {'frequency': 'r', 'id': 777, 'synset': 'parchment.n.01', 'synonyms': ['parchment'], 'def': 'a superior paper resembling sheepskin', 'name': 'parchment'}, {'frequency': 'r', 'id': 778, 'synset': 'parka.n.01', 'synonyms': ['parka', 'anorak'], 'def': "a kind of heavy jacket (`windcheater' is a British term)", 'name': 'parka'}, {'frequency': 'f', 'id': 779, 'synset': 'parking_meter.n.01', 'synonyms': ['parking_meter'], 'def': 'a coin-operated timer located next to a parking space', 'name': 'parking_meter'}, {'frequency': 'c', 'id': 780, 'synset': 'parrot.n.01', 'synonyms': ['parrot'], 'def': 'usually brightly colored tropical birds with short hooked beaks and the ability to mimic sounds', 'name': 'parrot'}, {'frequency': 'c', 'id': 781, 'synset': 'passenger_car.n.01', 'synonyms': ['passenger_car_(part_of_a_train)', 'coach_(part_of_a_train)'], 'def': 'a railcar where passengers ride', 'name': 'passenger_car_(part_of_a_train)'}, {'frequency': 'r', 'id': 782, 'synset': 'passenger_ship.n.01', 'synonyms': ['passenger_ship'], 'def': 'a ship built to carry passengers', 'name': 'passenger_ship'}, {'frequency': 'r', 'id': 783, 'synset': 'passport.n.02', 'synonyms': ['passport'], 'def': 'a document issued by a country to a citizen allowing that person to travel abroad and re-enter the home country', 'name': 'passport'}, {'frequency': 'f', 'id': 784, 'synset': 'pastry.n.02', 'synonyms': ['pastry'], 'def': 'any of various baked foods made of dough or batter', 'name': 'pastry'}, {'frequency': 'r', 'id': 785, 'synset': 'patty.n.01', 'synonyms': ['patty_(food)'], 'def': 'small flat mass of chopped food', 'name': 'patty_(food)'}, {'frequency': 'c', 'id': 786, 'synset': 'pea.n.01', 'synonyms': ['pea_(food)'], 'def': 'seed of a pea plant used for food', 'name': 'pea_(food)'}, {'frequency': 'c', 'id': 787, 'synset': 'peach.n.03', 'synonyms': ['peach'], 'def': 'downy juicy fruit with sweet yellowish or whitish flesh', 'name': 'peach'}, {'frequency': 'c', 'id': 788, 'synset': 'peanut_butter.n.01', 'synonyms': ['peanut_butter'], 'def': 'a spread made from ground peanuts', 'name': 'peanut_butter'}, {'frequency': 'c', 'id': 789, 'synset': 'pear.n.01', 'synonyms': ['pear'], 'def': 'sweet juicy gritty-textured fruit available in many varieties', 'name': 'pear'}, {'frequency': 'r', 'id': 790, 'synset': 'peeler.n.03', 'synonyms': ['peeler_(tool_for_fruit_and_vegetables)'], 'def': 'a device for peeling vegetables or fruits', 'name': 'peeler_(tool_for_fruit_and_vegetables)'}, {'frequency': 'r', 'id': 791, 'synset': 'pegboard.n.01', 'synonyms': ['pegboard'], 'def': 'a board perforated with regularly spaced holes into which pegs can be fitted', 'name': 'pegboard'}, {'frequency': 'c', 'id': 792, 'synset': 'pelican.n.01', 'synonyms': ['pelican'], 'def': 'large long-winged warm-water seabird having a large bill with a distensible pouch for fish', 'name': 'pelican'}, {'frequency': 'f', 'id': 793, 'synset': 'pen.n.01', 'synonyms': ['pen'], 'def': 'a writing implement with a point from which ink flows', 'name': 'pen'}, {'frequency': 'c', 'id': 794, 'synset': 'pencil.n.01', 'synonyms': ['pencil'], 'def': 'a thin cylindrical pointed writing implement made of wood and graphite', 'name': 'pencil'}, {'frequency': 'r', 'id': 795, 'synset': 'pencil_box.n.01', 'synonyms': ['pencil_box', 'pencil_case'], 'def': 'a box for holding pencils', 'name': 'pencil_box'}, {'frequency': 'r', 'id': 796, 'synset': 'pencil_sharpener.n.01', 'synonyms': ['pencil_sharpener'], 'def': 'a rotary implement for sharpening the point on pencils', 'name': 'pencil_sharpener'}, {'frequency': 'r', 'id': 797, 'synset': 'pendulum.n.01', 'synonyms': ['pendulum'], 'def': 'an apparatus consisting of an object mounted so that it swings freely under the influence of gravity', 'name': 'pendulum'}, {'frequency': 'c', 'id': 798, 'synset': 'penguin.n.01', 'synonyms': ['penguin'], 'def': 'short-legged flightless birds of cold southern regions having webbed feet and wings modified as flippers', 'name': 'penguin'}, {'frequency': 'r', 'id': 799, 'synset': 'pennant.n.02', 'synonyms': ['pennant'], 'def': 'a flag longer than it is wide (and often tapering)', 'name': 'pennant'}, {'frequency': 'r', 'id': 800, 'synset': 'penny.n.02', 'synonyms': ['penny_(coin)'], 'def': 'a coin worth one-hundredth of the value of the basic unit', 'name': 'penny_(coin)'}, {'frequency': 'c', 'id': 801, 'synset': 'pepper.n.03', 'synonyms': ['pepper', 'peppercorn'], 'def': 'pungent seasoning from the berry of the common pepper plant; whole or ground', 'name': 'pepper'}, {'frequency': 'c', 'id': 802, 'synset': 'pepper_mill.n.01', 'synonyms': ['pepper_mill', 'pepper_grinder'], 'def': 'a mill for grinding pepper', 'name': 'pepper_mill'}, {'frequency': 'c', 'id': 803, 'synset': 'perfume.n.02', 'synonyms': ['perfume'], 'def': 'a toiletry that emits and diffuses a fragrant odor', 'name': 'perfume'}, {'frequency': 'r', 'id': 804, 'synset': 'persimmon.n.02', 'synonyms': ['persimmon'], 'def': 'orange fruit resembling a plum; edible when fully ripe', 'name': 'persimmon'}, {'frequency': 'f', 'id': 805, 'synset': 'person.n.01', 'synonyms': ['baby', 'child', 'boy', 'girl', 'man', 'woman', 'person', 'human'], 'def': 'a human being', 'name': 'baby'}, {'frequency': 'r', 'id': 806, 'synset': 'pet.n.01', 'synonyms': ['pet'], 'def': 'a domesticated animal kept for companionship or amusement', 'name': 'pet'}, {'frequency': 'r', 'id': 807, 'synset': 'petfood.n.01', 'synonyms': ['petfood', 'pet-food'], 'def': 'food prepared for animal pets', 'name': 'petfood'}, {'frequency': 'r', 'id': 808, 'synset': 'pew.n.01', 'synonyms': ['pew_(church_bench)', 'church_bench'], 'def': 'long bench with backs; used in church by the congregation', 'name': 'pew_(church_bench)'}, {'frequency': 'r', 'id': 809, 'synset': 'phonebook.n.01', 'synonyms': ['phonebook', 'telephone_book', 'telephone_directory'], 'def': 'a directory containing an alphabetical list of telephone subscribers and their telephone numbers', 'name': 'phonebook'}, {'frequency': 'c', 'id': 810, 'synset': 'phonograph_record.n.01', 'synonyms': ['phonograph_record', 'phonograph_recording', 'record_(phonograph_recording)'], 'def': 'sound recording consisting of a typically black disk with a continuous groove', 'name': 'phonograph_record'}, {'frequency': 'c', 'id': 811, 'synset': 'piano.n.01', 'synonyms': ['piano'], 'def': 'a keyboard instrument that is played by depressing keys that cause hammers to strike tuned strings and produce sounds', 'name': 'piano'}, {'frequency': 'f', 'id': 812, 'synset': 'pickle.n.01', 'synonyms': ['pickle'], 'def': 'vegetables (especially cucumbers) preserved in brine or vinegar', 'name': 'pickle'}, {'frequency': 'f', 'id': 813, 'synset': 'pickup.n.01', 'synonyms': ['pickup_truck'], 'def': 'a light truck with an open body and low sides and a tailboard', 'name': 'pickup_truck'}, {'frequency': 'c', 'id': 814, 'synset': 'pie.n.01', 'synonyms': ['pie'], 'def': 'dish baked in pastry-lined pan often with a pastry top', 'name': 'pie'}, {'frequency': 'c', 'id': 815, 'synset': 'pigeon.n.01', 'synonyms': ['pigeon'], 'def': 'wild and domesticated birds having a heavy body and short legs', 'name': 'pigeon'}, {'frequency': 'r', 'id': 816, 'synset': 'piggy_bank.n.01', 'synonyms': ['piggy_bank', 'penny_bank'], 'def': "a child's coin bank (often shaped like a pig)", 'name': 'piggy_bank'}, {'frequency': 'f', 'id': 817, 'synset': 'pillow.n.01', 'synonyms': ['pillow'], 'def': 'a cushion to support the head of a sleeping person', 'name': 'pillow'}, {'frequency': 'r', 'id': 818, 'synset': 'pin.n.09', 'synonyms': ['pin_(non_jewelry)'], 'def': 'a small slender (often pointed) piece of wood or metal used to support or fasten or attach things', 'name': 'pin_(non_jewelry)'}, {'frequency': 'f', 'id': 819, 'synset': 'pineapple.n.02', 'synonyms': ['pineapple'], 'def': 'large sweet fleshy tropical fruit with a tuft of stiff leaves', 'name': 'pineapple'}, {'frequency': 'c', 'id': 820, 'synset': 'pinecone.n.01', 'synonyms': ['pinecone'], 'def': 'the seed-producing cone of a pine tree', 'name': 'pinecone'}, {'frequency': 'r', 'id': 821, 'synset': 'ping-pong_ball.n.01', 'synonyms': ['ping-pong_ball'], 'def': 'light hollow ball used in playing table tennis', 'name': 'ping-pong_ball'}, {'frequency': 'r', 'id': 822, 'synset': 'pinwheel.n.03', 'synonyms': ['pinwheel'], 'def': 'a toy consisting of vanes of colored paper or plastic that is pinned to a stick and spins when it is pointed into the wind', 'name': 'pinwheel'}, {'frequency': 'r', 'id': 823, 'synset': 'pipe.n.01', 'synonyms': ['tobacco_pipe'], 'def': 'a tube with a small bowl at one end; used for smoking tobacco', 'name': 'tobacco_pipe'}, {'frequency': 'f', 'id': 824, 'synset': 'pipe.n.02', 'synonyms': ['pipe', 'piping'], 'def': 'a long tube made of metal or plastic that is used to carry water or oil or gas etc.', 'name': 'pipe'}, {'frequency': 'r', 'id': 825, 'synset': 'pistol.n.01', 'synonyms': ['pistol', 'handgun'], 'def': 'a firearm that is held and fired with one hand', 'name': 'pistol'}, {'frequency': 'r', 'id': 826, 'synset': 'pita.n.01', 'synonyms': ['pita_(bread)', 'pocket_bread'], 'def': 'usually small round bread that can open into a pocket for filling', 'name': 'pita_(bread)'}, {'frequency': 'f', 'id': 827, 'synset': 'pitcher.n.02', 'synonyms': ['pitcher_(vessel_for_liquid)', 'ewer'], 'def': 'an open vessel with a handle and a spout for pouring', 'name': 'pitcher_(vessel_for_liquid)'}, {'frequency': 'r', 'id': 828, 'synset': 'pitchfork.n.01', 'synonyms': ['pitchfork'], 'def': 'a long-handled hand tool with sharp widely spaced prongs for lifting and pitching hay', 'name': 'pitchfork'}, {'frequency': 'f', 'id': 829, 'synset': 'pizza.n.01', 'synonyms': ['pizza'], 'def': 'Italian open pie made of thin bread dough spread with a spiced mixture of e.g. tomato sauce and cheese', 'name': 'pizza'}, {'frequency': 'f', 'id': 830, 'synset': 'place_mat.n.01', 'synonyms': ['place_mat'], 'def': 'a mat placed on a table for an individual place setting', 'name': 'place_mat'}, {'frequency': 'f', 'id': 831, 'synset': 'plate.n.04', 'synonyms': ['plate'], 'def': 'dish on which food is served or from which food is eaten', 'name': 'plate'}, {'frequency': 'c', 'id': 832, 'synset': 'platter.n.01', 'synonyms': ['platter'], 'def': 'a large shallow dish used for serving food', 'name': 'platter'}, {'frequency': 'r', 'id': 833, 'synset': 'playing_card.n.01', 'synonyms': ['playing_card'], 'def': 'one of a pack of cards that are used to play card games', 'name': 'playing_card'}, {'frequency': 'r', 'id': 834, 'synset': 'playpen.n.01', 'synonyms': ['playpen'], 'def': 'a portable enclosure in which babies may be left to play', 'name': 'playpen'}, {'frequency': 'c', 'id': 835, 'synset': 'pliers.n.01', 'synonyms': ['pliers', 'plyers'], 'def': 'a gripping hand tool with two hinged arms and (usually) serrated jaws', 'name': 'pliers'}, {'frequency': 'r', 'id': 836, 'synset': 'plow.n.01', 'synonyms': ['plow_(farm_equipment)', 'plough_(farm_equipment)'], 'def': 'a farm tool having one or more heavy blades to break the soil and cut a furrow prior to sowing', 'name': 'plow_(farm_equipment)'}, {'frequency': 'r', 'id': 837, 'synset': 'pocket_watch.n.01', 'synonyms': ['pocket_watch'], 'def': 'a watch that is carried in a small watch pocket', 'name': 'pocket_watch'}, {'frequency': 'c', 'id': 838, 'synset': 'pocketknife.n.01', 'synonyms': ['pocketknife'], 'def': 'a knife with a blade that folds into the handle; suitable for carrying in the pocket', 'name': 'pocketknife'}, {'frequency': 'c', 'id': 839, 'synset': 'poker.n.01', 'synonyms': ['poker_(fire_stirring_tool)', 'stove_poker', 'fire_hook'], 'def': 'fire iron consisting of a metal rod with a handle; used to stir a fire', 'name': 'poker_(fire_stirring_tool)'}, {'frequency': 'f', 'id': 840, 'synset': 'pole.n.01', 'synonyms': ['pole', 'post'], 'def': 'a long (usually round) rod of wood or metal or plastic', 'name': 'pole'}, {'frequency': 'r', 'id': 841, 'synset': 'police_van.n.01', 'synonyms': ['police_van', 'police_wagon', 'paddy_wagon', 'patrol_wagon'], 'def': 'van used by police to transport prisoners', 'name': 'police_van'}, {'frequency': 'f', 'id': 842, 'synset': 'polo_shirt.n.01', 'synonyms': ['polo_shirt', 'sport_shirt'], 'def': 'a shirt with short sleeves designed for comfort and casual wear', 'name': 'polo_shirt'}, {'frequency': 'r', 'id': 843, 'synset': 'poncho.n.01', 'synonyms': ['poncho'], 'def': 'a blanket-like cloak with a hole in the center for the head', 'name': 'poncho'}, {'frequency': 'c', 'id': 844, 'synset': 'pony.n.05', 'synonyms': ['pony'], 'def': 'any of various breeds of small gentle horses usually less than five feet high at the shoulder', 'name': 'pony'}, {'frequency': 'r', 'id': 845, 'synset': 'pool_table.n.01', 'synonyms': ['pool_table', 'billiard_table', 'snooker_table'], 'def': 'game equipment consisting of a heavy table on which pool is played', 'name': 'pool_table'}, {'frequency': 'f', 'id': 846, 'synset': 'pop.n.02', 'synonyms': ['pop_(soda)', 'soda_(pop)', 'tonic', 'soft_drink'], 'def': 'a sweet drink containing carbonated water and flavoring', 'name': 'pop_(soda)'}, {'frequency': 'r', 'id': 847, 'synset': 'portrait.n.02', 'synonyms': ['portrait', 'portrayal'], 'def': 'any likeness of a person, in any medium', 'name': 'portrait'}, {'frequency': 'c', 'id': 848, 'synset': 'postbox.n.01', 'synonyms': ['postbox_(public)', 'mailbox_(public)'], 'def': 'public box for deposit of mail', 'name': 'postbox_(public)'}, {'frequency': 'c', 'id': 849, 'synset': 'postcard.n.01', 'synonyms': ['postcard', 'postal_card', 'mailing-card'], 'def': 'a card for sending messages by post without an envelope', 'name': 'postcard'}, {'frequency': 'f', 'id': 850, 'synset': 'poster.n.01', 'synonyms': ['poster', 'placard'], 'def': 'a sign posted in a public place as an advertisement', 'name': 'poster'}, {'frequency': 'f', 'id': 851, 'synset': 'pot.n.01', 'synonyms': ['pot'], 'def': 'metal or earthenware cooking vessel that is usually round and deep; often has a handle and lid', 'name': 'pot'}, {'frequency': 'f', 'id': 852, 'synset': 'pot.n.04', 'synonyms': ['flowerpot'], 'def': 'a container in which plants are cultivated', 'name': 'flowerpot'}, {'frequency': 'f', 'id': 853, 'synset': 'potato.n.01', 'synonyms': ['potato'], 'def': 'an edible tuber native to South America', 'name': 'potato'}, {'frequency': 'c', 'id': 854, 'synset': 'potholder.n.01', 'synonyms': ['potholder'], 'def': 'an insulated pad for holding hot pots', 'name': 'potholder'}, {'frequency': 'c', 'id': 855, 'synset': 'pottery.n.01', 'synonyms': ['pottery', 'clayware'], 'def': 'ceramic ware made from clay and baked in a kiln', 'name': 'pottery'}, {'frequency': 'c', 'id': 856, 'synset': 'pouch.n.01', 'synonyms': ['pouch'], 'def': 'a small or medium size container for holding or carrying things', 'name': 'pouch'}, {'frequency': 'r', 'id': 857, 'synset': 'power_shovel.n.01', 'synonyms': ['power_shovel', 'excavator', 'digger'], 'def': 'a machine for excavating', 'name': 'power_shovel'}, {'frequency': 'c', 'id': 858, 'synset': 'prawn.n.01', 'synonyms': ['prawn', 'shrimp'], 'def': 'any of various edible decapod crustaceans', 'name': 'prawn'}, {'frequency': 'f', 'id': 859, 'synset': 'printer.n.03', 'synonyms': ['printer', 'printing_machine'], 'def': 'a machine that prints', 'name': 'printer'}, {'frequency': 'c', 'id': 860, 'synset': 'projectile.n.01', 'synonyms': ['projectile_(weapon)', 'missile'], 'def': 'a weapon that is forcibly thrown or projected at a targets', 'name': 'projectile_(weapon)'}, {'frequency': 'c', 'id': 861, 'synset': 'projector.n.02', 'synonyms': ['projector'], 'def': 'an optical instrument that projects an enlarged image onto a screen', 'name': 'projector'}, {'frequency': 'f', 'id': 862, 'synset': 'propeller.n.01', 'synonyms': ['propeller', 'propellor'], 'def': 'a mechanical device that rotates to push against air or water', 'name': 'propeller'}, {'frequency': 'r', 'id': 863, 'synset': 'prune.n.01', 'synonyms': ['prune'], 'def': 'dried plum', 'name': 'prune'}, {'frequency': 'r', 'id': 864, 'synset': 'pudding.n.01', 'synonyms': ['pudding'], 'def': 'any of various soft thick unsweetened baked dishes', 'name': 'pudding'}, {'frequency': 'r', 'id': 865, 'synset': 'puffer.n.02', 'synonyms': ['puffer_(fish)', 'pufferfish', 'blowfish', 'globefish'], 'def': 'fishes whose elongated spiny body can inflate itself with water or air to form a globe', 'name': 'puffer_(fish)'}, {'frequency': 'r', 'id': 866, 'synset': 'puffin.n.01', 'synonyms': ['puffin'], 'def': 'seabirds having short necks and brightly colored compressed bills', 'name': 'puffin'}, {'frequency': 'r', 'id': 867, 'synset': 'pug.n.01', 'synonyms': ['pug-dog'], 'def': 'small compact smooth-coated breed of Asiatic origin having a tightly curled tail and broad flat wrinkled muzzle', 'name': 'pug-dog'}, {'frequency': 'c', 'id': 868, 'synset': 'pumpkin.n.02', 'synonyms': ['pumpkin'], 'def': 'usually large pulpy deep-yellow round fruit of the squash family maturing in late summer or early autumn', 'name': 'pumpkin'}, {'frequency': 'r', 'id': 869, 'synset': 'punch.n.03', 'synonyms': ['puncher'], 'def': 'a tool for making holes or indentations', 'name': 'puncher'}, {'frequency': 'r', 'id': 870, 'synset': 'puppet.n.01', 'synonyms': ['puppet', 'marionette'], 'def': 'a small figure of a person operated from above with strings by a puppeteer', 'name': 'puppet'}, {'frequency': 'r', 'id': 871, 'synset': 'puppy.n.01', 'synonyms': ['puppy'], 'def': 'a young dog', 'name': 'puppy'}, {'frequency': 'r', 'id': 872, 'synset': 'quesadilla.n.01', 'synonyms': ['quesadilla'], 'def': 'a tortilla that is filled with cheese and heated', 'name': 'quesadilla'}, {'frequency': 'r', 'id': 873, 'synset': 'quiche.n.02', 'synonyms': ['quiche'], 'def': 'a tart filled with rich unsweetened custard; often contains other ingredients (as cheese or ham or seafood or vegetables)', 'name': 'quiche'}, {'frequency': 'f', 'id': 874, 'synset': 'quilt.n.01', 'synonyms': ['quilt', 'comforter'], 'def': 'bedding made of two layers of cloth filled with stuffing and stitched together', 'name': 'quilt'}, {'frequency': 'c', 'id': 875, 'synset': 'rabbit.n.01', 'synonyms': ['rabbit'], 'def': 'any of various burrowing animals of the family Leporidae having long ears and short tails', 'name': 'rabbit'}, {'frequency': 'r', 'id': 876, 'synset': 'racer.n.02', 'synonyms': ['race_car', 'racing_car'], 'def': 'a fast car that competes in races', 'name': 'race_car'}, {'frequency': 'c', 'id': 877, 'synset': 'racket.n.04', 'synonyms': ['racket', 'racquet'], 'def': 'a sports implement used to strike a ball in various games', 'name': 'racket'}, {'frequency': 'r', 'id': 878, 'synset': 'radar.n.01', 'synonyms': ['radar'], 'def': 'measuring instrument in which the echo of a pulse of microwave radiation is used to detect and locate distant objects', 'name': 'radar'}, {'frequency': 'c', 'id': 879, 'synset': 'radiator.n.03', 'synonyms': ['radiator'], 'def': 'a mechanism consisting of a metal honeycomb through which hot fluids circulate', 'name': 'radiator'}, {'frequency': 'c', 'id': 880, 'synset': 'radio_receiver.n.01', 'synonyms': ['radio_receiver', 'radio_set', 'radio', 'tuner_(radio)'], 'def': 'an electronic receiver that detects and demodulates and amplifies transmitted radio signals', 'name': 'radio_receiver'}, {'frequency': 'c', 'id': 881, 'synset': 'radish.n.03', 'synonyms': ['radish', 'daikon'], 'def': 'pungent edible root of any of various cultivated radish plants', 'name': 'radish'}, {'frequency': 'c', 'id': 882, 'synset': 'raft.n.01', 'synonyms': ['raft'], 'def': 'a flat float (usually made of logs or planks) that can be used for transport or as a platform for swimmers', 'name': 'raft'}, {'frequency': 'r', 'id': 883, 'synset': 'rag_doll.n.01', 'synonyms': ['rag_doll'], 'def': 'a cloth doll that is stuffed and (usually) painted', 'name': 'rag_doll'}, {'frequency': 'c', 'id': 884, 'synset': 'raincoat.n.01', 'synonyms': ['raincoat', 'waterproof_jacket'], 'def': 'a water-resistant coat', 'name': 'raincoat'}, {'frequency': 'c', 'id': 885, 'synset': 'ram.n.05', 'synonyms': ['ram_(animal)'], 'def': 'uncastrated adult male sheep', 'name': 'ram_(animal)'}, {'frequency': 'c', 'id': 886, 'synset': 'raspberry.n.02', 'synonyms': ['raspberry'], 'def': 'red or black edible aggregate berries usually smaller than the related blackberries', 'name': 'raspberry'}, {'frequency': 'r', 'id': 887, 'synset': 'rat.n.01', 'synonyms': ['rat'], 'def': 'any of various long-tailed rodents similar to but larger than a mouse', 'name': 'rat'}, {'frequency': 'c', 'id': 888, 'synset': 'razorblade.n.01', 'synonyms': ['razorblade'], 'def': 'a blade that has very sharp edge', 'name': 'razorblade'}, {'frequency': 'c', 'id': 889, 'synset': 'reamer.n.01', 'synonyms': ['reamer_(juicer)', 'juicer', 'juice_reamer'], 'def': 'a squeezer with a conical ridged center that is used for squeezing juice from citrus fruit', 'name': 'reamer_(juicer)'}, {'frequency': 'f', 'id': 890, 'synset': 'rearview_mirror.n.01', 'synonyms': ['rearview_mirror'], 'def': 'car mirror that reflects the view out of the rear window', 'name': 'rearview_mirror'}, {'frequency': 'c', 'id': 891, 'synset': 'receipt.n.02', 'synonyms': ['receipt'], 'def': 'an acknowledgment (usually tangible) that payment has been made', 'name': 'receipt'}, {'frequency': 'c', 'id': 892, 'synset': 'recliner.n.01', 'synonyms': ['recliner', 'reclining_chair', 'lounger_(chair)'], 'def': 'an armchair whose back can be lowered and foot can be raised to allow the sitter to recline in it', 'name': 'recliner'}, {'frequency': 'r', 'id': 893, 'synset': 'record_player.n.01', 'synonyms': ['record_player', 'phonograph_(record_player)', 'turntable'], 'def': 'machine in which rotating records cause a stylus to vibrate and the vibrations are amplified acoustically or electronically', 'name': 'record_player'}, {'frequency': 'r', 'id': 894, 'synset': 'red_cabbage.n.02', 'synonyms': ['red_cabbage'], 'def': 'compact head of purplish-red leaves', 'name': 'red_cabbage'}, {'frequency': 'f', 'id': 895, 'synset': 'reflector.n.01', 'synonyms': ['reflector'], 'def': 'device that reflects light, radiation, etc.', 'name': 'reflector'}, {'frequency': 'f', 'id': 896, 'synset': 'remote_control.n.01', 'synonyms': ['remote_control'], 'def': 'a device that can be used to control a machine or apparatus from a distance', 'name': 'remote_control'}, {'frequency': 'c', 'id': 897, 'synset': 'rhinoceros.n.01', 'synonyms': ['rhinoceros'], 'def': 'massive powerful herbivorous odd-toed ungulate of southeast Asia and Africa having very thick skin and one or two horns on the snout', 'name': 'rhinoceros'}, {'frequency': 'r', 'id': 898, 'synset': 'rib.n.03', 'synonyms': ['rib_(food)'], 'def': 'cut of meat including one or more ribs', 'name': 'rib_(food)'}, {'frequency': 'r', 'id': 899, 'synset': 'rifle.n.01', 'synonyms': ['rifle'], 'def': 'a shoulder firearm with a long barrel', 'name': 'rifle'}, {'frequency': 'f', 'id': 900, 'synset': 'ring.n.08', 'synonyms': ['ring'], 'def': 'jewelry consisting of a circlet of precious metal (often set with jewels) worn on the finger', 'name': 'ring'}, {'frequency': 'r', 'id': 901, 'synset': 'river_boat.n.01', 'synonyms': ['river_boat'], 'def': 'a boat used on rivers or to ply a river', 'name': 'river_boat'}, {'frequency': 'r', 'id': 902, 'synset': 'road_map.n.02', 'synonyms': ['road_map'], 'def': '(NOT A ROAD) a MAP showing roads (for automobile travel)', 'name': 'road_map'}, {'frequency': 'c', 'id': 903, 'synset': 'robe.n.01', 'synonyms': ['robe'], 'def': 'any loose flowing garment', 'name': 'robe'}, {'frequency': 'c', 'id': 904, 'synset': 'rocking_chair.n.01', 'synonyms': ['rocking_chair'], 'def': 'a chair mounted on rockers', 'name': 'rocking_chair'}, {'frequency': 'r', 'id': 905, 'synset': 'roller_skate.n.01', 'synonyms': ['roller_skate'], 'def': 'a shoe with pairs of rollers (small hard wheels) fixed to the sole', 'name': 'roller_skate'}, {'frequency': 'r', 'id': 906, 'synset': 'rollerblade.n.01', 'synonyms': ['Rollerblade'], 'def': 'an in-line variant of a roller skate', 'name': 'Rollerblade'}, {'frequency': 'c', 'id': 907, 'synset': 'rolling_pin.n.01', 'synonyms': ['rolling_pin'], 'def': 'utensil consisting of a cylinder (usually of wood) with a handle at each end; used to roll out dough', 'name': 'rolling_pin'}, {'frequency': 'r', 'id': 908, 'synset': 'root_beer.n.01', 'synonyms': ['root_beer'], 'def': 'carbonated drink containing extracts of roots and herbs', 'name': 'root_beer'}, {'frequency': 'c', 'id': 909, 'synset': 'router.n.02', 'synonyms': ['router_(computer_equipment)'], 'def': 'a device that forwards data packets between computer networks', 'name': 'router_(computer_equipment)'}, {'frequency': 'f', 'id': 910, 'synset': 'rubber_band.n.01', 'synonyms': ['rubber_band', 'elastic_band'], 'def': 'a narrow band of elastic rubber used to hold things (such as papers) together', 'name': 'rubber_band'}, {'frequency': 'c', 'id': 911, 'synset': 'runner.n.08', 'synonyms': ['runner_(carpet)'], 'def': 'a long narrow carpet', 'name': 'runner_(carpet)'}, {'frequency': 'f', 'id': 912, 'synset': 'sack.n.01', 'synonyms': ['plastic_bag', 'paper_bag'], 'def': "a bag made of paper or plastic for holding customer's purchases", 'name': 'plastic_bag'}, {'frequency': 'f', 'id': 913, 'synset': 'saddle.n.01', 'synonyms': ['saddle_(on_an_animal)'], 'def': 'a seat for the rider of a horse or camel', 'name': 'saddle_(on_an_animal)'}, {'frequency': 'f', 'id': 914, 'synset': 'saddle_blanket.n.01', 'synonyms': ['saddle_blanket', 'saddlecloth', 'horse_blanket'], 'def': 'stable gear consisting of a blanket placed under the saddle', 'name': 'saddle_blanket'}, {'frequency': 'c', 'id': 915, 'synset': 'saddlebag.n.01', 'synonyms': ['saddlebag'], 'def': 'a large bag (or pair of bags) hung over a saddle', 'name': 'saddlebag'}, {'frequency': 'r', 'id': 916, 'synset': 'safety_pin.n.01', 'synonyms': ['safety_pin'], 'def': 'a pin in the form of a clasp; has a guard so the point of the pin will not stick the user', 'name': 'safety_pin'}, {'frequency': 'c', 'id': 917, 'synset': 'sail.n.01', 'synonyms': ['sail'], 'def': 'a large piece of fabric by means of which wind is used to propel a sailing vessel', 'name': 'sail'}, {'frequency': 'c', 'id': 918, 'synset': 'salad.n.01', 'synonyms': ['salad'], 'def': 'food mixtures either arranged on a plate or tossed and served with a moist dressing; usually consisting of or including greens', 'name': 'salad'}, {'frequency': 'r', 'id': 919, 'synset': 'salad_plate.n.01', 'synonyms': ['salad_plate', 'salad_bowl'], 'def': 'a plate or bowl for individual servings of salad', 'name': 'salad_plate'}, {'frequency': 'r', 'id': 920, 'synset': 'salami.n.01', 'synonyms': ['salami'], 'def': 'highly seasoned fatty sausage of pork and beef usually dried', 'name': 'salami'}, {'frequency': 'r', 'id': 921, 'synset': 'salmon.n.01', 'synonyms': ['salmon_(fish)'], 'def': 'any of various large food and game fishes of northern waters', 'name': 'salmon_(fish)'}, {'frequency': 'r', 'id': 922, 'synset': 'salmon.n.03', 'synonyms': ['salmon_(food)'], 'def': 'flesh of any of various marine or freshwater fish of the family Salmonidae', 'name': 'salmon_(food)'}, {'frequency': 'r', 'id': 923, 'synset': 'salsa.n.01', 'synonyms': ['salsa'], 'def': 'spicy sauce of tomatoes and onions and chili peppers to accompany Mexican foods', 'name': 'salsa'}, {'frequency': 'f', 'id': 924, 'synset': 'saltshaker.n.01', 'synonyms': ['saltshaker'], 'def': 'a shaker with a perforated top for sprinkling salt', 'name': 'saltshaker'}, {'frequency': 'f', 'id': 925, 'synset': 'sandal.n.01', 'synonyms': ['sandal_(type_of_shoe)'], 'def': 'a shoe consisting of a sole fastened by straps to the foot', 'name': 'sandal_(type_of_shoe)'}, {'frequency': 'f', 'id': 926, 'synset': 'sandwich.n.01', 'synonyms': ['sandwich'], 'def': 'two (or more) slices of bread with a filling between them', 'name': 'sandwich'}, {'frequency': 'r', 'id': 927, 'synset': 'satchel.n.01', 'synonyms': ['satchel'], 'def': 'luggage consisting of a small case with a flat bottom and (usually) a shoulder strap', 'name': 'satchel'}, {'frequency': 'r', 'id': 928, 'synset': 'saucepan.n.01', 'synonyms': ['saucepan'], 'def': 'a deep pan with a handle; used for stewing or boiling', 'name': 'saucepan'}, {'frequency': 'f', 'id': 929, 'synset': 'saucer.n.02', 'synonyms': ['saucer'], 'def': 'a small shallow dish for holding a cup at the table', 'name': 'saucer'}, {'frequency': 'f', 'id': 930, 'synset': 'sausage.n.01', 'synonyms': ['sausage'], 'def': 'highly seasoned minced meat stuffed in casings', 'name': 'sausage'}, {'frequency': 'r', 'id': 931, 'synset': 'sawhorse.n.01', 'synonyms': ['sawhorse', 'sawbuck'], 'def': 'a framework for holding wood that is being sawed', 'name': 'sawhorse'}, {'frequency': 'r', 'id': 932, 'synset': 'sax.n.02', 'synonyms': ['saxophone'], 'def': "a wind instrument with a `J'-shaped form typically made of brass", 'name': 'saxophone'}, {'frequency': 'f', 'id': 933, 'synset': 'scale.n.07', 'synonyms': ['scale_(measuring_instrument)'], 'def': 'a measuring instrument for weighing; shows amount of mass', 'name': 'scale_(measuring_instrument)'}, {'frequency': 'r', 'id': 934, 'synset': 'scarecrow.n.01', 'synonyms': ['scarecrow', 'strawman'], 'def': 'an effigy in the shape of a man to frighten birds away from seeds', 'name': 'scarecrow'}, {'frequency': 'f', 'id': 935, 'synset': 'scarf.n.01', 'synonyms': ['scarf'], 'def': 'a garment worn around the head or neck or shoulders for warmth or decoration', 'name': 'scarf'}, {'frequency': 'c', 'id': 936, 'synset': 'school_bus.n.01', 'synonyms': ['school_bus'], 'def': 'a bus used to transport children to or from school', 'name': 'school_bus'}, {'frequency': 'f', 'id': 937, 'synset': 'scissors.n.01', 'synonyms': ['scissors'], 'def': 'a tool having two crossed pivoting blades with looped handles', 'name': 'scissors'}, {'frequency': 'c', 'id': 938, 'synset': 'scoreboard.n.01', 'synonyms': ['scoreboard'], 'def': 'a large board for displaying the score of a contest (and some other information)', 'name': 'scoreboard'}, {'frequency': 'c', 'id': 939, 'synset': 'scrambled_eggs.n.01', 'synonyms': ['scrambled_eggs'], 'def': 'eggs beaten and cooked to a soft firm consistency while stirring', 'name': 'scrambled_eggs'}, {'frequency': 'r', 'id': 940, 'synset': 'scraper.n.01', 'synonyms': ['scraper'], 'def': 'any of various hand tools for scraping', 'name': 'scraper'}, {'frequency': 'r', 'id': 941, 'synset': 'scratcher.n.03', 'synonyms': ['scratcher'], 'def': 'a device used for scratching', 'name': 'scratcher'}, {'frequency': 'c', 'id': 942, 'synset': 'screwdriver.n.01', 'synonyms': ['screwdriver'], 'def': 'a hand tool for driving screws; has a tip that fits into the head of a screw', 'name': 'screwdriver'}, {'frequency': 'c', 'id': 943, 'synset': 'scrub_brush.n.01', 'synonyms': ['scrubbing_brush'], 'def': 'a brush with short stiff bristles for heavy cleaning', 'name': 'scrubbing_brush'}, {'frequency': 'c', 'id': 944, 'synset': 'sculpture.n.01', 'synonyms': ['sculpture'], 'def': 'a three-dimensional work of art', 'name': 'sculpture'}, {'frequency': 'r', 'id': 945, 'synset': 'seabird.n.01', 'synonyms': ['seabird', 'seafowl'], 'def': 'a bird that frequents coastal waters and the open ocean: gulls; pelicans; gannets; cormorants; albatrosses; petrels; etc.', 'name': 'seabird'}, {'frequency': 'r', 'id': 946, 'synset': 'seahorse.n.02', 'synonyms': ['seahorse'], 'def': 'small fish with horse-like heads bent sharply downward and curled tails', 'name': 'seahorse'}, {'frequency': 'r', 'id': 947, 'synset': 'seaplane.n.01', 'synonyms': ['seaplane', 'hydroplane'], 'def': 'an airplane that can land on or take off from water', 'name': 'seaplane'}, {'frequency': 'c', 'id': 948, 'synset': 'seashell.n.01', 'synonyms': ['seashell'], 'def': 'the shell of a marine organism', 'name': 'seashell'}, {'frequency': 'r', 'id': 949, 'synset': 'seedling.n.01', 'synonyms': ['seedling'], 'def': 'young plant or tree grown from a seed', 'name': 'seedling'}, {'frequency': 'c', 'id': 950, 'synset': 'serving_dish.n.01', 'synonyms': ['serving_dish'], 'def': 'a dish used for serving food', 'name': 'serving_dish'}, {'frequency': 'r', 'id': 951, 'synset': 'sewing_machine.n.01', 'synonyms': ['sewing_machine'], 'def': 'a textile machine used as a home appliance for sewing', 'name': 'sewing_machine'}, {'frequency': 'r', 'id': 952, 'synset': 'shaker.n.03', 'synonyms': ['shaker'], 'def': 'a container in which something can be shaken', 'name': 'shaker'}, {'frequency': 'c', 'id': 953, 'synset': 'shampoo.n.01', 'synonyms': ['shampoo'], 'def': 'cleansing agent consisting of soaps or detergents used for washing the hair', 'name': 'shampoo'}, {'frequency': 'r', 'id': 954, 'synset': 'shark.n.01', 'synonyms': ['shark'], 'def': 'typically large carnivorous fishes with sharpe teeth', 'name': 'shark'}, {'frequency': 'r', 'id': 955, 'synset': 'sharpener.n.01', 'synonyms': ['sharpener'], 'def': 'any implement that is used to make something (an edge or a point) sharper', 'name': 'sharpener'}, {'frequency': 'r', 'id': 956, 'synset': 'sharpie.n.03', 'synonyms': ['Sharpie'], 'def': 'a pen with indelible ink that will write on any surface', 'name': 'Sharpie'}, {'frequency': 'r', 'id': 957, 'synset': 'shaver.n.03', 'synonyms': ['shaver_(electric)', 'electric_shaver', 'electric_razor'], 'def': 'a razor powered by an electric motor', 'name': 'shaver_(electric)'}, {'frequency': 'c', 'id': 958, 'synset': 'shaving_cream.n.01', 'synonyms': ['shaving_cream', 'shaving_soap'], 'def': 'toiletry consisting that forms a rich lather for softening the beard before shaving', 'name': 'shaving_cream'}, {'frequency': 'r', 'id': 959, 'synset': 'shawl.n.01', 'synonyms': ['shawl'], 'def': 'cloak consisting of an oblong piece of cloth used to cover the head and shoulders', 'name': 'shawl'}, {'frequency': 'r', 'id': 960, 'synset': 'shears.n.01', 'synonyms': ['shears'], 'def': 'large scissors with strong blades', 'name': 'shears'}, {'frequency': 'f', 'id': 961, 'synset': 'sheep.n.01', 'synonyms': ['sheep'], 'def': 'woolly usually horned ruminant mammal related to the goat', 'name': 'sheep'}, {'frequency': 'r', 'id': 962, 'synset': 'shepherd_dog.n.01', 'synonyms': ['shepherd_dog', 'sheepdog'], 'def': 'any of various usually long-haired breeds of dog reared to herd and guard sheep', 'name': 'shepherd_dog'}, {'frequency': 'r', 'id': 963, 'synset': 'sherbert.n.01', 'synonyms': ['sherbert', 'sherbet'], 'def': 'a frozen dessert made primarily of fruit juice and sugar', 'name': 'sherbert'}, {'frequency': 'r', 'id': 964, 'synset': 'shield.n.02', 'synonyms': ['shield'], 'def': 'armor carried on the arm to intercept blows', 'name': 'shield'}, {'frequency': 'f', 'id': 965, 'synset': 'shirt.n.01', 'synonyms': ['shirt'], 'def': 'a garment worn on the upper half of the body', 'name': 'shirt'}, {'frequency': 'f', 'id': 966, 'synset': 'shoe.n.01', 'synonyms': ['shoe', 'sneaker_(type_of_shoe)', 'tennis_shoe'], 'def': 'common footwear covering the foot', 'name': 'shoe'}, {'frequency': 'c', 'id': 967, 'synset': 'shopping_bag.n.01', 'synonyms': ['shopping_bag'], 'def': 'a bag made of plastic or strong paper (often with handles); used to transport goods after shopping', 'name': 'shopping_bag'}, {'frequency': 'c', 'id': 968, 'synset': 'shopping_cart.n.01', 'synonyms': ['shopping_cart'], 'def': 'a handcart that holds groceries or other goods while shopping', 'name': 'shopping_cart'}, {'frequency': 'f', 'id': 969, 'synset': 'short_pants.n.01', 'synonyms': ['short_pants', 'shorts_(clothing)', 'trunks_(clothing)'], 'def': 'trousers that end at or above the knee', 'name': 'short_pants'}, {'frequency': 'r', 'id': 970, 'synset': 'shot_glass.n.01', 'synonyms': ['shot_glass'], 'def': 'a small glass adequate to hold a single swallow of whiskey', 'name': 'shot_glass'}, {'frequency': 'c', 'id': 971, 'synset': 'shoulder_bag.n.01', 'synonyms': ['shoulder_bag'], 'def': 'a large handbag that can be carried by a strap looped over the shoulder', 'name': 'shoulder_bag'}, {'frequency': 'c', 'id': 972, 'synset': 'shovel.n.01', 'synonyms': ['shovel'], 'def': 'a hand tool for lifting loose material such as snow, dirt, etc.', 'name': 'shovel'}, {'frequency': 'f', 'id': 973, 'synset': 'shower.n.01', 'synonyms': ['shower_head'], 'def': 'a plumbing fixture that sprays water over you', 'name': 'shower_head'}, {'frequency': 'f', 'id': 974, 'synset': 'shower_curtain.n.01', 'synonyms': ['shower_curtain'], 'def': 'a curtain that keeps water from splashing out of the shower area', 'name': 'shower_curtain'}, {'frequency': 'r', 'id': 975, 'synset': 'shredder.n.01', 'synonyms': ['shredder_(for_paper)'], 'def': 'a device that shreds documents', 'name': 'shredder_(for_paper)'}, {'frequency': 'r', 'id': 976, 'synset': 'sieve.n.01', 'synonyms': ['sieve', 'screen_(sieve)'], 'def': 'a strainer for separating lumps from powdered material or grading particles', 'name': 'sieve'}, {'frequency': 'f', 'id': 977, 'synset': 'signboard.n.01', 'synonyms': ['signboard'], 'def': 'structure displaying a board on which advertisements can be posted', 'name': 'signboard'}, {'frequency': 'c', 'id': 978, 'synset': 'silo.n.01', 'synonyms': ['silo'], 'def': 'a cylindrical tower used for storing goods', 'name': 'silo'}, {'frequency': 'f', 'id': 979, 'synset': 'sink.n.01', 'synonyms': ['sink'], 'def': 'plumbing fixture consisting of a water basin fixed to a wall or floor and having a drainpipe', 'name': 'sink'}, {'frequency': 'f', 'id': 980, 'synset': 'skateboard.n.01', 'synonyms': ['skateboard'], 'def': 'a board with wheels that is ridden in a standing or crouching position and propelled by foot', 'name': 'skateboard'}, {'frequency': 'c', 'id': 981, 'synset': 'skewer.n.01', 'synonyms': ['skewer'], 'def': 'a long pin for holding meat in position while it is being roasted', 'name': 'skewer'}, {'frequency': 'f', 'id': 982, 'synset': 'ski.n.01', 'synonyms': ['ski'], 'def': 'sports equipment for skiing on snow', 'name': 'ski'}, {'frequency': 'f', 'id': 983, 'synset': 'ski_boot.n.01', 'synonyms': ['ski_boot'], 'def': 'a stiff boot that is fastened to a ski with a ski binding', 'name': 'ski_boot'}, {'frequency': 'f', 'id': 984, 'synset': 'ski_parka.n.01', 'synonyms': ['ski_parka', 'ski_jacket'], 'def': 'a parka to be worn while skiing', 'name': 'ski_parka'}, {'frequency': 'f', 'id': 985, 'synset': 'ski_pole.n.01', 'synonyms': ['ski_pole'], 'def': 'a pole with metal points used as an aid in skiing', 'name': 'ski_pole'}, {'frequency': 'f', 'id': 986, 'synset': 'skirt.n.02', 'synonyms': ['skirt'], 'def': 'a garment hanging from the waist; worn mainly by girls and women', 'name': 'skirt'}, {'frequency': 'c', 'id': 987, 'synset': 'sled.n.01', 'synonyms': ['sled', 'sledge', 'sleigh'], 'def': 'a vehicle or flat object for transportation over snow by sliding or pulled by dogs, etc.', 'name': 'sled'}, {'frequency': 'c', 'id': 988, 'synset': 'sleeping_bag.n.01', 'synonyms': ['sleeping_bag'], 'def': 'large padded bag designed to be slept in outdoors', 'name': 'sleeping_bag'}, {'frequency': 'r', 'id': 989, 'synset': 'sling.n.05', 'synonyms': ['sling_(bandage)', 'triangular_bandage'], 'def': 'bandage to support an injured forearm; slung over the shoulder or neck', 'name': 'sling_(bandage)'}, {'frequency': 'c', 'id': 990, 'synset': 'slipper.n.01', 'synonyms': ['slipper_(footwear)', 'carpet_slipper_(footwear)'], 'def': 'low footwear that can be slipped on and off easily; usually worn indoors', 'name': 'slipper_(footwear)'}, {'frequency': 'r', 'id': 991, 'synset': 'smoothie.n.02', 'synonyms': ['smoothie'], 'def': 'a thick smooth drink consisting of fresh fruit pureed with ice cream or yoghurt or milk', 'name': 'smoothie'}, {'frequency': 'r', 'id': 992, 'synset': 'snake.n.01', 'synonyms': ['snake', 'serpent'], 'def': 'limbless scaly elongate reptile; some are venomous', 'name': 'snake'}, {'frequency': 'f', 'id': 993, 'synset': 'snowboard.n.01', 'synonyms': ['snowboard'], 'def': 'a board that resembles a broad ski or a small surfboard; used in a standing position to slide down snow-covered slopes', 'name': 'snowboard'}, {'frequency': 'c', 'id': 994, 'synset': 'snowman.n.01', 'synonyms': ['snowman'], 'def': 'a figure of a person made of packed snow', 'name': 'snowman'}, {'frequency': 'c', 'id': 995, 'synset': 'snowmobile.n.01', 'synonyms': ['snowmobile'], 'def': 'tracked vehicle for travel on snow having skis in front', 'name': 'snowmobile'}, {'frequency': 'f', 'id': 996, 'synset': 'soap.n.01', 'synonyms': ['soap'], 'def': 'a cleansing agent made from the salts of vegetable or animal fats', 'name': 'soap'}, {'frequency': 'f', 'id': 997, 'synset': 'soccer_ball.n.01', 'synonyms': ['soccer_ball'], 'def': "an inflated ball used in playing soccer (called `football' outside of the United States)", 'name': 'soccer_ball'}, {'frequency': 'f', 'id': 998, 'synset': 'sock.n.01', 'synonyms': ['sock'], 'def': 'cloth covering for the foot; worn inside the shoe; reaches to between the ankle and the knee', 'name': 'sock'}, {'frequency': 'r', 'id': 999, 'synset': 'soda_fountain.n.02', 'synonyms': ['soda_fountain'], 'def': 'an apparatus for dispensing soda water', 'name': 'soda_fountain'}, {'frequency': 'r', 'id': 1000, 'synset': 'soda_water.n.01', 'synonyms': ['carbonated_water', 'club_soda', 'seltzer', 'sparkling_water'], 'def': 'effervescent beverage artificially charged with carbon dioxide', 'name': 'carbonated_water'}, {'frequency': 'f', 'id': 1001, 'synset': 'sofa.n.01', 'synonyms': ['sofa', 'couch', 'lounge'], 'def': 'an upholstered seat for more than one person', 'name': 'sofa'}, {'frequency': 'r', 'id': 1002, 'synset': 'softball.n.01', 'synonyms': ['softball'], 'def': 'ball used in playing softball', 'name': 'softball'}, {'frequency': 'c', 'id': 1003, 'synset': 'solar_array.n.01', 'synonyms': ['solar_array', 'solar_battery', 'solar_panel'], 'def': 'electrical device consisting of a large array of connected solar cells', 'name': 'solar_array'}, {'frequency': 'r', 'id': 1004, 'synset': 'sombrero.n.02', 'synonyms': ['sombrero'], 'def': 'a straw hat with a tall crown and broad brim; worn in American southwest and in Mexico', 'name': 'sombrero'}, {'frequency': 'c', 'id': 1005, 'synset': 'soup.n.01', 'synonyms': ['soup'], 'def': 'liquid food especially of meat or fish or vegetable stock often containing pieces of solid food', 'name': 'soup'}, {'frequency': 'r', 'id': 1006, 'synset': 'soup_bowl.n.01', 'synonyms': ['soup_bowl'], 'def': 'a bowl for serving soup', 'name': 'soup_bowl'}, {'frequency': 'c', 'id': 1007, 'synset': 'soupspoon.n.01', 'synonyms': ['soupspoon'], 'def': 'a spoon with a rounded bowl for eating soup', 'name': 'soupspoon'}, {'frequency': 'c', 'id': 1008, 'synset': 'sour_cream.n.01', 'synonyms': ['sour_cream', 'soured_cream'], 'def': 'soured light cream', 'name': 'sour_cream'}, {'frequency': 'r', 'id': 1009, 'synset': 'soya_milk.n.01', 'synonyms': ['soya_milk', 'soybean_milk', 'soymilk'], 'def': 'a milk substitute containing soybean flour and water; used in some infant formulas and in making tofu', 'name': 'soya_milk'}, {'frequency': 'r', 'id': 1010, 'synset': 'space_shuttle.n.01', 'synonyms': ['space_shuttle'], 'def': "a reusable spacecraft with wings for a controlled descent through the Earth's atmosphere", 'name': 'space_shuttle'}, {'frequency': 'r', 'id': 1011, 'synset': 'sparkler.n.02', 'synonyms': ['sparkler_(fireworks)'], 'def': 'a firework that burns slowly and throws out a shower of sparks', 'name': 'sparkler_(fireworks)'}, {'frequency': 'f', 'id': 1012, 'synset': 'spatula.n.02', 'synonyms': ['spatula'], 'def': 'a hand tool with a thin flexible blade used to mix or spread soft substances', 'name': 'spatula'}, {'frequency': 'r', 'id': 1013, 'synset': 'spear.n.01', 'synonyms': ['spear', 'lance'], 'def': 'a long pointed rod used as a tool or weapon', 'name': 'spear'}, {'frequency': 'f', 'id': 1014, 'synset': 'spectacles.n.01', 'synonyms': ['spectacles', 'specs', 'eyeglasses', 'glasses'], 'def': 'optical instrument consisting of a frame that holds a pair of lenses for correcting defective vision', 'name': 'spectacles'}, {'frequency': 'c', 'id': 1015, 'synset': 'spice_rack.n.01', 'synonyms': ['spice_rack'], 'def': 'a rack for displaying containers filled with spices', 'name': 'spice_rack'}, {'frequency': 'r', 'id': 1016, 'synset': 'spider.n.01', 'synonyms': ['spider'], 'def': 'predatory arachnid with eight legs, two poison fangs, two feelers, and usually two silk-spinning organs at the back end of the body', 'name': 'spider'}, {'frequency': 'c', 'id': 1017, 'synset': 'sponge.n.01', 'synonyms': ['sponge'], 'def': 'a porous mass usable to absorb water typically used for cleaning', 'name': 'sponge'}, {'frequency': 'f', 'id': 1018, 'synset': 'spoon.n.01', 'synonyms': ['spoon'], 'def': 'a piece of cutlery with a shallow bowl-shaped container and a handle', 'name': 'spoon'}, {'frequency': 'c', 'id': 1019, 'synset': 'sportswear.n.01', 'synonyms': ['sportswear', 'athletic_wear', 'activewear'], 'def': 'attire worn for sport or for casual wear', 'name': 'sportswear'}, {'frequency': 'c', 'id': 1020, 'synset': 'spotlight.n.02', 'synonyms': ['spotlight'], 'def': 'a lamp that produces a strong beam of light to illuminate a restricted area; used to focus attention of a stage performer', 'name': 'spotlight'}, {'frequency': 'r', 'id': 1021, 'synset': 'squirrel.n.01', 'synonyms': ['squirrel'], 'def': 'a kind of arboreal rodent having a long bushy tail', 'name': 'squirrel'}, {'frequency': 'c', 'id': 1022, 'synset': 'stapler.n.01', 'synonyms': ['stapler_(stapling_machine)'], 'def': 'a machine that inserts staples into sheets of paper in order to fasten them together', 'name': 'stapler_(stapling_machine)'}, {'frequency': 'r', 'id': 1023, 'synset': 'starfish.n.01', 'synonyms': ['starfish', 'sea_star'], 'def': 'echinoderms characterized by five arms extending from a central disk', 'name': 'starfish'}, {'frequency': 'f', 'id': 1024, 'synset': 'statue.n.01', 'synonyms': ['statue_(sculpture)'], 'def': 'a sculpture representing a human or animal', 'name': 'statue_(sculpture)'}, {'frequency': 'c', 'id': 1025, 'synset': 'steak.n.01', 'synonyms': ['steak_(food)'], 'def': 'a slice of meat cut from the fleshy part of an animal or large fish', 'name': 'steak_(food)'}, {'frequency': 'r', 'id': 1026, 'synset': 'steak_knife.n.01', 'synonyms': ['steak_knife'], 'def': 'a sharp table knife used in eating steak', 'name': 'steak_knife'}, {'frequency': 'r', 'id': 1027, 'synset': 'steamer.n.02', 'synonyms': ['steamer_(kitchen_appliance)'], 'def': 'a cooking utensil that can be used to cook food by steaming it', 'name': 'steamer_(kitchen_appliance)'}, {'frequency': 'f', 'id': 1028, 'synset': 'steering_wheel.n.01', 'synonyms': ['steering_wheel'], 'def': 'a handwheel that is used for steering', 'name': 'steering_wheel'}, {'frequency': 'r', 'id': 1029, 'synset': 'stencil.n.01', 'synonyms': ['stencil'], 'def': 'a sheet of material (metal, plastic, etc.) that has been perforated with a pattern; ink or paint can pass through the perforations to create the printed pattern on the surface below', 'name': 'stencil'}, {'frequency': 'r', 'id': 1030, 'synset': 'step_ladder.n.01', 'synonyms': ['stepladder'], 'def': 'a folding portable ladder hinged at the top', 'name': 'stepladder'}, {'frequency': 'c', 'id': 1031, 'synset': 'step_stool.n.01', 'synonyms': ['step_stool'], 'def': 'a stool that has one or two steps that fold under the seat', 'name': 'step_stool'}, {'frequency': 'c', 'id': 1032, 'synset': 'stereo.n.01', 'synonyms': ['stereo_(sound_system)'], 'def': 'electronic device for playing audio', 'name': 'stereo_(sound_system)'}, {'frequency': 'r', 'id': 1033, 'synset': 'stew.n.02', 'synonyms': ['stew'], 'def': 'food prepared by stewing especially meat or fish with vegetables', 'name': 'stew'}, {'frequency': 'r', 'id': 1034, 'synset': 'stirrer.n.02', 'synonyms': ['stirrer'], 'def': 'an implement used for stirring', 'name': 'stirrer'}, {'frequency': 'f', 'id': 1035, 'synset': 'stirrup.n.01', 'synonyms': ['stirrup'], 'def': "support consisting of metal loops into which rider's feet go", 'name': 'stirrup'}, {'frequency': 'c', 'id': 1036, 'synset': 'stocking.n.01', 'synonyms': ['stockings_(leg_wear)'], 'def': 'close-fitting hosiery to cover the foot and leg; come in matched pairs', 'name': 'stockings_(leg_wear)'}, {'frequency': 'f', 'id': 1037, 'synset': 'stool.n.01', 'synonyms': ['stool'], 'def': 'a simple seat without a back or arms', 'name': 'stool'}, {'frequency': 'f', 'id': 1038, 'synset': 'stop_sign.n.01', 'synonyms': ['stop_sign'], 'def': 'a traffic sign to notify drivers that they must come to a complete stop', 'name': 'stop_sign'}, {'frequency': 'f', 'id': 1039, 'synset': 'stoplight.n.01', 'synonyms': ['brake_light'], 'def': 'a red light on the rear of a motor vehicle that signals when the brakes are applied', 'name': 'brake_light'}, {'frequency': 'f', 'id': 1040, 'synset': 'stove.n.01', 'synonyms': ['stove', 'kitchen_stove', 'range_(kitchen_appliance)', 'kitchen_range', 'cooking_stove'], 'def': 'a kitchen appliance used for cooking food', 'name': 'stove'}, {'frequency': 'c', 'id': 1041, 'synset': 'strainer.n.01', 'synonyms': ['strainer'], 'def': 'a filter to retain larger pieces while smaller pieces and liquids pass through', 'name': 'strainer'}, {'frequency': 'f', 'id': 1042, 'synset': 'strap.n.01', 'synonyms': ['strap'], 'def': 'an elongated strip of material for binding things together or holding', 'name': 'strap'}, {'frequency': 'f', 'id': 1043, 'synset': 'straw.n.04', 'synonyms': ['straw_(for_drinking)', 'drinking_straw'], 'def': 'a thin paper or plastic tube used to suck liquids into the mouth', 'name': 'straw_(for_drinking)'}, {'frequency': 'f', 'id': 1044, 'synset': 'strawberry.n.01', 'synonyms': ['strawberry'], 'def': 'sweet fleshy red fruit', 'name': 'strawberry'}, {'frequency': 'f', 'id': 1045, 'synset': 'street_sign.n.01', 'synonyms': ['street_sign'], 'def': 'a sign visible from the street', 'name': 'street_sign'}, {'frequency': 'f', 'id': 1046, 'synset': 'streetlight.n.01', 'synonyms': ['streetlight', 'street_lamp'], 'def': 'a lamp supported on a lamppost; for illuminating a street', 'name': 'streetlight'}, {'frequency': 'r', 'id': 1047, 'synset': 'string_cheese.n.01', 'synonyms': ['string_cheese'], 'def': 'cheese formed in long strings twisted together', 'name': 'string_cheese'}, {'frequency': 'r', 'id': 1048, 'synset': 'stylus.n.02', 'synonyms': ['stylus'], 'def': 'a pointed tool for writing or drawing or engraving', 'name': 'stylus'}, {'frequency': 'r', 'id': 1049, 'synset': 'subwoofer.n.01', 'synonyms': ['subwoofer'], 'def': 'a loudspeaker that is designed to reproduce very low bass frequencies', 'name': 'subwoofer'}, {'frequency': 'r', 'id': 1050, 'synset': 'sugar_bowl.n.01', 'synonyms': ['sugar_bowl'], 'def': 'a dish in which sugar is served', 'name': 'sugar_bowl'}, {'frequency': 'r', 'id': 1051, 'synset': 'sugarcane.n.01', 'synonyms': ['sugarcane_(plant)'], 'def': 'juicy canes whose sap is a source of molasses and commercial sugar; fresh canes are sometimes chewed for the juice', 'name': 'sugarcane_(plant)'}, {'frequency': 'c', 'id': 1052, 'synset': 'suit.n.01', 'synonyms': ['suit_(clothing)'], 'def': 'a set of garments (usually including a jacket and trousers or skirt) for outerwear all of the same fabric and color', 'name': 'suit_(clothing)'}, {'frequency': 'c', 'id': 1053, 'synset': 'sunflower.n.01', 'synonyms': ['sunflower'], 'def': 'any plant of the genus Helianthus having large flower heads with dark disk florets and showy yellow rays', 'name': 'sunflower'}, {'frequency': 'f', 'id': 1054, 'synset': 'sunglasses.n.01', 'synonyms': ['sunglasses'], 'def': 'spectacles that are darkened or polarized to protect the eyes from the glare of the sun', 'name': 'sunglasses'}, {'frequency': 'c', 'id': 1055, 'synset': 'sunhat.n.01', 'synonyms': ['sunhat'], 'def': 'a hat with a broad brim that protects the face from direct exposure to the sun', 'name': 'sunhat'}, {'frequency': 'r', 'id': 1056, 'synset': 'sunscreen.n.01', 'synonyms': ['sunscreen', 'sunblock'], 'def': 'a cream spread on the skin; contains a chemical to filter out ultraviolet light and so protect from sunburn', 'name': 'sunscreen'}, {'frequency': 'f', 'id': 1057, 'synset': 'surfboard.n.01', 'synonyms': ['surfboard'], 'def': 'a narrow buoyant board for riding surf', 'name': 'surfboard'}, {'frequency': 'c', 'id': 1058, 'synset': 'sushi.n.01', 'synonyms': ['sushi'], 'def': 'rice (with raw fish) wrapped in seaweed', 'name': 'sushi'}, {'frequency': 'c', 'id': 1059, 'synset': 'swab.n.02', 'synonyms': ['mop'], 'def': 'cleaning implement consisting of absorbent material fastened to a handle; for cleaning floors', 'name': 'mop'}, {'frequency': 'c', 'id': 1060, 'synset': 'sweat_pants.n.01', 'synonyms': ['sweat_pants'], 'def': 'loose-fitting trousers with elastic cuffs; worn by athletes', 'name': 'sweat_pants'}, {'frequency': 'c', 'id': 1061, 'synset': 'sweatband.n.02', 'synonyms': ['sweatband'], 'def': 'a band of material tied around the forehead or wrist to absorb sweat', 'name': 'sweatband'}, {'frequency': 'f', 'id': 1062, 'synset': 'sweater.n.01', 'synonyms': ['sweater'], 'def': 'a crocheted or knitted garment covering the upper part of the body', 'name': 'sweater'}, {'frequency': 'f', 'id': 1063, 'synset': 'sweatshirt.n.01', 'synonyms': ['sweatshirt'], 'def': 'cotton knit pullover with long sleeves worn during athletic activity', 'name': 'sweatshirt'}, {'frequency': 'c', 'id': 1064, 'synset': 'sweet_potato.n.02', 'synonyms': ['sweet_potato'], 'def': 'the edible tuberous root of the sweet potato vine', 'name': 'sweet_potato'}, {'frequency': 'f', 'id': 1065, 'synset': 'swimsuit.n.01', 'synonyms': ['swimsuit', 'swimwear', 'bathing_suit', 'swimming_costume', 'bathing_costume', 'swimming_trunks', 'bathing_trunks'], 'def': 'garment worn for swimming', 'name': 'swimsuit'}, {'frequency': 'c', 'id': 1066, 'synset': 'sword.n.01', 'synonyms': ['sword'], 'def': 'a cutting or thrusting weapon that has a long metal blade', 'name': 'sword'}, {'frequency': 'r', 'id': 1067, 'synset': 'syringe.n.01', 'synonyms': ['syringe'], 'def': 'a medical instrument used to inject or withdraw fluids', 'name': 'syringe'}, {'frequency': 'r', 'id': 1068, 'synset': 'tabasco.n.02', 'synonyms': ['Tabasco_sauce'], 'def': 'very spicy sauce (trade name Tabasco) made from fully-aged red peppers', 'name': 'Tabasco_sauce'}, {'frequency': 'r', 'id': 1069, 'synset': 'table-tennis_table.n.01', 'synonyms': ['table-tennis_table', 'ping-pong_table'], 'def': 'a table used for playing table tennis', 'name': 'table-tennis_table'}, {'frequency': 'f', 'id': 1070, 'synset': 'table.n.02', 'synonyms': ['table'], 'def': 'a piece of furniture having a smooth flat top that is usually supported by one or more vertical legs', 'name': 'table'}, {'frequency': 'c', 'id': 1071, 'synset': 'table_lamp.n.01', 'synonyms': ['table_lamp'], 'def': 'a lamp that sits on a table', 'name': 'table_lamp'}, {'frequency': 'f', 'id': 1072, 'synset': 'tablecloth.n.01', 'synonyms': ['tablecloth'], 'def': 'a covering spread over a dining table', 'name': 'tablecloth'}, {'frequency': 'r', 'id': 1073, 'synset': 'tachometer.n.01', 'synonyms': ['tachometer'], 'def': 'measuring instrument for indicating speed of rotation', 'name': 'tachometer'}, {'frequency': 'r', 'id': 1074, 'synset': 'taco.n.02', 'synonyms': ['taco'], 'def': 'a small tortilla cupped around a filling', 'name': 'taco'}, {'frequency': 'f', 'id': 1075, 'synset': 'tag.n.02', 'synonyms': ['tag'], 'def': 'a label associated with something for the purpose of identification or information', 'name': 'tag'}, {'frequency': 'f', 'id': 1076, 'synset': 'taillight.n.01', 'synonyms': ['taillight', 'rear_light'], 'def': 'lamp (usually red) mounted at the rear of a motor vehicle', 'name': 'taillight'}, {'frequency': 'r', 'id': 1077, 'synset': 'tambourine.n.01', 'synonyms': ['tambourine'], 'def': 'a shallow drum with a single drumhead and with metallic disks in the sides', 'name': 'tambourine'}, {'frequency': 'r', 'id': 1078, 'synset': 'tank.n.01', 'synonyms': ['army_tank', 'armored_combat_vehicle', 'armoured_combat_vehicle'], 'def': 'an enclosed armored military vehicle; has a cannon and moves on caterpillar treads', 'name': 'army_tank'}, {'frequency': 'c', 'id': 1079, 'synset': 'tank.n.02', 'synonyms': ['tank_(storage_vessel)', 'storage_tank'], 'def': 'a large (usually metallic) vessel for holding gases or liquids', 'name': 'tank_(storage_vessel)'}, {'frequency': 'f', 'id': 1080, 'synset': 'tank_top.n.01', 'synonyms': ['tank_top_(clothing)'], 'def': 'a tight-fitting sleeveless shirt with wide shoulder straps and low neck and no front opening', 'name': 'tank_top_(clothing)'}, {'frequency': 'c', 'id': 1081, 'synset': 'tape.n.01', 'synonyms': ['tape_(sticky_cloth_or_paper)'], 'def': 'a long thin piece of cloth or paper as used for binding or fastening', 'name': 'tape_(sticky_cloth_or_paper)'}, {'frequency': 'c', 'id': 1082, 'synset': 'tape.n.04', 'synonyms': ['tape_measure', 'measuring_tape'], 'def': 'measuring instrument consisting of a narrow strip (cloth or metal) marked in inches or centimeters and used for measuring lengths', 'name': 'tape_measure'}, {'frequency': 'c', 'id': 1083, 'synset': 'tapestry.n.02', 'synonyms': ['tapestry'], 'def': 'a heavy textile with a woven design; used for curtains and upholstery', 'name': 'tapestry'}, {'frequency': 'f', 'id': 1084, 'synset': 'tarpaulin.n.01', 'synonyms': ['tarp'], 'def': 'waterproofed canvas', 'name': 'tarp'}, {'frequency': 'c', 'id': 1085, 'synset': 'tartan.n.01', 'synonyms': ['tartan', 'plaid'], 'def': 'a cloth having a crisscross design', 'name': 'tartan'}, {'frequency': 'c', 'id': 1086, 'synset': 'tassel.n.01', 'synonyms': ['tassel'], 'def': 'adornment consisting of a bunch of cords fastened at one end', 'name': 'tassel'}, {'frequency': 'r', 'id': 1087, 'synset': 'tea_bag.n.01', 'synonyms': ['tea_bag'], 'def': 'a measured amount of tea in a bag for an individual serving of tea', 'name': 'tea_bag'}, {'frequency': 'c', 'id': 1088, 'synset': 'teacup.n.02', 'synonyms': ['teacup'], 'def': 'a cup from which tea is drunk', 'name': 'teacup'}, {'frequency': 'c', 'id': 1089, 'synset': 'teakettle.n.01', 'synonyms': ['teakettle'], 'def': 'kettle for boiling water to make tea', 'name': 'teakettle'}, {'frequency': 'c', 'id': 1090, 'synset': 'teapot.n.01', 'synonyms': ['teapot'], 'def': 'pot for brewing tea; usually has a spout and handle', 'name': 'teapot'}, {'frequency': 'f', 'id': 1091, 'synset': 'teddy.n.01', 'synonyms': ['teddy_bear'], 'def': "plaything consisting of a child's toy bear (usually plush and stuffed with soft materials)", 'name': 'teddy_bear'}, {'frequency': 'f', 'id': 1092, 'synset': 'telephone.n.01', 'synonyms': ['telephone', 'phone', 'telephone_set'], 'def': 'electronic device for communicating by voice over long distances', 'name': 'telephone'}, {'frequency': 'c', 'id': 1093, 'synset': 'telephone_booth.n.01', 'synonyms': ['telephone_booth', 'phone_booth', 'call_box', 'telephone_box', 'telephone_kiosk'], 'def': 'booth for using a telephone', 'name': 'telephone_booth'}, {'frequency': 'f', 'id': 1094, 'synset': 'telephone_pole.n.01', 'synonyms': ['telephone_pole', 'telegraph_pole', 'telegraph_post'], 'def': 'tall pole supporting telephone wires', 'name': 'telephone_pole'}, {'frequency': 'r', 'id': 1095, 'synset': 'telephoto_lens.n.01', 'synonyms': ['telephoto_lens', 'zoom_lens'], 'def': 'a camera lens that magnifies the image', 'name': 'telephoto_lens'}, {'frequency': 'c', 'id': 1096, 'synset': 'television_camera.n.01', 'synonyms': ['television_camera', 'tv_camera'], 'def': 'television equipment for capturing and recording video', 'name': 'television_camera'}, {'frequency': 'f', 'id': 1097, 'synset': 'television_receiver.n.01', 'synonyms': ['television_set', 'tv', 'tv_set'], 'def': 'an electronic device that receives television signals and displays them on a screen', 'name': 'television_set'}, {'frequency': 'f', 'id': 1098, 'synset': 'tennis_ball.n.01', 'synonyms': ['tennis_ball'], 'def': 'ball about the size of a fist used in playing tennis', 'name': 'tennis_ball'}, {'frequency': 'f', 'id': 1099, 'synset': 'tennis_racket.n.01', 'synonyms': ['tennis_racket'], 'def': 'a racket used to play tennis', 'name': 'tennis_racket'}, {'frequency': 'r', 'id': 1100, 'synset': 'tequila.n.01', 'synonyms': ['tequila'], 'def': 'Mexican liquor made from fermented juices of an agave plant', 'name': 'tequila'}, {'frequency': 'c', 'id': 1101, 'synset': 'thermometer.n.01', 'synonyms': ['thermometer'], 'def': 'measuring instrument for measuring temperature', 'name': 'thermometer'}, {'frequency': 'c', 'id': 1102, 'synset': 'thermos.n.01', 'synonyms': ['thermos_bottle'], 'def': 'vacuum flask that preserves temperature of hot or cold drinks', 'name': 'thermos_bottle'}, {'frequency': 'c', 'id': 1103, 'synset': 'thermostat.n.01', 'synonyms': ['thermostat'], 'def': 'a regulator for automatically regulating temperature by starting or stopping the supply of heat', 'name': 'thermostat'}, {'frequency': 'r', 'id': 1104, 'synset': 'thimble.n.02', 'synonyms': ['thimble'], 'def': 'a small metal cap to protect the finger while sewing; can be used as a small container', 'name': 'thimble'}, {'frequency': 'c', 'id': 1105, 'synset': 'thread.n.01', 'synonyms': ['thread', 'yarn'], 'def': 'a fine cord of twisted fibers (of cotton or silk or wool or nylon etc.) used in sewing and weaving', 'name': 'thread'}, {'frequency': 'c', 'id': 1106, 'synset': 'thumbtack.n.01', 'synonyms': ['thumbtack', 'drawing_pin', 'pushpin'], 'def': 'a tack for attaching papers to a bulletin board or drawing board', 'name': 'thumbtack'}, {'frequency': 'c', 'id': 1107, 'synset': 'tiara.n.01', 'synonyms': ['tiara'], 'def': 'a jeweled headdress worn by women on formal occasions', 'name': 'tiara'}, {'frequency': 'c', 'id': 1108, 'synset': 'tiger.n.02', 'synonyms': ['tiger'], 'def': 'large feline of forests in most of Asia having a tawny coat with black stripes', 'name': 'tiger'}, {'frequency': 'c', 'id': 1109, 'synset': 'tights.n.01', 'synonyms': ['tights_(clothing)', 'leotards'], 'def': 'skintight knit hose covering the body from the waist to the feet worn by acrobats and dancers and as stockings by women and girls', 'name': 'tights_(clothing)'}, {'frequency': 'c', 'id': 1110, 'synset': 'timer.n.01', 'synonyms': ['timer', 'stopwatch'], 'def': 'a timepiece that measures a time interval and signals its end', 'name': 'timer'}, {'frequency': 'f', 'id': 1111, 'synset': 'tinfoil.n.01', 'synonyms': ['tinfoil'], 'def': 'foil made of tin or an alloy of tin and lead', 'name': 'tinfoil'}, {'frequency': 'r', 'id': 1112, 'synset': 'tinsel.n.01', 'synonyms': ['tinsel'], 'def': 'a showy decoration that is basically valueless', 'name': 'tinsel'}, {'frequency': 'f', 'id': 1113, 'synset': 'tissue.n.02', 'synonyms': ['tissue_paper'], 'def': 'a soft thin (usually translucent) paper', 'name': 'tissue_paper'}, {'frequency': 'c', 'id': 1114, 'synset': 'toast.n.01', 'synonyms': ['toast_(food)'], 'def': 'slice of bread that has been toasted', 'name': 'toast_(food)'}, {'frequency': 'f', 'id': 1115, 'synset': 'toaster.n.02', 'synonyms': ['toaster'], 'def': 'a kitchen appliance (usually electric) for toasting bread', 'name': 'toaster'}, {'frequency': 'c', 'id': 1116, 'synset': 'toaster_oven.n.01', 'synonyms': ['toaster_oven'], 'def': 'kitchen appliance consisting of a small electric oven for toasting or warming food', 'name': 'toaster_oven'}, {'frequency': 'f', 'id': 1117, 'synset': 'toilet.n.02', 'synonyms': ['toilet'], 'def': 'a plumbing fixture for defecation and urination', 'name': 'toilet'}, {'frequency': 'f', 'id': 1118, 'synset': 'toilet_tissue.n.01', 'synonyms': ['toilet_tissue', 'toilet_paper', 'bathroom_tissue'], 'def': 'a soft thin absorbent paper for use in toilets', 'name': 'toilet_tissue'}, {'frequency': 'f', 'id': 1119, 'synset': 'tomato.n.01', 'synonyms': ['tomato'], 'def': 'mildly acid red or yellow pulpy fruit eaten as a vegetable', 'name': 'tomato'}, {'frequency': 'c', 'id': 1120, 'synset': 'tongs.n.01', 'synonyms': ['tongs'], 'def': 'any of various devices for taking hold of objects; usually have two hinged legs with handles above and pointed hooks below', 'name': 'tongs'}, {'frequency': 'c', 'id': 1121, 'synset': 'toolbox.n.01', 'synonyms': ['toolbox'], 'def': 'a box or chest or cabinet for holding hand tools', 'name': 'toolbox'}, {'frequency': 'f', 'id': 1122, 'synset': 'toothbrush.n.01', 'synonyms': ['toothbrush'], 'def': 'small brush; has long handle; used to clean teeth', 'name': 'toothbrush'}, {'frequency': 'f', 'id': 1123, 'synset': 'toothpaste.n.01', 'synonyms': ['toothpaste'], 'def': 'a dentifrice in the form of a paste', 'name': 'toothpaste'}, {'frequency': 'c', 'id': 1124, 'synset': 'toothpick.n.01', 'synonyms': ['toothpick'], 'def': 'pick consisting of a small strip of wood or plastic; used to pick food from between the teeth', 'name': 'toothpick'}, {'frequency': 'c', 'id': 1125, 'synset': 'top.n.09', 'synonyms': ['cover'], 'def': 'covering for a hole (especially a hole in the top of a container)', 'name': 'cover'}, {'frequency': 'c', 'id': 1126, 'synset': 'tortilla.n.01', 'synonyms': ['tortilla'], 'def': 'thin unleavened pancake made from cornmeal or wheat flour', 'name': 'tortilla'}, {'frequency': 'c', 'id': 1127, 'synset': 'tow_truck.n.01', 'synonyms': ['tow_truck'], 'def': 'a truck equipped to hoist and pull wrecked cars (or to remove cars from no-parking zones)', 'name': 'tow_truck'}, {'frequency': 'f', 'id': 1128, 'synset': 'towel.n.01', 'synonyms': ['towel'], 'def': 'a rectangular piece of absorbent cloth (or paper) for drying or wiping', 'name': 'towel'}, {'frequency': 'f', 'id': 1129, 'synset': 'towel_rack.n.01', 'synonyms': ['towel_rack', 'towel_rail', 'towel_bar'], 'def': 'a rack consisting of one or more bars on which towels can be hung', 'name': 'towel_rack'}, {'frequency': 'f', 'id': 1130, 'synset': 'toy.n.03', 'synonyms': ['toy'], 'def': 'a device regarded as providing amusement', 'name': 'toy'}, {'frequency': 'c', 'id': 1131, 'synset': 'tractor.n.01', 'synonyms': ['tractor_(farm_equipment)'], 'def': 'a wheeled vehicle with large wheels; used in farming and other applications', 'name': 'tractor_(farm_equipment)'}, {'frequency': 'f', 'id': 1132, 'synset': 'traffic_light.n.01', 'synonyms': ['traffic_light'], 'def': 'a device to control vehicle traffic often consisting of three or more lights', 'name': 'traffic_light'}, {'frequency': 'r', 'id': 1133, 'synset': 'trail_bike.n.01', 'synonyms': ['dirt_bike'], 'def': 'a lightweight motorcycle equipped with rugged tires and suspension for off-road use', 'name': 'dirt_bike'}, {'frequency': 'c', 'id': 1134, 'synset': 'trailer_truck.n.01', 'synonyms': ['trailer_truck', 'tractor_trailer', 'trucking_rig', 'articulated_lorry', 'semi_truck'], 'def': 'a truck consisting of a tractor and trailer together', 'name': 'trailer_truck'}, {'frequency': 'f', 'id': 1135, 'synset': 'train.n.01', 'synonyms': ['train_(railroad_vehicle)', 'railroad_train'], 'def': 'public or private transport provided by a line of railway cars coupled together and drawn by a locomotive', 'name': 'train_(railroad_vehicle)'}, {'frequency': 'r', 'id': 1136, 'synset': 'trampoline.n.01', 'synonyms': ['trampoline'], 'def': 'gymnastic apparatus consisting of a strong canvas sheet attached with springs to a metal frame', 'name': 'trampoline'}, {'frequency': 'f', 'id': 1137, 'synset': 'tray.n.01', 'synonyms': ['tray'], 'def': 'an open receptacle for holding or displaying or serving articles or food', 'name': 'tray'}, {'frequency': 'r', 'id': 1138, 'synset': 'tree_house.n.01', 'synonyms': ['tree_house'], 'def': '(NOT A TREE) a PLAYHOUSE built in the branches of a tree', 'name': 'tree_house'}, {'frequency': 'r', 'id': 1139, 'synset': 'trench_coat.n.01', 'synonyms': ['trench_coat'], 'def': 'a military style raincoat; belted with deep pockets', 'name': 'trench_coat'}, {'frequency': 'r', 'id': 1140, 'synset': 'triangle.n.05', 'synonyms': ['triangle_(musical_instrument)'], 'def': 'a percussion instrument consisting of a metal bar bent in the shape of an open triangle', 'name': 'triangle_(musical_instrument)'}, {'frequency': 'r', 'id': 1141, 'synset': 'tricycle.n.01', 'synonyms': ['tricycle'], 'def': 'a vehicle with three wheels that is moved by foot pedals', 'name': 'tricycle'}, {'frequency': 'c', 'id': 1142, 'synset': 'tripod.n.01', 'synonyms': ['tripod'], 'def': 'a three-legged rack used for support', 'name': 'tripod'}, {'frequency': 'f', 'id': 1143, 'synset': 'trouser.n.01', 'synonyms': ['trousers', 'pants_(clothing)'], 'def': 'a garment extending from the waist to the knee or ankle, covering each leg separately', 'name': 'trousers'}, {'frequency': 'f', 'id': 1144, 'synset': 'truck.n.01', 'synonyms': ['truck'], 'def': 'an automotive vehicle suitable for hauling', 'name': 'truck'}, {'frequency': 'r', 'id': 1145, 'synset': 'truffle.n.03', 'synonyms': ['truffle_(chocolate)', 'chocolate_truffle'], 'def': 'creamy chocolate candy', 'name': 'truffle_(chocolate)'}, {'frequency': 'c', 'id': 1146, 'synset': 'trunk.n.02', 'synonyms': ['trunk'], 'def': 'luggage consisting of a large strong case used when traveling or for storage', 'name': 'trunk'}, {'frequency': 'r', 'id': 1147, 'synset': 'tub.n.02', 'synonyms': ['vat'], 'def': 'a large open vessel for holding or storing liquids', 'name': 'vat'}, {'frequency': 'c', 'id': 1148, 'synset': 'turban.n.01', 'synonyms': ['turban'], 'def': 'a traditional headdress consisting of a long scarf wrapped around the head', 'name': 'turban'}, {'frequency': 'r', 'id': 1149, 'synset': 'turkey.n.01', 'synonyms': ['turkey_(bird)'], 'def': 'large gallinaceous bird with fan-shaped tail; widely domesticated for food', 'name': 'turkey_(bird)'}, {'frequency': 'c', 'id': 1150, 'synset': 'turkey.n.04', 'synonyms': ['turkey_(food)'], 'def': 'flesh of large domesticated fowl usually roasted', 'name': 'turkey_(food)'}, {'frequency': 'r', 'id': 1151, 'synset': 'turnip.n.01', 'synonyms': ['turnip'], 'def': 'widely cultivated plant having a large fleshy edible white or yellow root', 'name': 'turnip'}, {'frequency': 'c', 'id': 1152, 'synset': 'turtle.n.02', 'synonyms': ['turtle'], 'def': 'any of various aquatic and land reptiles having a bony shell and flipper-like limbs for swimming', 'name': 'turtle'}, {'frequency': 'r', 'id': 1153, 'synset': 'turtleneck.n.01', 'synonyms': ['turtleneck_(clothing)', 'polo-neck'], 'def': 'a sweater or jersey with a high close-fitting collar', 'name': 'turtleneck_(clothing)'}, {'frequency': 'r', 'id': 1154, 'synset': 'typewriter.n.01', 'synonyms': ['typewriter'], 'def': 'hand-operated character printer for printing written messages one character at a time', 'name': 'typewriter'}, {'frequency': 'f', 'id': 1155, 'synset': 'umbrella.n.01', 'synonyms': ['umbrella'], 'def': 'a lightweight handheld collapsible canopy', 'name': 'umbrella'}, {'frequency': 'c', 'id': 1156, 'synset': 'underwear.n.01', 'synonyms': ['underwear', 'underclothes', 'underclothing', 'underpants'], 'def': 'undergarment worn next to the skin and under the outer garments', 'name': 'underwear'}, {'frequency': 'r', 'id': 1157, 'synset': 'unicycle.n.01', 'synonyms': ['unicycle'], 'def': 'a vehicle with a single wheel that is driven by pedals', 'name': 'unicycle'}, {'frequency': 'c', 'id': 1158, 'synset': 'urinal.n.01', 'synonyms': ['urinal'], 'def': 'a plumbing fixture (usually attached to the wall) used by men to urinate', 'name': 'urinal'}, {'frequency': 'r', 'id': 1159, 'synset': 'urn.n.01', 'synonyms': ['urn'], 'def': 'a large vase that usually has a pedestal or feet', 'name': 'urn'}, {'frequency': 'c', 'id': 1160, 'synset': 'vacuum.n.04', 'synonyms': ['vacuum_cleaner'], 'def': 'an electrical home appliance that cleans by suction', 'name': 'vacuum_cleaner'}, {'frequency': 'c', 'id': 1161, 'synset': 'valve.n.03', 'synonyms': ['valve'], 'def': 'control consisting of a mechanical device for controlling the flow of a fluid', 'name': 'valve'}, {'frequency': 'f', 'id': 1162, 'synset': 'vase.n.01', 'synonyms': ['vase'], 'def': 'an open jar of glass or porcelain used as an ornament or to hold flowers', 'name': 'vase'}, {'frequency': 'c', 'id': 1163, 'synset': 'vending_machine.n.01', 'synonyms': ['vending_machine'], 'def': 'a slot machine for selling goods', 'name': 'vending_machine'}, {'frequency': 'f', 'id': 1164, 'synset': 'vent.n.01', 'synonyms': ['vent', 'blowhole', 'air_vent'], 'def': 'a hole for the escape of gas or air', 'name': 'vent'}, {'frequency': 'c', 'id': 1165, 'synset': 'videotape.n.01', 'synonyms': ['videotape'], 'def': 'a video recording made on magnetic tape', 'name': 'videotape'}, {'frequency': 'r', 'id': 1166, 'synset': 'vinegar.n.01', 'synonyms': ['vinegar'], 'def': 'sour-tasting liquid produced usually by oxidation of the alcohol in wine or cider and used as a condiment or food preservative', 'name': 'vinegar'}, {'frequency': 'r', 'id': 1167, 'synset': 'violin.n.01', 'synonyms': ['violin', 'fiddle'], 'def': 'bowed stringed instrument that is the highest member of the violin family', 'name': 'violin'}, {'frequency': 'r', 'id': 1168, 'synset': 'vodka.n.01', 'synonyms': ['vodka'], 'def': 'unaged colorless liquor originating in Russia', 'name': 'vodka'}, {'frequency': 'r', 'id': 1169, 'synset': 'volleyball.n.02', 'synonyms': ['volleyball'], 'def': 'an inflated ball used in playing volleyball', 'name': 'volleyball'}, {'frequency': 'r', 'id': 1170, 'synset': 'vulture.n.01', 'synonyms': ['vulture'], 'def': 'any of various large birds of prey having naked heads and weak claws and feeding chiefly on carrion', 'name': 'vulture'}, {'frequency': 'c', 'id': 1171, 'synset': 'waffle.n.01', 'synonyms': ['waffle'], 'def': 'pancake batter baked in a waffle iron', 'name': 'waffle'}, {'frequency': 'r', 'id': 1172, 'synset': 'waffle_iron.n.01', 'synonyms': ['waffle_iron'], 'def': 'a kitchen appliance for baking waffles', 'name': 'waffle_iron'}, {'frequency': 'c', 'id': 1173, 'synset': 'wagon.n.01', 'synonyms': ['wagon'], 'def': 'any of various kinds of wheeled vehicles drawn by an animal or a tractor', 'name': 'wagon'}, {'frequency': 'c', 'id': 1174, 'synset': 'wagon_wheel.n.01', 'synonyms': ['wagon_wheel'], 'def': 'a wheel of a wagon', 'name': 'wagon_wheel'}, {'frequency': 'c', 'id': 1175, 'synset': 'walking_stick.n.01', 'synonyms': ['walking_stick'], 'def': 'a stick carried in the hand for support in walking', 'name': 'walking_stick'}, {'frequency': 'c', 'id': 1176, 'synset': 'wall_clock.n.01', 'synonyms': ['wall_clock'], 'def': 'a clock mounted on a wall', 'name': 'wall_clock'}, {'frequency': 'f', 'id': 1177, 'synset': 'wall_socket.n.01', 'synonyms': ['wall_socket', 'wall_plug', 'electric_outlet', 'electrical_outlet', 'outlet', 'electric_receptacle'], 'def': 'receptacle providing a place in a wiring system where current can be taken to run electrical devices', 'name': 'wall_socket'}, {'frequency': 'c', 'id': 1178, 'synset': 'wallet.n.01', 'synonyms': ['wallet', 'billfold'], 'def': 'a pocket-size case for holding papers and paper money', 'name': 'wallet'}, {'frequency': 'r', 'id': 1179, 'synset': 'walrus.n.01', 'synonyms': ['walrus'], 'def': 'either of two large northern marine mammals having ivory tusks and tough hide over thick blubber', 'name': 'walrus'}, {'frequency': 'r', 'id': 1180, 'synset': 'wardrobe.n.01', 'synonyms': ['wardrobe'], 'def': 'a tall piece of furniture that provides storage space for clothes; has a door and rails or hooks for hanging clothes', 'name': 'wardrobe'}, {'frequency': 'r', 'id': 1181, 'synset': 'wasabi.n.02', 'synonyms': ['wasabi'], 'def': 'the thick green root of the wasabi plant that the Japanese use in cooking and that tastes like strong horseradish', 'name': 'wasabi'}, {'frequency': 'c', 'id': 1182, 'synset': 'washer.n.03', 'synonyms': ['automatic_washer', 'washing_machine'], 'def': 'a home appliance for washing clothes and linens automatically', 'name': 'automatic_washer'}, {'frequency': 'f', 'id': 1183, 'synset': 'watch.n.01', 'synonyms': ['watch', 'wristwatch'], 'def': 'a small, portable timepiece', 'name': 'watch'}, {'frequency': 'f', 'id': 1184, 'synset': 'water_bottle.n.01', 'synonyms': ['water_bottle'], 'def': 'a bottle for holding water', 'name': 'water_bottle'}, {'frequency': 'c', 'id': 1185, 'synset': 'water_cooler.n.01', 'synonyms': ['water_cooler'], 'def': 'a device for cooling and dispensing drinking water', 'name': 'water_cooler'}, {'frequency': 'c', 'id': 1186, 'synset': 'water_faucet.n.01', 'synonyms': ['water_faucet', 'water_tap', 'tap_(water_faucet)'], 'def': 'a faucet for drawing water from a pipe or cask', 'name': 'water_faucet'}, {'frequency': 'r', 'id': 1187, 'synset': 'water_filter.n.01', 'synonyms': ['water_filter'], 'def': 'a filter to remove impurities from the water supply', 'name': 'water_filter'}, {'frequency': 'r', 'id': 1188, 'synset': 'water_heater.n.01', 'synonyms': ['water_heater', 'hot-water_heater'], 'def': 'a heater and storage tank to supply heated water', 'name': 'water_heater'}, {'frequency': 'r', 'id': 1189, 'synset': 'water_jug.n.01', 'synonyms': ['water_jug'], 'def': 'a jug that holds water', 'name': 'water_jug'}, {'frequency': 'r', 'id': 1190, 'synset': 'water_pistol.n.01', 'synonyms': ['water_gun', 'squirt_gun'], 'def': 'plaything consisting of a toy pistol that squirts water', 'name': 'water_gun'}, {'frequency': 'c', 'id': 1191, 'synset': 'water_scooter.n.01', 'synonyms': ['water_scooter', 'sea_scooter', 'jet_ski'], 'def': 'a motorboat resembling a motor scooter (NOT A SURFBOARD OR WATER SKI)', 'name': 'water_scooter'}, {'frequency': 'c', 'id': 1192, 'synset': 'water_ski.n.01', 'synonyms': ['water_ski'], 'def': 'broad ski for skimming over water towed by a speedboat (DO NOT MARK WATER)', 'name': 'water_ski'}, {'frequency': 'c', 'id': 1193, 'synset': 'water_tower.n.01', 'synonyms': ['water_tower'], 'def': 'a large reservoir for water', 'name': 'water_tower'}, {'frequency': 'c', 'id': 1194, 'synset': 'watering_can.n.01', 'synonyms': ['watering_can'], 'def': 'a container with a handle and a spout with a perforated nozzle; used to sprinkle water over plants', 'name': 'watering_can'}, {'frequency': 'c', 'id': 1195, 'synset': 'watermelon.n.02', 'synonyms': ['watermelon'], 'def': 'large oblong or roundish melon with a hard green rind and sweet watery red or occasionally yellowish pulp', 'name': 'watermelon'}, {'frequency': 'f', 'id': 1196, 'synset': 'weathervane.n.01', 'synonyms': ['weathervane', 'vane_(weathervane)', 'wind_vane'], 'def': 'mechanical device attached to an elevated structure; rotates freely to show the direction of the wind', 'name': 'weathervane'}, {'frequency': 'c', 'id': 1197, 'synset': 'webcam.n.01', 'synonyms': ['webcam'], 'def': 'a digital camera designed to take digital photographs and transmit them over the internet', 'name': 'webcam'}, {'frequency': 'c', 'id': 1198, 'synset': 'wedding_cake.n.01', 'synonyms': ['wedding_cake', 'bridecake'], 'def': 'a rich cake with two or more tiers and covered with frosting and decorations; served at a wedding reception', 'name': 'wedding_cake'}, {'frequency': 'c', 'id': 1199, 'synset': 'wedding_ring.n.01', 'synonyms': ['wedding_ring', 'wedding_band'], 'def': 'a ring given to the bride and/or groom at the wedding', 'name': 'wedding_ring'}, {'frequency': 'f', 'id': 1200, 'synset': 'wet_suit.n.01', 'synonyms': ['wet_suit'], 'def': 'a close-fitting garment made of a permeable material; worn in cold water to retain body heat', 'name': 'wet_suit'}, {'frequency': 'f', 'id': 1201, 'synset': 'wheel.n.01', 'synonyms': ['wheel'], 'def': 'a circular frame with spokes (or a solid disc) that can rotate on a shaft or axle', 'name': 'wheel'}, {'frequency': 'c', 'id': 1202, 'synset': 'wheelchair.n.01', 'synonyms': ['wheelchair'], 'def': 'a movable chair mounted on large wheels', 'name': 'wheelchair'}, {'frequency': 'c', 'id': 1203, 'synset': 'whipped_cream.n.01', 'synonyms': ['whipped_cream'], 'def': 'cream that has been beaten until light and fluffy', 'name': 'whipped_cream'}, {'frequency': 'r', 'id': 1204, 'synset': 'whiskey.n.01', 'synonyms': ['whiskey'], 'def': 'a liquor made from fermented mash of grain', 'name': 'whiskey'}, {'frequency': 'r', 'id': 1205, 'synset': 'whistle.n.03', 'synonyms': ['whistle'], 'def': 'a small wind instrument that produces a whistling sound by blowing into it', 'name': 'whistle'}, {'frequency': 'r', 'id': 1206, 'synset': 'wick.n.02', 'synonyms': ['wick'], 'def': 'a loosely woven cord in a candle or oil lamp that is lit on fire', 'name': 'wick'}, {'frequency': 'c', 'id': 1207, 'synset': 'wig.n.01', 'synonyms': ['wig'], 'def': 'hairpiece covering the head and made of real or synthetic hair', 'name': 'wig'}, {'frequency': 'c', 'id': 1208, 'synset': 'wind_chime.n.01', 'synonyms': ['wind_chime'], 'def': 'a decorative arrangement of pieces of metal or glass or pottery that hang together loosely so the wind can cause them to tinkle', 'name': 'wind_chime'}, {'frequency': 'c', 'id': 1209, 'synset': 'windmill.n.01', 'synonyms': ['windmill'], 'def': 'a mill that is powered by the wind', 'name': 'windmill'}, {'frequency': 'c', 'id': 1210, 'synset': 'window_box.n.01', 'synonyms': ['window_box_(for_plants)'], 'def': 'a container for growing plants on a windowsill', 'name': 'window_box_(for_plants)'}, {'frequency': 'f', 'id': 1211, 'synset': 'windshield_wiper.n.01', 'synonyms': ['windshield_wiper', 'windscreen_wiper', 'wiper_(for_windshield/screen)'], 'def': 'a mechanical device that cleans the windshield', 'name': 'windshield_wiper'}, {'frequency': 'c', 'id': 1212, 'synset': 'windsock.n.01', 'synonyms': ['windsock', 'air_sock', 'air-sleeve', 'wind_sleeve', 'wind_cone'], 'def': 'a truncated cloth cone mounted on a mast/pole; shows wind direction', 'name': 'windsock'}, {'frequency': 'f', 'id': 1213, 'synset': 'wine_bottle.n.01', 'synonyms': ['wine_bottle'], 'def': 'a bottle for holding wine', 'name': 'wine_bottle'}, {'frequency': 'r', 'id': 1214, 'synset': 'wine_bucket.n.01', 'synonyms': ['wine_bucket', 'wine_cooler'], 'def': 'a bucket of ice used to chill a bottle of wine', 'name': 'wine_bucket'}, {'frequency': 'f', 'id': 1215, 'synset': 'wineglass.n.01', 'synonyms': ['wineglass'], 'def': 'a glass that has a stem and in which wine is served', 'name': 'wineglass'}, {'frequency': 'r', 'id': 1216, 'synset': 'wing_chair.n.01', 'synonyms': ['wing_chair'], 'def': 'easy chair having wings on each side of a high back', 'name': 'wing_chair'}, {'frequency': 'c', 'id': 1217, 'synset': 'winker.n.02', 'synonyms': ['blinder_(for_horses)'], 'def': 'blinds that prevent a horse from seeing something on either side', 'name': 'blinder_(for_horses)'}, {'frequency': 'c', 'id': 1218, 'synset': 'wok.n.01', 'synonyms': ['wok'], 'def': 'pan with a convex bottom; used for frying in Chinese cooking', 'name': 'wok'}, {'frequency': 'r', 'id': 1219, 'synset': 'wolf.n.01', 'synonyms': ['wolf'], 'def': 'a wild carnivorous mammal of the dog family, living and hunting in packs', 'name': 'wolf'}, {'frequency': 'c', 'id': 1220, 'synset': 'wooden_spoon.n.02', 'synonyms': ['wooden_spoon'], 'def': 'a spoon made of wood', 'name': 'wooden_spoon'}, {'frequency': 'c', 'id': 1221, 'synset': 'wreath.n.01', 'synonyms': ['wreath'], 'def': 'an arrangement of flowers, leaves, or stems fastened in a ring', 'name': 'wreath'}, {'frequency': 'c', 'id': 1222, 'synset': 'wrench.n.03', 'synonyms': ['wrench', 'spanner'], 'def': 'a hand tool that is used to hold or twist a nut or bolt', 'name': 'wrench'}, {'frequency': 'c', 'id': 1223, 'synset': 'wristband.n.01', 'synonyms': ['wristband'], 'def': 'band consisting of a part of a sleeve that covers the wrist', 'name': 'wristband'}, {'frequency': 'f', 'id': 1224, 'synset': 'wristlet.n.01', 'synonyms': ['wristlet', 'wrist_band'], 'def': 'a band or bracelet worn around the wrist', 'name': 'wristlet'}, {'frequency': 'r', 'id': 1225, 'synset': 'yacht.n.01', 'synonyms': ['yacht'], 'def': 'an expensive vessel propelled by sail or power and used for cruising or racing', 'name': 'yacht'}, {'frequency': 'r', 'id': 1226, 'synset': 'yak.n.02', 'synonyms': ['yak'], 'def': 'large long-haired wild ox of Tibet often domesticated', 'name': 'yak'}, {'frequency': 'c', 'id': 1227, 'synset': 'yogurt.n.01', 'synonyms': ['yogurt', 'yoghurt', 'yoghourt'], 'def': 'a custard-like food made from curdled milk', 'name': 'yogurt'}, {'frequency': 'r', 'id': 1228, 'synset': 'yoke.n.07', 'synonyms': ['yoke_(animal_equipment)'], 'def': 'gear joining two animals at the neck; NOT egg yolk', 'name': 'yoke_(animal_equipment)'}, {'frequency': 'f', 'id': 1229, 'synset': 'zebra.n.01', 'synonyms': ['zebra'], 'def': 'any of several fleet black-and-white striped African equines', 'name': 'zebra'}, {'frequency': 'c', 'id': 1230, 'synset': 'zucchini.n.02', 'synonyms': ['zucchini', 'courgette'], 'def': 'small cucumber-shaped vegetable marrow; typically dark green', 'name': 'zucchini'}] # noqa
+# fmt: on
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/lvis_v1_categories.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/lvis_v1_categories.py
new file mode 100644
index 0000000000000000000000000000000000000000..7374e6968bb006f5d8c49e75d9d3b31ea3d77d05
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/lvis_v1_categories.py
@@ -0,0 +1,16 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+# Autogen with
+# with open("lvis_v1_val.json", "r") as f:
+# a = json.load(f)
+# c = a["categories"]
+# for x in c:
+# del x["image_count"]
+# del x["instance_count"]
+# LVIS_CATEGORIES = repr(c) + " # noqa"
+# with open("/tmp/lvis_categories.py", "wt") as f:
+# f.write(f"LVIS_CATEGORIES = {LVIS_CATEGORIES}")
+# Then paste the contents of that file below
+
+# fmt: off
+LVIS_CATEGORIES = [{'frequency': 'c', 'synset': 'aerosol.n.02', 'synonyms': ['aerosol_can', 'spray_can'], 'id': 1, 'def': 'a dispenser that holds a substance under pressure', 'name': 'aerosol_can'}, {'frequency': 'f', 'synset': 'air_conditioner.n.01', 'synonyms': ['air_conditioner'], 'id': 2, 'def': 'a machine that keeps air cool and dry', 'name': 'air_conditioner'}, {'frequency': 'f', 'synset': 'airplane.n.01', 'synonyms': ['airplane', 'aeroplane'], 'id': 3, 'def': 'an aircraft that has a fixed wing and is powered by propellers or jets', 'name': 'airplane'}, {'frequency': 'f', 'synset': 'alarm_clock.n.01', 'synonyms': ['alarm_clock'], 'id': 4, 'def': 'a clock that wakes a sleeper at some preset time', 'name': 'alarm_clock'}, {'frequency': 'c', 'synset': 'alcohol.n.01', 'synonyms': ['alcohol', 'alcoholic_beverage'], 'id': 5, 'def': 'a liquor or brew containing alcohol as the active agent', 'name': 'alcohol'}, {'frequency': 'c', 'synset': 'alligator.n.02', 'synonyms': ['alligator', 'gator'], 'id': 6, 'def': 'amphibious reptiles related to crocodiles but with shorter broader snouts', 'name': 'alligator'}, {'frequency': 'c', 'synset': 'almond.n.02', 'synonyms': ['almond'], 'id': 7, 'def': 'oval-shaped edible seed of the almond tree', 'name': 'almond'}, {'frequency': 'c', 'synset': 'ambulance.n.01', 'synonyms': ['ambulance'], 'id': 8, 'def': 'a vehicle that takes people to and from hospitals', 'name': 'ambulance'}, {'frequency': 'c', 'synset': 'amplifier.n.01', 'synonyms': ['amplifier'], 'id': 9, 'def': 'electronic equipment that increases strength of signals', 'name': 'amplifier'}, {'frequency': 'c', 'synset': 'anklet.n.03', 'synonyms': ['anklet', 'ankle_bracelet'], 'id': 10, 'def': 'an ornament worn around the ankle', 'name': 'anklet'}, {'frequency': 'f', 'synset': 'antenna.n.01', 'synonyms': ['antenna', 'aerial', 'transmitting_aerial'], 'id': 11, 'def': 'an electrical device that sends or receives radio or television signals', 'name': 'antenna'}, {'frequency': 'f', 'synset': 'apple.n.01', 'synonyms': ['apple'], 'id': 12, 'def': 'fruit with red or yellow or green skin and sweet to tart crisp whitish flesh', 'name': 'apple'}, {'frequency': 'r', 'synset': 'applesauce.n.01', 'synonyms': ['applesauce'], 'id': 13, 'def': 'puree of stewed apples usually sweetened and spiced', 'name': 'applesauce'}, {'frequency': 'r', 'synset': 'apricot.n.02', 'synonyms': ['apricot'], 'id': 14, 'def': 'downy yellow to rosy-colored fruit resembling a small peach', 'name': 'apricot'}, {'frequency': 'f', 'synset': 'apron.n.01', 'synonyms': ['apron'], 'id': 15, 'def': 'a garment of cloth that is tied about the waist and worn to protect clothing', 'name': 'apron'}, {'frequency': 'c', 'synset': 'aquarium.n.01', 'synonyms': ['aquarium', 'fish_tank'], 'id': 16, 'def': 'a tank/pool/bowl filled with water for keeping live fish and underwater animals', 'name': 'aquarium'}, {'frequency': 'r', 'synset': 'arctic.n.02', 'synonyms': ['arctic_(type_of_shoe)', 'galosh', 'golosh', 'rubber_(type_of_shoe)', 'gumshoe'], 'id': 17, 'def': 'a waterproof overshoe that protects shoes from water or snow', 'name': 'arctic_(type_of_shoe)'}, {'frequency': 'c', 'synset': 'armband.n.02', 'synonyms': ['armband'], 'id': 18, 'def': 'a band worn around the upper arm', 'name': 'armband'}, {'frequency': 'f', 'synset': 'armchair.n.01', 'synonyms': ['armchair'], 'id': 19, 'def': 'chair with a support on each side for arms', 'name': 'armchair'}, {'frequency': 'r', 'synset': 'armoire.n.01', 'synonyms': ['armoire'], 'id': 20, 'def': 'a large wardrobe or cabinet', 'name': 'armoire'}, {'frequency': 'r', 'synset': 'armor.n.01', 'synonyms': ['armor', 'armour'], 'id': 21, 'def': 'protective covering made of metal and used in combat', 'name': 'armor'}, {'frequency': 'c', 'synset': 'artichoke.n.02', 'synonyms': ['artichoke'], 'id': 22, 'def': 'a thistlelike flower head with edible fleshy leaves and heart', 'name': 'artichoke'}, {'frequency': 'f', 'synset': 'ashcan.n.01', 'synonyms': ['trash_can', 'garbage_can', 'wastebin', 'dustbin', 'trash_barrel', 'trash_bin'], 'id': 23, 'def': 'a bin that holds rubbish until it is collected', 'name': 'trash_can'}, {'frequency': 'c', 'synset': 'ashtray.n.01', 'synonyms': ['ashtray'], 'id': 24, 'def': "a receptacle for the ash from smokers' cigars or cigarettes", 'name': 'ashtray'}, {'frequency': 'c', 'synset': 'asparagus.n.02', 'synonyms': ['asparagus'], 'id': 25, 'def': 'edible young shoots of the asparagus plant', 'name': 'asparagus'}, {'frequency': 'c', 'synset': 'atomizer.n.01', 'synonyms': ['atomizer', 'atomiser', 'spray', 'sprayer', 'nebulizer', 'nebuliser'], 'id': 26, 'def': 'a dispenser that turns a liquid (such as perfume) into a fine mist', 'name': 'atomizer'}, {'frequency': 'f', 'synset': 'avocado.n.01', 'synonyms': ['avocado'], 'id': 27, 'def': 'a pear-shaped fruit with green or blackish skin and rich yellowish pulp enclosing a single large seed', 'name': 'avocado'}, {'frequency': 'c', 'synset': 'award.n.02', 'synonyms': ['award', 'accolade'], 'id': 28, 'def': 'a tangible symbol signifying approval or distinction', 'name': 'award'}, {'frequency': 'f', 'synset': 'awning.n.01', 'synonyms': ['awning'], 'id': 29, 'def': 'a canopy made of canvas to shelter people or things from rain or sun', 'name': 'awning'}, {'frequency': 'r', 'synset': 'ax.n.01', 'synonyms': ['ax', 'axe'], 'id': 30, 'def': 'an edge tool with a heavy bladed head mounted across a handle', 'name': 'ax'}, {'frequency': 'r', 'synset': 'baboon.n.01', 'synonyms': ['baboon'], 'id': 31, 'def': 'large terrestrial monkeys having doglike muzzles', 'name': 'baboon'}, {'frequency': 'f', 'synset': 'baby_buggy.n.01', 'synonyms': ['baby_buggy', 'baby_carriage', 'perambulator', 'pram', 'stroller'], 'id': 32, 'def': 'a small vehicle with four wheels in which a baby or child is pushed around', 'name': 'baby_buggy'}, {'frequency': 'c', 'synset': 'backboard.n.01', 'synonyms': ['basketball_backboard'], 'id': 33, 'def': 'a raised vertical board with basket attached; used to play basketball', 'name': 'basketball_backboard'}, {'frequency': 'f', 'synset': 'backpack.n.01', 'synonyms': ['backpack', 'knapsack', 'packsack', 'rucksack', 'haversack'], 'id': 34, 'def': 'a bag carried by a strap on your back or shoulder', 'name': 'backpack'}, {'frequency': 'f', 'synset': 'bag.n.04', 'synonyms': ['handbag', 'purse', 'pocketbook'], 'id': 35, 'def': 'a container used for carrying money and small personal items or accessories', 'name': 'handbag'}, {'frequency': 'f', 'synset': 'bag.n.06', 'synonyms': ['suitcase', 'baggage', 'luggage'], 'id': 36, 'def': 'cases used to carry belongings when traveling', 'name': 'suitcase'}, {'frequency': 'c', 'synset': 'bagel.n.01', 'synonyms': ['bagel', 'beigel'], 'id': 37, 'def': 'glazed yeast-raised doughnut-shaped roll with hard crust', 'name': 'bagel'}, {'frequency': 'r', 'synset': 'bagpipe.n.01', 'synonyms': ['bagpipe'], 'id': 38, 'def': 'a tubular wind instrument; the player blows air into a bag and squeezes it out', 'name': 'bagpipe'}, {'frequency': 'r', 'synset': 'baguet.n.01', 'synonyms': ['baguet', 'baguette'], 'id': 39, 'def': 'narrow French stick loaf', 'name': 'baguet'}, {'frequency': 'r', 'synset': 'bait.n.02', 'synonyms': ['bait', 'lure'], 'id': 40, 'def': 'something used to lure fish or other animals into danger so they can be trapped or killed', 'name': 'bait'}, {'frequency': 'f', 'synset': 'ball.n.06', 'synonyms': ['ball'], 'id': 41, 'def': 'a spherical object used as a plaything', 'name': 'ball'}, {'frequency': 'r', 'synset': 'ballet_skirt.n.01', 'synonyms': ['ballet_skirt', 'tutu'], 'id': 42, 'def': 'very short skirt worn by ballerinas', 'name': 'ballet_skirt'}, {'frequency': 'f', 'synset': 'balloon.n.01', 'synonyms': ['balloon'], 'id': 43, 'def': 'large tough nonrigid bag filled with gas or heated air', 'name': 'balloon'}, {'frequency': 'c', 'synset': 'bamboo.n.02', 'synonyms': ['bamboo'], 'id': 44, 'def': 'woody tropical grass having hollow woody stems', 'name': 'bamboo'}, {'frequency': 'f', 'synset': 'banana.n.02', 'synonyms': ['banana'], 'id': 45, 'def': 'elongated crescent-shaped yellow fruit with soft sweet flesh', 'name': 'banana'}, {'frequency': 'c', 'synset': 'band_aid.n.01', 'synonyms': ['Band_Aid'], 'id': 46, 'def': 'trade name for an adhesive bandage to cover small cuts or blisters', 'name': 'Band_Aid'}, {'frequency': 'c', 'synset': 'bandage.n.01', 'synonyms': ['bandage'], 'id': 47, 'def': 'a piece of soft material that covers and protects an injured part of the body', 'name': 'bandage'}, {'frequency': 'f', 'synset': 'bandanna.n.01', 'synonyms': ['bandanna', 'bandana'], 'id': 48, 'def': 'large and brightly colored handkerchief; often used as a neckerchief', 'name': 'bandanna'}, {'frequency': 'r', 'synset': 'banjo.n.01', 'synonyms': ['banjo'], 'id': 49, 'def': 'a stringed instrument of the guitar family with a long neck and circular body', 'name': 'banjo'}, {'frequency': 'f', 'synset': 'banner.n.01', 'synonyms': ['banner', 'streamer'], 'id': 50, 'def': 'long strip of cloth or paper used for decoration or advertising', 'name': 'banner'}, {'frequency': 'r', 'synset': 'barbell.n.01', 'synonyms': ['barbell'], 'id': 51, 'def': 'a bar to which heavy discs are attached at each end; used in weightlifting', 'name': 'barbell'}, {'frequency': 'r', 'synset': 'barge.n.01', 'synonyms': ['barge'], 'id': 52, 'def': 'a flatbottom boat for carrying heavy loads (especially on canals)', 'name': 'barge'}, {'frequency': 'f', 'synset': 'barrel.n.02', 'synonyms': ['barrel', 'cask'], 'id': 53, 'def': 'a cylindrical container that holds liquids', 'name': 'barrel'}, {'frequency': 'c', 'synset': 'barrette.n.01', 'synonyms': ['barrette'], 'id': 54, 'def': "a pin for holding women's hair in place", 'name': 'barrette'}, {'frequency': 'c', 'synset': 'barrow.n.03', 'synonyms': ['barrow', 'garden_cart', 'lawn_cart', 'wheelbarrow'], 'id': 55, 'def': 'a cart for carrying small loads; has handles and one or more wheels', 'name': 'barrow'}, {'frequency': 'f', 'synset': 'base.n.03', 'synonyms': ['baseball_base'], 'id': 56, 'def': 'a place that the runner must touch before scoring', 'name': 'baseball_base'}, {'frequency': 'f', 'synset': 'baseball.n.02', 'synonyms': ['baseball'], 'id': 57, 'def': 'a ball used in playing baseball', 'name': 'baseball'}, {'frequency': 'f', 'synset': 'baseball_bat.n.01', 'synonyms': ['baseball_bat'], 'id': 58, 'def': 'an implement used in baseball by the batter', 'name': 'baseball_bat'}, {'frequency': 'f', 'synset': 'baseball_cap.n.01', 'synonyms': ['baseball_cap', 'jockey_cap', 'golf_cap'], 'id': 59, 'def': 'a cap with a bill', 'name': 'baseball_cap'}, {'frequency': 'f', 'synset': 'baseball_glove.n.01', 'synonyms': ['baseball_glove', 'baseball_mitt'], 'id': 60, 'def': 'the handwear used by fielders in playing baseball', 'name': 'baseball_glove'}, {'frequency': 'f', 'synset': 'basket.n.01', 'synonyms': ['basket', 'handbasket'], 'id': 61, 'def': 'a container that is usually woven and has handles', 'name': 'basket'}, {'frequency': 'c', 'synset': 'basketball.n.02', 'synonyms': ['basketball'], 'id': 62, 'def': 'an inflated ball used in playing basketball', 'name': 'basketball'}, {'frequency': 'r', 'synset': 'bass_horn.n.01', 'synonyms': ['bass_horn', 'sousaphone', 'tuba'], 'id': 63, 'def': 'the lowest brass wind instrument', 'name': 'bass_horn'}, {'frequency': 'c', 'synset': 'bat.n.01', 'synonyms': ['bat_(animal)'], 'id': 64, 'def': 'nocturnal mouselike mammal with forelimbs modified to form membranous wings', 'name': 'bat_(animal)'}, {'frequency': 'f', 'synset': 'bath_mat.n.01', 'synonyms': ['bath_mat'], 'id': 65, 'def': 'a heavy towel or mat to stand on while drying yourself after a bath', 'name': 'bath_mat'}, {'frequency': 'f', 'synset': 'bath_towel.n.01', 'synonyms': ['bath_towel'], 'id': 66, 'def': 'a large towel; to dry yourself after a bath', 'name': 'bath_towel'}, {'frequency': 'c', 'synset': 'bathrobe.n.01', 'synonyms': ['bathrobe'], 'id': 67, 'def': 'a loose-fitting robe of towelling; worn after a bath or swim', 'name': 'bathrobe'}, {'frequency': 'f', 'synset': 'bathtub.n.01', 'synonyms': ['bathtub', 'bathing_tub'], 'id': 68, 'def': 'a large open container that you fill with water and use to wash the body', 'name': 'bathtub'}, {'frequency': 'r', 'synset': 'batter.n.02', 'synonyms': ['batter_(food)'], 'id': 69, 'def': 'a liquid or semiliquid mixture, as of flour, eggs, and milk, used in cooking', 'name': 'batter_(food)'}, {'frequency': 'c', 'synset': 'battery.n.02', 'synonyms': ['battery'], 'id': 70, 'def': 'a portable device that produces electricity', 'name': 'battery'}, {'frequency': 'r', 'synset': 'beach_ball.n.01', 'synonyms': ['beachball'], 'id': 71, 'def': 'large and light ball; for play at the seaside', 'name': 'beachball'}, {'frequency': 'c', 'synset': 'bead.n.01', 'synonyms': ['bead'], 'id': 72, 'def': 'a small ball with a hole through the middle used for ornamentation, jewellery, etc.', 'name': 'bead'}, {'frequency': 'c', 'synset': 'bean_curd.n.01', 'synonyms': ['bean_curd', 'tofu'], 'id': 73, 'def': 'cheeselike food made of curdled soybean milk', 'name': 'bean_curd'}, {'frequency': 'c', 'synset': 'beanbag.n.01', 'synonyms': ['beanbag'], 'id': 74, 'def': 'a bag filled with dried beans or similar items; used in games or to sit on', 'name': 'beanbag'}, {'frequency': 'f', 'synset': 'beanie.n.01', 'synonyms': ['beanie', 'beany'], 'id': 75, 'def': 'a small skullcap; formerly worn by schoolboys and college freshmen', 'name': 'beanie'}, {'frequency': 'f', 'synset': 'bear.n.01', 'synonyms': ['bear'], 'id': 76, 'def': 'large carnivorous or omnivorous mammals with shaggy coats and claws', 'name': 'bear'}, {'frequency': 'f', 'synset': 'bed.n.01', 'synonyms': ['bed'], 'id': 77, 'def': 'a piece of furniture that provides a place to sleep', 'name': 'bed'}, {'frequency': 'r', 'synset': 'bedpan.n.01', 'synonyms': ['bedpan'], 'id': 78, 'def': 'a shallow vessel used by a bedridden patient for defecation and urination', 'name': 'bedpan'}, {'frequency': 'f', 'synset': 'bedspread.n.01', 'synonyms': ['bedspread', 'bedcover', 'bed_covering', 'counterpane', 'spread'], 'id': 79, 'def': 'decorative cover for a bed', 'name': 'bedspread'}, {'frequency': 'f', 'synset': 'beef.n.01', 'synonyms': ['cow'], 'id': 80, 'def': 'cattle/cow', 'name': 'cow'}, {'frequency': 'f', 'synset': 'beef.n.02', 'synonyms': ['beef_(food)', 'boeuf_(food)'], 'id': 81, 'def': 'meat from an adult domestic bovine', 'name': 'beef_(food)'}, {'frequency': 'r', 'synset': 'beeper.n.01', 'synonyms': ['beeper', 'pager'], 'id': 82, 'def': 'an device that beeps when the person carrying it is being paged', 'name': 'beeper'}, {'frequency': 'f', 'synset': 'beer_bottle.n.01', 'synonyms': ['beer_bottle'], 'id': 83, 'def': 'a bottle that holds beer', 'name': 'beer_bottle'}, {'frequency': 'c', 'synset': 'beer_can.n.01', 'synonyms': ['beer_can'], 'id': 84, 'def': 'a can that holds beer', 'name': 'beer_can'}, {'frequency': 'r', 'synset': 'beetle.n.01', 'synonyms': ['beetle'], 'id': 85, 'def': 'insect with hard wing covers', 'name': 'beetle'}, {'frequency': 'f', 'synset': 'bell.n.01', 'synonyms': ['bell'], 'id': 86, 'def': 'a hollow device made of metal that makes a ringing sound when struck', 'name': 'bell'}, {'frequency': 'f', 'synset': 'bell_pepper.n.02', 'synonyms': ['bell_pepper', 'capsicum'], 'id': 87, 'def': 'large bell-shaped sweet pepper in green or red or yellow or orange or black varieties', 'name': 'bell_pepper'}, {'frequency': 'f', 'synset': 'belt.n.02', 'synonyms': ['belt'], 'id': 88, 'def': 'a band to tie or buckle around the body (usually at the waist)', 'name': 'belt'}, {'frequency': 'f', 'synset': 'belt_buckle.n.01', 'synonyms': ['belt_buckle'], 'id': 89, 'def': 'the buckle used to fasten a belt', 'name': 'belt_buckle'}, {'frequency': 'f', 'synset': 'bench.n.01', 'synonyms': ['bench'], 'id': 90, 'def': 'a long seat for more than one person', 'name': 'bench'}, {'frequency': 'c', 'synset': 'beret.n.01', 'synonyms': ['beret'], 'id': 91, 'def': 'a cap with no brim or bill; made of soft cloth', 'name': 'beret'}, {'frequency': 'c', 'synset': 'bib.n.02', 'synonyms': ['bib'], 'id': 92, 'def': 'a napkin tied under the chin of a child while eating', 'name': 'bib'}, {'frequency': 'r', 'synset': 'bible.n.01', 'synonyms': ['Bible'], 'id': 93, 'def': 'the sacred writings of the Christian religions', 'name': 'Bible'}, {'frequency': 'f', 'synset': 'bicycle.n.01', 'synonyms': ['bicycle', 'bike_(bicycle)'], 'id': 94, 'def': 'a wheeled vehicle that has two wheels and is moved by foot pedals', 'name': 'bicycle'}, {'frequency': 'f', 'synset': 'bill.n.09', 'synonyms': ['visor', 'vizor'], 'id': 95, 'def': 'a brim that projects to the front to shade the eyes', 'name': 'visor'}, {'frequency': 'f', 'synset': 'billboard.n.01', 'synonyms': ['billboard'], 'id': 96, 'def': 'large outdoor signboard', 'name': 'billboard'}, {'frequency': 'c', 'synset': 'binder.n.03', 'synonyms': ['binder', 'ring-binder'], 'id': 97, 'def': 'holds loose papers or magazines', 'name': 'binder'}, {'frequency': 'c', 'synset': 'binoculars.n.01', 'synonyms': ['binoculars', 'field_glasses', 'opera_glasses'], 'id': 98, 'def': 'an optical instrument designed for simultaneous use by both eyes', 'name': 'binoculars'}, {'frequency': 'f', 'synset': 'bird.n.01', 'synonyms': ['bird'], 'id': 99, 'def': 'animal characterized by feathers and wings', 'name': 'bird'}, {'frequency': 'c', 'synset': 'bird_feeder.n.01', 'synonyms': ['birdfeeder'], 'id': 100, 'def': 'an outdoor device that supplies food for wild birds', 'name': 'birdfeeder'}, {'frequency': 'c', 'synset': 'birdbath.n.01', 'synonyms': ['birdbath'], 'id': 101, 'def': 'an ornamental basin (usually in a garden) for birds to bathe in', 'name': 'birdbath'}, {'frequency': 'c', 'synset': 'birdcage.n.01', 'synonyms': ['birdcage'], 'id': 102, 'def': 'a cage in which a bird can be kept', 'name': 'birdcage'}, {'frequency': 'c', 'synset': 'birdhouse.n.01', 'synonyms': ['birdhouse'], 'id': 103, 'def': 'a shelter for birds', 'name': 'birdhouse'}, {'frequency': 'f', 'synset': 'birthday_cake.n.01', 'synonyms': ['birthday_cake'], 'id': 104, 'def': 'decorated cake served at a birthday party', 'name': 'birthday_cake'}, {'frequency': 'r', 'synset': 'birthday_card.n.01', 'synonyms': ['birthday_card'], 'id': 105, 'def': 'a card expressing a birthday greeting', 'name': 'birthday_card'}, {'frequency': 'r', 'synset': 'black_flag.n.01', 'synonyms': ['pirate_flag'], 'id': 106, 'def': 'a flag usually bearing a white skull and crossbones on a black background', 'name': 'pirate_flag'}, {'frequency': 'c', 'synset': 'black_sheep.n.02', 'synonyms': ['black_sheep'], 'id': 107, 'def': 'sheep with a black coat', 'name': 'black_sheep'}, {'frequency': 'c', 'synset': 'blackberry.n.01', 'synonyms': ['blackberry'], 'id': 108, 'def': 'large sweet black or very dark purple edible aggregate fruit', 'name': 'blackberry'}, {'frequency': 'f', 'synset': 'blackboard.n.01', 'synonyms': ['blackboard', 'chalkboard'], 'id': 109, 'def': 'sheet of slate; for writing with chalk', 'name': 'blackboard'}, {'frequency': 'f', 'synset': 'blanket.n.01', 'synonyms': ['blanket'], 'id': 110, 'def': 'bedding that keeps a person warm in bed', 'name': 'blanket'}, {'frequency': 'c', 'synset': 'blazer.n.01', 'synonyms': ['blazer', 'sport_jacket', 'sport_coat', 'sports_jacket', 'sports_coat'], 'id': 111, 'def': 'lightweight jacket; often striped in the colors of a club or school', 'name': 'blazer'}, {'frequency': 'f', 'synset': 'blender.n.01', 'synonyms': ['blender', 'liquidizer', 'liquidiser'], 'id': 112, 'def': 'an electrically powered mixer that mix or chop or liquefy foods', 'name': 'blender'}, {'frequency': 'r', 'synset': 'blimp.n.02', 'synonyms': ['blimp'], 'id': 113, 'def': 'a small nonrigid airship used for observation or as a barrage balloon', 'name': 'blimp'}, {'frequency': 'f', 'synset': 'blinker.n.01', 'synonyms': ['blinker', 'flasher'], 'id': 114, 'def': 'a light that flashes on and off; used as a signal or to send messages', 'name': 'blinker'}, {'frequency': 'f', 'synset': 'blouse.n.01', 'synonyms': ['blouse'], 'id': 115, 'def': 'a top worn by women', 'name': 'blouse'}, {'frequency': 'f', 'synset': 'blueberry.n.02', 'synonyms': ['blueberry'], 'id': 116, 'def': 'sweet edible dark-blue berries of blueberry plants', 'name': 'blueberry'}, {'frequency': 'r', 'synset': 'board.n.09', 'synonyms': ['gameboard'], 'id': 117, 'def': 'a flat portable surface (usually rectangular) designed for board games', 'name': 'gameboard'}, {'frequency': 'f', 'synset': 'boat.n.01', 'synonyms': ['boat', 'ship_(boat)'], 'id': 118, 'def': 'a vessel for travel on water', 'name': 'boat'}, {'frequency': 'r', 'synset': 'bob.n.05', 'synonyms': ['bob', 'bobber', 'bobfloat'], 'id': 119, 'def': 'a small float usually made of cork; attached to a fishing line', 'name': 'bob'}, {'frequency': 'c', 'synset': 'bobbin.n.01', 'synonyms': ['bobbin', 'spool', 'reel'], 'id': 120, 'def': 'a thing around which thread/tape/film or other flexible materials can be wound', 'name': 'bobbin'}, {'frequency': 'c', 'synset': 'bobby_pin.n.01', 'synonyms': ['bobby_pin', 'hairgrip'], 'id': 121, 'def': 'a flat wire hairpin used to hold bobbed hair in place', 'name': 'bobby_pin'}, {'frequency': 'c', 'synset': 'boiled_egg.n.01', 'synonyms': ['boiled_egg', 'coddled_egg'], 'id': 122, 'def': 'egg cooked briefly in the shell in gently boiling water', 'name': 'boiled_egg'}, {'frequency': 'r', 'synset': 'bolo_tie.n.01', 'synonyms': ['bolo_tie', 'bolo', 'bola_tie', 'bola'], 'id': 123, 'def': 'a cord fastened around the neck with an ornamental clasp and worn as a necktie', 'name': 'bolo_tie'}, {'frequency': 'c', 'synset': 'bolt.n.03', 'synonyms': ['deadbolt'], 'id': 124, 'def': 'the part of a lock that is engaged or withdrawn with a key', 'name': 'deadbolt'}, {'frequency': 'f', 'synset': 'bolt.n.06', 'synonyms': ['bolt'], 'id': 125, 'def': 'a screw that screws into a nut to form a fastener', 'name': 'bolt'}, {'frequency': 'r', 'synset': 'bonnet.n.01', 'synonyms': ['bonnet'], 'id': 126, 'def': 'a hat tied under the chin', 'name': 'bonnet'}, {'frequency': 'f', 'synset': 'book.n.01', 'synonyms': ['book'], 'id': 127, 'def': 'a written work or composition that has been published', 'name': 'book'}, {'frequency': 'c', 'synset': 'bookcase.n.01', 'synonyms': ['bookcase'], 'id': 128, 'def': 'a piece of furniture with shelves for storing books', 'name': 'bookcase'}, {'frequency': 'c', 'synset': 'booklet.n.01', 'synonyms': ['booklet', 'brochure', 'leaflet', 'pamphlet'], 'id': 129, 'def': 'a small book usually having a paper cover', 'name': 'booklet'}, {'frequency': 'r', 'synset': 'bookmark.n.01', 'synonyms': ['bookmark', 'bookmarker'], 'id': 130, 'def': 'a marker (a piece of paper or ribbon) placed between the pages of a book', 'name': 'bookmark'}, {'frequency': 'r', 'synset': 'boom.n.04', 'synonyms': ['boom_microphone', 'microphone_boom'], 'id': 131, 'def': 'a pole carrying an overhead microphone projected over a film or tv set', 'name': 'boom_microphone'}, {'frequency': 'f', 'synset': 'boot.n.01', 'synonyms': ['boot'], 'id': 132, 'def': 'footwear that covers the whole foot and lower leg', 'name': 'boot'}, {'frequency': 'f', 'synset': 'bottle.n.01', 'synonyms': ['bottle'], 'id': 133, 'def': 'a glass or plastic vessel used for storing drinks or other liquids', 'name': 'bottle'}, {'frequency': 'c', 'synset': 'bottle_opener.n.01', 'synonyms': ['bottle_opener'], 'id': 134, 'def': 'an opener for removing caps or corks from bottles', 'name': 'bottle_opener'}, {'frequency': 'c', 'synset': 'bouquet.n.01', 'synonyms': ['bouquet'], 'id': 135, 'def': 'an arrangement of flowers that is usually given as a present', 'name': 'bouquet'}, {'frequency': 'r', 'synset': 'bow.n.04', 'synonyms': ['bow_(weapon)'], 'id': 136, 'def': 'a weapon for shooting arrows', 'name': 'bow_(weapon)'}, {'frequency': 'f', 'synset': 'bow.n.08', 'synonyms': ['bow_(decorative_ribbons)'], 'id': 137, 'def': 'a decorative interlacing of ribbons', 'name': 'bow_(decorative_ribbons)'}, {'frequency': 'f', 'synset': 'bow_tie.n.01', 'synonyms': ['bow-tie', 'bowtie'], 'id': 138, 'def': "a man's tie that ties in a bow", 'name': 'bow-tie'}, {'frequency': 'f', 'synset': 'bowl.n.03', 'synonyms': ['bowl'], 'id': 139, 'def': 'a dish that is round and open at the top for serving foods', 'name': 'bowl'}, {'frequency': 'r', 'synset': 'bowl.n.08', 'synonyms': ['pipe_bowl'], 'id': 140, 'def': 'a small round container that is open at the top for holding tobacco', 'name': 'pipe_bowl'}, {'frequency': 'c', 'synset': 'bowler_hat.n.01', 'synonyms': ['bowler_hat', 'bowler', 'derby_hat', 'derby', 'plug_hat'], 'id': 141, 'def': 'a felt hat that is round and hard with a narrow brim', 'name': 'bowler_hat'}, {'frequency': 'r', 'synset': 'bowling_ball.n.01', 'synonyms': ['bowling_ball'], 'id': 142, 'def': 'a large ball with finger holes used in the sport of bowling', 'name': 'bowling_ball'}, {'frequency': 'f', 'synset': 'box.n.01', 'synonyms': ['box'], 'id': 143, 'def': 'a (usually rectangular) container; may have a lid', 'name': 'box'}, {'frequency': 'r', 'synset': 'boxing_glove.n.01', 'synonyms': ['boxing_glove'], 'id': 144, 'def': 'large glove coverings the fists of a fighter worn for the sport of boxing', 'name': 'boxing_glove'}, {'frequency': 'c', 'synset': 'brace.n.06', 'synonyms': ['suspenders'], 'id': 145, 'def': 'elastic straps that hold trousers up (usually used in the plural)', 'name': 'suspenders'}, {'frequency': 'f', 'synset': 'bracelet.n.02', 'synonyms': ['bracelet', 'bangle'], 'id': 146, 'def': 'jewelry worn around the wrist for decoration', 'name': 'bracelet'}, {'frequency': 'r', 'synset': 'brass.n.07', 'synonyms': ['brass_plaque'], 'id': 147, 'def': 'a memorial made of brass', 'name': 'brass_plaque'}, {'frequency': 'c', 'synset': 'brassiere.n.01', 'synonyms': ['brassiere', 'bra', 'bandeau'], 'id': 148, 'def': 'an undergarment worn by women to support their breasts', 'name': 'brassiere'}, {'frequency': 'c', 'synset': 'bread-bin.n.01', 'synonyms': ['bread-bin', 'breadbox'], 'id': 149, 'def': 'a container used to keep bread or cake in', 'name': 'bread-bin'}, {'frequency': 'f', 'synset': 'bread.n.01', 'synonyms': ['bread'], 'id': 150, 'def': 'food made from dough of flour or meal and usually raised with yeast or baking powder and then baked', 'name': 'bread'}, {'frequency': 'r', 'synset': 'breechcloth.n.01', 'synonyms': ['breechcloth', 'breechclout', 'loincloth'], 'id': 151, 'def': 'a garment that provides covering for the loins', 'name': 'breechcloth'}, {'frequency': 'f', 'synset': 'bridal_gown.n.01', 'synonyms': ['bridal_gown', 'wedding_gown', 'wedding_dress'], 'id': 152, 'def': 'a gown worn by the bride at a wedding', 'name': 'bridal_gown'}, {'frequency': 'c', 'synset': 'briefcase.n.01', 'synonyms': ['briefcase'], 'id': 153, 'def': 'a case with a handle; for carrying papers or files or books', 'name': 'briefcase'}, {'frequency': 'f', 'synset': 'broccoli.n.01', 'synonyms': ['broccoli'], 'id': 154, 'def': 'plant with dense clusters of tight green flower buds', 'name': 'broccoli'}, {'frequency': 'r', 'synset': 'brooch.n.01', 'synonyms': ['broach'], 'id': 155, 'def': 'a decorative pin worn by women', 'name': 'broach'}, {'frequency': 'c', 'synset': 'broom.n.01', 'synonyms': ['broom'], 'id': 156, 'def': 'bundle of straws or twigs attached to a long handle; used for cleaning', 'name': 'broom'}, {'frequency': 'c', 'synset': 'brownie.n.03', 'synonyms': ['brownie'], 'id': 157, 'def': 'square or bar of very rich chocolate cake usually with nuts', 'name': 'brownie'}, {'frequency': 'c', 'synset': 'brussels_sprouts.n.01', 'synonyms': ['brussels_sprouts'], 'id': 158, 'def': 'the small edible cabbage-like buds growing along a stalk', 'name': 'brussels_sprouts'}, {'frequency': 'r', 'synset': 'bubble_gum.n.01', 'synonyms': ['bubble_gum'], 'id': 159, 'def': 'a kind of chewing gum that can be blown into bubbles', 'name': 'bubble_gum'}, {'frequency': 'f', 'synset': 'bucket.n.01', 'synonyms': ['bucket', 'pail'], 'id': 160, 'def': 'a roughly cylindrical vessel that is open at the top', 'name': 'bucket'}, {'frequency': 'r', 'synset': 'buggy.n.01', 'synonyms': ['horse_buggy'], 'id': 161, 'def': 'a small lightweight carriage; drawn by a single horse', 'name': 'horse_buggy'}, {'frequency': 'c', 'synset': 'bull.n.11', 'synonyms': ['horned_cow'], 'id': 162, 'def': 'a cow with horns', 'name': 'bull'}, {'frequency': 'c', 'synset': 'bulldog.n.01', 'synonyms': ['bulldog'], 'id': 163, 'def': 'a thickset short-haired dog with a large head and strong undershot lower jaw', 'name': 'bulldog'}, {'frequency': 'r', 'synset': 'bulldozer.n.01', 'synonyms': ['bulldozer', 'dozer'], 'id': 164, 'def': 'large powerful tractor; a large blade in front flattens areas of ground', 'name': 'bulldozer'}, {'frequency': 'c', 'synset': 'bullet_train.n.01', 'synonyms': ['bullet_train'], 'id': 165, 'def': 'a high-speed passenger train', 'name': 'bullet_train'}, {'frequency': 'c', 'synset': 'bulletin_board.n.02', 'synonyms': ['bulletin_board', 'notice_board'], 'id': 166, 'def': 'a board that hangs on a wall; displays announcements', 'name': 'bulletin_board'}, {'frequency': 'r', 'synset': 'bulletproof_vest.n.01', 'synonyms': ['bulletproof_vest'], 'id': 167, 'def': 'a vest capable of resisting the impact of a bullet', 'name': 'bulletproof_vest'}, {'frequency': 'c', 'synset': 'bullhorn.n.01', 'synonyms': ['bullhorn', 'megaphone'], 'id': 168, 'def': 'a portable loudspeaker with built-in microphone and amplifier', 'name': 'bullhorn'}, {'frequency': 'f', 'synset': 'bun.n.01', 'synonyms': ['bun', 'roll'], 'id': 169, 'def': 'small rounded bread either plain or sweet', 'name': 'bun'}, {'frequency': 'c', 'synset': 'bunk_bed.n.01', 'synonyms': ['bunk_bed'], 'id': 170, 'def': 'beds built one above the other', 'name': 'bunk_bed'}, {'frequency': 'f', 'synset': 'buoy.n.01', 'synonyms': ['buoy'], 'id': 171, 'def': 'a float attached by rope to the seabed to mark channels in a harbor or underwater hazards', 'name': 'buoy'}, {'frequency': 'r', 'synset': 'burrito.n.01', 'synonyms': ['burrito'], 'id': 172, 'def': 'a flour tortilla folded around a filling', 'name': 'burrito'}, {'frequency': 'f', 'synset': 'bus.n.01', 'synonyms': ['bus_(vehicle)', 'autobus', 'charabanc', 'double-decker', 'motorbus', 'motorcoach'], 'id': 173, 'def': 'a vehicle carrying many passengers; used for public transport', 'name': 'bus_(vehicle)'}, {'frequency': 'c', 'synset': 'business_card.n.01', 'synonyms': ['business_card'], 'id': 174, 'def': "a card on which are printed the person's name and business affiliation", 'name': 'business_card'}, {'frequency': 'f', 'synset': 'butter.n.01', 'synonyms': ['butter'], 'id': 175, 'def': 'an edible emulsion of fat globules made by churning milk or cream; for cooking and table use', 'name': 'butter'}, {'frequency': 'c', 'synset': 'butterfly.n.01', 'synonyms': ['butterfly'], 'id': 176, 'def': 'insect typically having a slender body with knobbed antennae and broad colorful wings', 'name': 'butterfly'}, {'frequency': 'f', 'synset': 'button.n.01', 'synonyms': ['button'], 'id': 177, 'def': 'a round fastener sewn to shirts and coats etc to fit through buttonholes', 'name': 'button'}, {'frequency': 'f', 'synset': 'cab.n.03', 'synonyms': ['cab_(taxi)', 'taxi', 'taxicab'], 'id': 178, 'def': 'a car that takes passengers where they want to go in exchange for money', 'name': 'cab_(taxi)'}, {'frequency': 'r', 'synset': 'cabana.n.01', 'synonyms': ['cabana'], 'id': 179, 'def': 'a small tent used as a dressing room beside the sea or a swimming pool', 'name': 'cabana'}, {'frequency': 'c', 'synset': 'cabin_car.n.01', 'synonyms': ['cabin_car', 'caboose'], 'id': 180, 'def': 'a car on a freight train for use of the train crew; usually the last car on the train', 'name': 'cabin_car'}, {'frequency': 'f', 'synset': 'cabinet.n.01', 'synonyms': ['cabinet'], 'id': 181, 'def': 'a piece of furniture resembling a cupboard with doors and shelves and drawers', 'name': 'cabinet'}, {'frequency': 'r', 'synset': 'cabinet.n.03', 'synonyms': ['locker', 'storage_locker'], 'id': 182, 'def': 'a storage compartment for clothes and valuables; usually it has a lock', 'name': 'locker'}, {'frequency': 'f', 'synset': 'cake.n.03', 'synonyms': ['cake'], 'id': 183, 'def': 'baked goods made from or based on a mixture of flour, sugar, eggs, and fat', 'name': 'cake'}, {'frequency': 'c', 'synset': 'calculator.n.02', 'synonyms': ['calculator'], 'id': 184, 'def': 'a small machine that is used for mathematical calculations', 'name': 'calculator'}, {'frequency': 'f', 'synset': 'calendar.n.02', 'synonyms': ['calendar'], 'id': 185, 'def': 'a list or register of events (appointments/social events/court cases, etc)', 'name': 'calendar'}, {'frequency': 'c', 'synset': 'calf.n.01', 'synonyms': ['calf'], 'id': 186, 'def': 'young of domestic cattle', 'name': 'calf'}, {'frequency': 'c', 'synset': 'camcorder.n.01', 'synonyms': ['camcorder'], 'id': 187, 'def': 'a portable television camera and videocassette recorder', 'name': 'camcorder'}, {'frequency': 'c', 'synset': 'camel.n.01', 'synonyms': ['camel'], 'id': 188, 'def': 'cud-chewing mammal used as a draft or saddle animal in desert regions', 'name': 'camel'}, {'frequency': 'f', 'synset': 'camera.n.01', 'synonyms': ['camera'], 'id': 189, 'def': 'equipment for taking photographs', 'name': 'camera'}, {'frequency': 'c', 'synset': 'camera_lens.n.01', 'synonyms': ['camera_lens'], 'id': 190, 'def': 'a lens that focuses the image in a camera', 'name': 'camera_lens'}, {'frequency': 'c', 'synset': 'camper.n.02', 'synonyms': ['camper_(vehicle)', 'camping_bus', 'motor_home'], 'id': 191, 'def': 'a recreational vehicle equipped for camping out while traveling', 'name': 'camper_(vehicle)'}, {'frequency': 'f', 'synset': 'can.n.01', 'synonyms': ['can', 'tin_can'], 'id': 192, 'def': 'airtight sealed metal container for food or drink or paint etc.', 'name': 'can'}, {'frequency': 'c', 'synset': 'can_opener.n.01', 'synonyms': ['can_opener', 'tin_opener'], 'id': 193, 'def': 'a device for cutting cans open', 'name': 'can_opener'}, {'frequency': 'f', 'synset': 'candle.n.01', 'synonyms': ['candle', 'candlestick'], 'id': 194, 'def': 'stick of wax with a wick in the middle', 'name': 'candle'}, {'frequency': 'f', 'synset': 'candlestick.n.01', 'synonyms': ['candle_holder'], 'id': 195, 'def': 'a holder with sockets for candles', 'name': 'candle_holder'}, {'frequency': 'r', 'synset': 'candy_bar.n.01', 'synonyms': ['candy_bar'], 'id': 196, 'def': 'a candy shaped as a bar', 'name': 'candy_bar'}, {'frequency': 'c', 'synset': 'candy_cane.n.01', 'synonyms': ['candy_cane'], 'id': 197, 'def': 'a hard candy in the shape of a rod (usually with stripes)', 'name': 'candy_cane'}, {'frequency': 'c', 'synset': 'cane.n.01', 'synonyms': ['walking_cane'], 'id': 198, 'def': 'a stick that people can lean on to help them walk', 'name': 'walking_cane'}, {'frequency': 'c', 'synset': 'canister.n.02', 'synonyms': ['canister', 'cannister'], 'id': 199, 'def': 'metal container for storing dry foods such as tea or flour', 'name': 'canister'}, {'frequency': 'c', 'synset': 'canoe.n.01', 'synonyms': ['canoe'], 'id': 200, 'def': 'small and light boat; pointed at both ends; propelled with a paddle', 'name': 'canoe'}, {'frequency': 'c', 'synset': 'cantaloup.n.02', 'synonyms': ['cantaloup', 'cantaloupe'], 'id': 201, 'def': 'the fruit of a cantaloup vine; small to medium-sized melon with yellowish flesh', 'name': 'cantaloup'}, {'frequency': 'r', 'synset': 'canteen.n.01', 'synonyms': ['canteen'], 'id': 202, 'def': 'a flask for carrying water; used by soldiers or travelers', 'name': 'canteen'}, {'frequency': 'f', 'synset': 'cap.n.01', 'synonyms': ['cap_(headwear)'], 'id': 203, 'def': 'a tight-fitting headwear', 'name': 'cap_(headwear)'}, {'frequency': 'f', 'synset': 'cap.n.02', 'synonyms': ['bottle_cap', 'cap_(container_lid)'], 'id': 204, 'def': 'a top (as for a bottle)', 'name': 'bottle_cap'}, {'frequency': 'c', 'synset': 'cape.n.02', 'synonyms': ['cape'], 'id': 205, 'def': 'a sleeveless garment like a cloak but shorter', 'name': 'cape'}, {'frequency': 'c', 'synset': 'cappuccino.n.01', 'synonyms': ['cappuccino', 'coffee_cappuccino'], 'id': 206, 'def': 'equal parts of espresso and steamed milk', 'name': 'cappuccino'}, {'frequency': 'f', 'synset': 'car.n.01', 'synonyms': ['car_(automobile)', 'auto_(automobile)', 'automobile'], 'id': 207, 'def': 'a motor vehicle with four wheels', 'name': 'car_(automobile)'}, {'frequency': 'f', 'synset': 'car.n.02', 'synonyms': ['railcar_(part_of_a_train)', 'railway_car_(part_of_a_train)', 'railroad_car_(part_of_a_train)'], 'id': 208, 'def': 'a wheeled vehicle adapted to the rails of railroad (mark each individual railcar separately)', 'name': 'railcar_(part_of_a_train)'}, {'frequency': 'r', 'synset': 'car.n.04', 'synonyms': ['elevator_car'], 'id': 209, 'def': 'where passengers ride up and down', 'name': 'elevator_car'}, {'frequency': 'r', 'synset': 'car_battery.n.01', 'synonyms': ['car_battery', 'automobile_battery'], 'id': 210, 'def': 'a battery in a motor vehicle', 'name': 'car_battery'}, {'frequency': 'c', 'synset': 'card.n.02', 'synonyms': ['identity_card'], 'id': 211, 'def': 'a card certifying the identity of the bearer', 'name': 'identity_card'}, {'frequency': 'c', 'synset': 'card.n.03', 'synonyms': ['card'], 'id': 212, 'def': 'a rectangular piece of paper used to send messages (e.g. greetings or pictures)', 'name': 'card'}, {'frequency': 'c', 'synset': 'cardigan.n.01', 'synonyms': ['cardigan'], 'id': 213, 'def': 'knitted jacket that is fastened up the front with buttons or a zipper', 'name': 'cardigan'}, {'frequency': 'r', 'synset': 'cargo_ship.n.01', 'synonyms': ['cargo_ship', 'cargo_vessel'], 'id': 214, 'def': 'a ship designed to carry cargo', 'name': 'cargo_ship'}, {'frequency': 'r', 'synset': 'carnation.n.01', 'synonyms': ['carnation'], 'id': 215, 'def': 'plant with pink to purple-red spice-scented usually double flowers', 'name': 'carnation'}, {'frequency': 'c', 'synset': 'carriage.n.02', 'synonyms': ['horse_carriage'], 'id': 216, 'def': 'a vehicle with wheels drawn by one or more horses', 'name': 'horse_carriage'}, {'frequency': 'f', 'synset': 'carrot.n.01', 'synonyms': ['carrot'], 'id': 217, 'def': 'deep orange edible root of the cultivated carrot plant', 'name': 'carrot'}, {'frequency': 'f', 'synset': 'carryall.n.01', 'synonyms': ['tote_bag'], 'id': 218, 'def': 'a capacious bag or basket', 'name': 'tote_bag'}, {'frequency': 'c', 'synset': 'cart.n.01', 'synonyms': ['cart'], 'id': 219, 'def': 'a heavy open wagon usually having two wheels and drawn by an animal', 'name': 'cart'}, {'frequency': 'c', 'synset': 'carton.n.02', 'synonyms': ['carton'], 'id': 220, 'def': 'a container made of cardboard for holding food or drink', 'name': 'carton'}, {'frequency': 'c', 'synset': 'cash_register.n.01', 'synonyms': ['cash_register', 'register_(for_cash_transactions)'], 'id': 221, 'def': 'a cashbox with an adding machine to register transactions', 'name': 'cash_register'}, {'frequency': 'r', 'synset': 'casserole.n.01', 'synonyms': ['casserole'], 'id': 222, 'def': 'food cooked and served in a casserole', 'name': 'casserole'}, {'frequency': 'r', 'synset': 'cassette.n.01', 'synonyms': ['cassette'], 'id': 223, 'def': 'a container that holds a magnetic tape used for recording or playing sound or video', 'name': 'cassette'}, {'frequency': 'c', 'synset': 'cast.n.05', 'synonyms': ['cast', 'plaster_cast', 'plaster_bandage'], 'id': 224, 'def': 'bandage consisting of a firm covering that immobilizes broken bones while they heal', 'name': 'cast'}, {'frequency': 'f', 'synset': 'cat.n.01', 'synonyms': ['cat'], 'id': 225, 'def': 'a domestic house cat', 'name': 'cat'}, {'frequency': 'f', 'synset': 'cauliflower.n.02', 'synonyms': ['cauliflower'], 'id': 226, 'def': 'edible compact head of white undeveloped flowers', 'name': 'cauliflower'}, {'frequency': 'c', 'synset': 'cayenne.n.02', 'synonyms': ['cayenne_(spice)', 'cayenne_pepper_(spice)', 'red_pepper_(spice)'], 'id': 227, 'def': 'ground pods and seeds of pungent red peppers of the genus Capsicum', 'name': 'cayenne_(spice)'}, {'frequency': 'c', 'synset': 'cd_player.n.01', 'synonyms': ['CD_player'], 'id': 228, 'def': 'electronic equipment for playing compact discs (CDs)', 'name': 'CD_player'}, {'frequency': 'f', 'synset': 'celery.n.01', 'synonyms': ['celery'], 'id': 229, 'def': 'widely cultivated herb with aromatic leaf stalks that are eaten raw or cooked', 'name': 'celery'}, {'frequency': 'f', 'synset': 'cellular_telephone.n.01', 'synonyms': ['cellular_telephone', 'cellular_phone', 'cellphone', 'mobile_phone', 'smart_phone'], 'id': 230, 'def': 'a hand-held mobile telephone', 'name': 'cellular_telephone'}, {'frequency': 'r', 'synset': 'chain_mail.n.01', 'synonyms': ['chain_mail', 'ring_mail', 'chain_armor', 'chain_armour', 'ring_armor', 'ring_armour'], 'id': 231, 'def': '(Middle Ages) flexible armor made of interlinked metal rings', 'name': 'chain_mail'}, {'frequency': 'f', 'synset': 'chair.n.01', 'synonyms': ['chair'], 'id': 232, 'def': 'a seat for one person, with a support for the back', 'name': 'chair'}, {'frequency': 'r', 'synset': 'chaise_longue.n.01', 'synonyms': ['chaise_longue', 'chaise', 'daybed'], 'id': 233, 'def': 'a long chair; for reclining', 'name': 'chaise_longue'}, {'frequency': 'r', 'synset': 'chalice.n.01', 'synonyms': ['chalice'], 'id': 234, 'def': 'a bowl-shaped drinking vessel; especially the Eucharistic cup', 'name': 'chalice'}, {'frequency': 'f', 'synset': 'chandelier.n.01', 'synonyms': ['chandelier'], 'id': 235, 'def': 'branched lighting fixture; often ornate; hangs from the ceiling', 'name': 'chandelier'}, {'frequency': 'r', 'synset': 'chap.n.04', 'synonyms': ['chap'], 'id': 236, 'def': 'leather leggings without a seat; worn over trousers by cowboys to protect their legs', 'name': 'chap'}, {'frequency': 'r', 'synset': 'checkbook.n.01', 'synonyms': ['checkbook', 'chequebook'], 'id': 237, 'def': 'a book issued to holders of checking accounts', 'name': 'checkbook'}, {'frequency': 'r', 'synset': 'checkerboard.n.01', 'synonyms': ['checkerboard'], 'id': 238, 'def': 'a board having 64 squares of two alternating colors', 'name': 'checkerboard'}, {'frequency': 'c', 'synset': 'cherry.n.03', 'synonyms': ['cherry'], 'id': 239, 'def': 'a red fruit with a single hard stone', 'name': 'cherry'}, {'frequency': 'r', 'synset': 'chessboard.n.01', 'synonyms': ['chessboard'], 'id': 240, 'def': 'a checkerboard used to play chess', 'name': 'chessboard'}, {'frequency': 'c', 'synset': 'chicken.n.02', 'synonyms': ['chicken_(animal)'], 'id': 241, 'def': 'a domestic fowl bred for flesh or eggs', 'name': 'chicken_(animal)'}, {'frequency': 'c', 'synset': 'chickpea.n.01', 'synonyms': ['chickpea', 'garbanzo'], 'id': 242, 'def': 'the seed of the chickpea plant; usually dried', 'name': 'chickpea'}, {'frequency': 'c', 'synset': 'chili.n.02', 'synonyms': ['chili_(vegetable)', 'chili_pepper_(vegetable)', 'chilli_(vegetable)', 'chilly_(vegetable)', 'chile_(vegetable)'], 'id': 243, 'def': 'very hot and finely tapering pepper of special pungency', 'name': 'chili_(vegetable)'}, {'frequency': 'r', 'synset': 'chime.n.01', 'synonyms': ['chime', 'gong'], 'id': 244, 'def': 'an instrument consisting of a set of bells that are struck with a hammer', 'name': 'chime'}, {'frequency': 'r', 'synset': 'chinaware.n.01', 'synonyms': ['chinaware'], 'id': 245, 'def': 'dishware made of high quality porcelain', 'name': 'chinaware'}, {'frequency': 'c', 'synset': 'chip.n.04', 'synonyms': ['crisp_(potato_chip)', 'potato_chip'], 'id': 246, 'def': 'a thin crisp slice of potato fried in deep fat', 'name': 'crisp_(potato_chip)'}, {'frequency': 'r', 'synset': 'chip.n.06', 'synonyms': ['poker_chip'], 'id': 247, 'def': 'a small disk-shaped counter used to represent money when gambling', 'name': 'poker_chip'}, {'frequency': 'c', 'synset': 'chocolate_bar.n.01', 'synonyms': ['chocolate_bar'], 'id': 248, 'def': 'a bar of chocolate candy', 'name': 'chocolate_bar'}, {'frequency': 'c', 'synset': 'chocolate_cake.n.01', 'synonyms': ['chocolate_cake'], 'id': 249, 'def': 'cake containing chocolate', 'name': 'chocolate_cake'}, {'frequency': 'r', 'synset': 'chocolate_milk.n.01', 'synonyms': ['chocolate_milk'], 'id': 250, 'def': 'milk flavored with chocolate syrup', 'name': 'chocolate_milk'}, {'frequency': 'r', 'synset': 'chocolate_mousse.n.01', 'synonyms': ['chocolate_mousse'], 'id': 251, 'def': 'dessert mousse made with chocolate', 'name': 'chocolate_mousse'}, {'frequency': 'f', 'synset': 'choker.n.03', 'synonyms': ['choker', 'collar', 'neckband'], 'id': 252, 'def': 'shirt collar, animal collar, or tight-fitting necklace', 'name': 'choker'}, {'frequency': 'f', 'synset': 'chopping_board.n.01', 'synonyms': ['chopping_board', 'cutting_board', 'chopping_block'], 'id': 253, 'def': 'a wooden board where meats or vegetables can be cut', 'name': 'chopping_board'}, {'frequency': 'f', 'synset': 'chopstick.n.01', 'synonyms': ['chopstick'], 'id': 254, 'def': 'one of a pair of slender sticks used as oriental tableware to eat food with', 'name': 'chopstick'}, {'frequency': 'f', 'synset': 'christmas_tree.n.05', 'synonyms': ['Christmas_tree'], 'id': 255, 'def': 'an ornamented evergreen used as a Christmas decoration', 'name': 'Christmas_tree'}, {'frequency': 'c', 'synset': 'chute.n.02', 'synonyms': ['slide'], 'id': 256, 'def': 'sloping channel through which things can descend', 'name': 'slide'}, {'frequency': 'r', 'synset': 'cider.n.01', 'synonyms': ['cider', 'cyder'], 'id': 257, 'def': 'a beverage made from juice pressed from apples', 'name': 'cider'}, {'frequency': 'r', 'synset': 'cigar_box.n.01', 'synonyms': ['cigar_box'], 'id': 258, 'def': 'a box for holding cigars', 'name': 'cigar_box'}, {'frequency': 'f', 'synset': 'cigarette.n.01', 'synonyms': ['cigarette'], 'id': 259, 'def': 'finely ground tobacco wrapped in paper; for smoking', 'name': 'cigarette'}, {'frequency': 'c', 'synset': 'cigarette_case.n.01', 'synonyms': ['cigarette_case', 'cigarette_pack'], 'id': 260, 'def': 'a small flat case for holding cigarettes', 'name': 'cigarette_case'}, {'frequency': 'f', 'synset': 'cistern.n.02', 'synonyms': ['cistern', 'water_tank'], 'id': 261, 'def': 'a tank that holds the water used to flush a toilet', 'name': 'cistern'}, {'frequency': 'r', 'synset': 'clarinet.n.01', 'synonyms': ['clarinet'], 'id': 262, 'def': 'a single-reed instrument with a straight tube', 'name': 'clarinet'}, {'frequency': 'c', 'synset': 'clasp.n.01', 'synonyms': ['clasp'], 'id': 263, 'def': 'a fastener (as a buckle or hook) that is used to hold two things together', 'name': 'clasp'}, {'frequency': 'c', 'synset': 'cleansing_agent.n.01', 'synonyms': ['cleansing_agent', 'cleanser', 'cleaner'], 'id': 264, 'def': 'a preparation used in cleaning something', 'name': 'cleansing_agent'}, {'frequency': 'r', 'synset': 'cleat.n.02', 'synonyms': ['cleat_(for_securing_rope)'], 'id': 265, 'def': 'a fastener (usually with two projecting horns) around which a rope can be secured', 'name': 'cleat_(for_securing_rope)'}, {'frequency': 'r', 'synset': 'clementine.n.01', 'synonyms': ['clementine'], 'id': 266, 'def': 'a variety of mandarin orange', 'name': 'clementine'}, {'frequency': 'c', 'synset': 'clip.n.03', 'synonyms': ['clip'], 'id': 267, 'def': 'any of various small fasteners used to hold loose articles together', 'name': 'clip'}, {'frequency': 'c', 'synset': 'clipboard.n.01', 'synonyms': ['clipboard'], 'id': 268, 'def': 'a small writing board with a clip at the top for holding papers', 'name': 'clipboard'}, {'frequency': 'r', 'synset': 'clipper.n.03', 'synonyms': ['clippers_(for_plants)'], 'id': 269, 'def': 'shears for cutting grass or shrubbery (often used in the plural)', 'name': 'clippers_(for_plants)'}, {'frequency': 'r', 'synset': 'cloak.n.02', 'synonyms': ['cloak'], 'id': 270, 'def': 'a loose outer garment', 'name': 'cloak'}, {'frequency': 'f', 'synset': 'clock.n.01', 'synonyms': ['clock', 'timepiece', 'timekeeper'], 'id': 271, 'def': 'a timepiece that shows the time of day', 'name': 'clock'}, {'frequency': 'f', 'synset': 'clock_tower.n.01', 'synonyms': ['clock_tower'], 'id': 272, 'def': 'a tower with a large clock visible high up on an outside face', 'name': 'clock_tower'}, {'frequency': 'c', 'synset': 'clothes_hamper.n.01', 'synonyms': ['clothes_hamper', 'laundry_basket', 'clothes_basket'], 'id': 273, 'def': 'a hamper that holds dirty clothes to be washed or wet clothes to be dried', 'name': 'clothes_hamper'}, {'frequency': 'c', 'synset': 'clothespin.n.01', 'synonyms': ['clothespin', 'clothes_peg'], 'id': 274, 'def': 'wood or plastic fastener; for holding clothes on a clothesline', 'name': 'clothespin'}, {'frequency': 'r', 'synset': 'clutch_bag.n.01', 'synonyms': ['clutch_bag'], 'id': 275, 'def': "a woman's strapless purse that is carried in the hand", 'name': 'clutch_bag'}, {'frequency': 'f', 'synset': 'coaster.n.03', 'synonyms': ['coaster'], 'id': 276, 'def': 'a covering (plate or mat) that protects the surface of a table', 'name': 'coaster'}, {'frequency': 'f', 'synset': 'coat.n.01', 'synonyms': ['coat'], 'id': 277, 'def': 'an outer garment that has sleeves and covers the body from shoulder down', 'name': 'coat'}, {'frequency': 'c', 'synset': 'coat_hanger.n.01', 'synonyms': ['coat_hanger', 'clothes_hanger', 'dress_hanger'], 'id': 278, 'def': "a hanger that is shaped like a person's shoulders", 'name': 'coat_hanger'}, {'frequency': 'c', 'synset': 'coatrack.n.01', 'synonyms': ['coatrack', 'hatrack'], 'id': 279, 'def': 'a rack with hooks for temporarily holding coats and hats', 'name': 'coatrack'}, {'frequency': 'c', 'synset': 'cock.n.04', 'synonyms': ['cock', 'rooster'], 'id': 280, 'def': 'adult male chicken', 'name': 'cock'}, {'frequency': 'r', 'synset': 'cockroach.n.01', 'synonyms': ['cockroach'], 'id': 281, 'def': 'any of numerous chiefly nocturnal insects; some are domestic pests', 'name': 'cockroach'}, {'frequency': 'r', 'synset': 'cocoa.n.01', 'synonyms': ['cocoa_(beverage)', 'hot_chocolate_(beverage)', 'drinking_chocolate'], 'id': 282, 'def': 'a beverage made from cocoa powder and milk and sugar; usually drunk hot', 'name': 'cocoa_(beverage)'}, {'frequency': 'c', 'synset': 'coconut.n.02', 'synonyms': ['coconut', 'cocoanut'], 'id': 283, 'def': 'large hard-shelled brown oval nut with a fibrous husk', 'name': 'coconut'}, {'frequency': 'f', 'synset': 'coffee_maker.n.01', 'synonyms': ['coffee_maker', 'coffee_machine'], 'id': 284, 'def': 'a kitchen appliance for brewing coffee automatically', 'name': 'coffee_maker'}, {'frequency': 'f', 'synset': 'coffee_table.n.01', 'synonyms': ['coffee_table', 'cocktail_table'], 'id': 285, 'def': 'low table where magazines can be placed and coffee or cocktails are served', 'name': 'coffee_table'}, {'frequency': 'c', 'synset': 'coffeepot.n.01', 'synonyms': ['coffeepot'], 'id': 286, 'def': 'tall pot in which coffee is brewed', 'name': 'coffeepot'}, {'frequency': 'r', 'synset': 'coil.n.05', 'synonyms': ['coil'], 'id': 287, 'def': 'tubing that is wound in a spiral', 'name': 'coil'}, {'frequency': 'c', 'synset': 'coin.n.01', 'synonyms': ['coin'], 'id': 288, 'def': 'a flat metal piece (usually a disc) used as money', 'name': 'coin'}, {'frequency': 'c', 'synset': 'colander.n.01', 'synonyms': ['colander', 'cullender'], 'id': 289, 'def': 'bowl-shaped strainer; used to wash or drain foods', 'name': 'colander'}, {'frequency': 'c', 'synset': 'coleslaw.n.01', 'synonyms': ['coleslaw', 'slaw'], 'id': 290, 'def': 'basically shredded cabbage', 'name': 'coleslaw'}, {'frequency': 'r', 'synset': 'coloring_material.n.01', 'synonyms': ['coloring_material', 'colouring_material'], 'id': 291, 'def': 'any material used for its color', 'name': 'coloring_material'}, {'frequency': 'r', 'synset': 'combination_lock.n.01', 'synonyms': ['combination_lock'], 'id': 292, 'def': 'lock that can be opened only by turning dials in a special sequence', 'name': 'combination_lock'}, {'frequency': 'c', 'synset': 'comforter.n.04', 'synonyms': ['pacifier', 'teething_ring'], 'id': 293, 'def': 'device used for an infant to suck or bite on', 'name': 'pacifier'}, {'frequency': 'r', 'synset': 'comic_book.n.01', 'synonyms': ['comic_book'], 'id': 294, 'def': 'a magazine devoted to comic strips', 'name': 'comic_book'}, {'frequency': 'r', 'synset': 'compass.n.01', 'synonyms': ['compass'], 'id': 295, 'def': 'navigational instrument for finding directions', 'name': 'compass'}, {'frequency': 'f', 'synset': 'computer_keyboard.n.01', 'synonyms': ['computer_keyboard', 'keyboard_(computer)'], 'id': 296, 'def': 'a keyboard that is a data input device for computers', 'name': 'computer_keyboard'}, {'frequency': 'f', 'synset': 'condiment.n.01', 'synonyms': ['condiment'], 'id': 297, 'def': 'a preparation (a sauce or relish or spice) to enhance flavor or enjoyment', 'name': 'condiment'}, {'frequency': 'f', 'synset': 'cone.n.01', 'synonyms': ['cone', 'traffic_cone'], 'id': 298, 'def': 'a cone-shaped object used to direct traffic', 'name': 'cone'}, {'frequency': 'f', 'synset': 'control.n.09', 'synonyms': ['control', 'controller'], 'id': 299, 'def': 'a mechanism that controls the operation of a machine', 'name': 'control'}, {'frequency': 'r', 'synset': 'convertible.n.01', 'synonyms': ['convertible_(automobile)'], 'id': 300, 'def': 'a car that has top that can be folded or removed', 'name': 'convertible_(automobile)'}, {'frequency': 'r', 'synset': 'convertible.n.03', 'synonyms': ['sofa_bed'], 'id': 301, 'def': 'a sofa that can be converted into a bed', 'name': 'sofa_bed'}, {'frequency': 'r', 'synset': 'cooker.n.01', 'synonyms': ['cooker'], 'id': 302, 'def': 'a utensil for cooking', 'name': 'cooker'}, {'frequency': 'f', 'synset': 'cookie.n.01', 'synonyms': ['cookie', 'cooky', 'biscuit_(cookie)'], 'id': 303, 'def': "any of various small flat sweet cakes (`biscuit' is the British term)", 'name': 'cookie'}, {'frequency': 'r', 'synset': 'cooking_utensil.n.01', 'synonyms': ['cooking_utensil'], 'id': 304, 'def': 'a kitchen utensil made of material that does not melt easily; used for cooking', 'name': 'cooking_utensil'}, {'frequency': 'f', 'synset': 'cooler.n.01', 'synonyms': ['cooler_(for_food)', 'ice_chest'], 'id': 305, 'def': 'an insulated box for storing food often with ice', 'name': 'cooler_(for_food)'}, {'frequency': 'f', 'synset': 'cork.n.04', 'synonyms': ['cork_(bottle_plug)', 'bottle_cork'], 'id': 306, 'def': 'the plug in the mouth of a bottle (especially a wine bottle)', 'name': 'cork_(bottle_plug)'}, {'frequency': 'r', 'synset': 'corkboard.n.01', 'synonyms': ['corkboard'], 'id': 307, 'def': 'a sheet consisting of cork granules', 'name': 'corkboard'}, {'frequency': 'c', 'synset': 'corkscrew.n.01', 'synonyms': ['corkscrew', 'bottle_screw'], 'id': 308, 'def': 'a bottle opener that pulls corks', 'name': 'corkscrew'}, {'frequency': 'f', 'synset': 'corn.n.03', 'synonyms': ['edible_corn', 'corn', 'maize'], 'id': 309, 'def': 'ears or kernels of corn that can be prepared and served for human food (only mark individual ears or kernels)', 'name': 'edible_corn'}, {'frequency': 'r', 'synset': 'cornbread.n.01', 'synonyms': ['cornbread'], 'id': 310, 'def': 'bread made primarily of cornmeal', 'name': 'cornbread'}, {'frequency': 'c', 'synset': 'cornet.n.01', 'synonyms': ['cornet', 'horn', 'trumpet'], 'id': 311, 'def': 'a brass musical instrument with a narrow tube and a flared bell and many valves', 'name': 'cornet'}, {'frequency': 'c', 'synset': 'cornice.n.01', 'synonyms': ['cornice', 'valance', 'valance_board', 'pelmet'], 'id': 312, 'def': 'a decorative framework to conceal curtain fixtures at the top of a window casing', 'name': 'cornice'}, {'frequency': 'r', 'synset': 'cornmeal.n.01', 'synonyms': ['cornmeal'], 'id': 313, 'def': 'coarsely ground corn', 'name': 'cornmeal'}, {'frequency': 'c', 'synset': 'corset.n.01', 'synonyms': ['corset', 'girdle'], 'id': 314, 'def': "a woman's close-fitting foundation garment", 'name': 'corset'}, {'frequency': 'c', 'synset': 'costume.n.04', 'synonyms': ['costume'], 'id': 315, 'def': 'the attire characteristic of a country or a time or a social class', 'name': 'costume'}, {'frequency': 'r', 'synset': 'cougar.n.01', 'synonyms': ['cougar', 'puma', 'catamount', 'mountain_lion', 'panther'], 'id': 316, 'def': 'large American feline resembling a lion', 'name': 'cougar'}, {'frequency': 'r', 'synset': 'coverall.n.01', 'synonyms': ['coverall'], 'id': 317, 'def': 'a loose-fitting protective garment that is worn over other clothing', 'name': 'coverall'}, {'frequency': 'c', 'synset': 'cowbell.n.01', 'synonyms': ['cowbell'], 'id': 318, 'def': 'a bell hung around the neck of cow so that the cow can be easily located', 'name': 'cowbell'}, {'frequency': 'f', 'synset': 'cowboy_hat.n.01', 'synonyms': ['cowboy_hat', 'ten-gallon_hat'], 'id': 319, 'def': 'a hat with a wide brim and a soft crown; worn by American ranch hands', 'name': 'cowboy_hat'}, {'frequency': 'c', 'synset': 'crab.n.01', 'synonyms': ['crab_(animal)'], 'id': 320, 'def': 'decapod having eyes on short stalks and a broad flattened shell and pincers', 'name': 'crab_(animal)'}, {'frequency': 'r', 'synset': 'crab.n.05', 'synonyms': ['crabmeat'], 'id': 321, 'def': 'the edible flesh of any of various crabs', 'name': 'crabmeat'}, {'frequency': 'c', 'synset': 'cracker.n.01', 'synonyms': ['cracker'], 'id': 322, 'def': 'a thin crisp wafer', 'name': 'cracker'}, {'frequency': 'r', 'synset': 'crape.n.01', 'synonyms': ['crape', 'crepe', 'French_pancake'], 'id': 323, 'def': 'small very thin pancake', 'name': 'crape'}, {'frequency': 'f', 'synset': 'crate.n.01', 'synonyms': ['crate'], 'id': 324, 'def': 'a rugged box (usually made of wood); used for shipping', 'name': 'crate'}, {'frequency': 'c', 'synset': 'crayon.n.01', 'synonyms': ['crayon', 'wax_crayon'], 'id': 325, 'def': 'writing or drawing implement made of a colored stick of composition wax', 'name': 'crayon'}, {'frequency': 'r', 'synset': 'cream_pitcher.n.01', 'synonyms': ['cream_pitcher'], 'id': 326, 'def': 'a small pitcher for serving cream', 'name': 'cream_pitcher'}, {'frequency': 'c', 'synset': 'crescent_roll.n.01', 'synonyms': ['crescent_roll', 'croissant'], 'id': 327, 'def': 'very rich flaky crescent-shaped roll', 'name': 'crescent_roll'}, {'frequency': 'c', 'synset': 'crib.n.01', 'synonyms': ['crib', 'cot'], 'id': 328, 'def': 'baby bed with high sides made of slats', 'name': 'crib'}, {'frequency': 'c', 'synset': 'crock.n.03', 'synonyms': ['crock_pot', 'earthenware_jar'], 'id': 329, 'def': 'an earthen jar (made of baked clay) or a modern electric crockpot', 'name': 'crock_pot'}, {'frequency': 'f', 'synset': 'crossbar.n.01', 'synonyms': ['crossbar'], 'id': 330, 'def': 'a horizontal bar that goes across something', 'name': 'crossbar'}, {'frequency': 'r', 'synset': 'crouton.n.01', 'synonyms': ['crouton'], 'id': 331, 'def': 'a small piece of toasted or fried bread; served in soup or salads', 'name': 'crouton'}, {'frequency': 'c', 'synset': 'crow.n.01', 'synonyms': ['crow'], 'id': 332, 'def': 'black birds having a raucous call', 'name': 'crow'}, {'frequency': 'r', 'synset': 'crowbar.n.01', 'synonyms': ['crowbar', 'wrecking_bar', 'pry_bar'], 'id': 333, 'def': 'a heavy iron lever with one end forged into a wedge', 'name': 'crowbar'}, {'frequency': 'c', 'synset': 'crown.n.04', 'synonyms': ['crown'], 'id': 334, 'def': 'an ornamental jeweled headdress signifying sovereignty', 'name': 'crown'}, {'frequency': 'c', 'synset': 'crucifix.n.01', 'synonyms': ['crucifix'], 'id': 335, 'def': 'representation of the cross on which Jesus died', 'name': 'crucifix'}, {'frequency': 'c', 'synset': 'cruise_ship.n.01', 'synonyms': ['cruise_ship', 'cruise_liner'], 'id': 336, 'def': 'a passenger ship used commercially for pleasure cruises', 'name': 'cruise_ship'}, {'frequency': 'c', 'synset': 'cruiser.n.01', 'synonyms': ['police_cruiser', 'patrol_car', 'police_car', 'squad_car'], 'id': 337, 'def': 'a car in which policemen cruise the streets', 'name': 'police_cruiser'}, {'frequency': 'f', 'synset': 'crumb.n.03', 'synonyms': ['crumb'], 'id': 338, 'def': 'small piece of e.g. bread or cake', 'name': 'crumb'}, {'frequency': 'c', 'synset': 'crutch.n.01', 'synonyms': ['crutch'], 'id': 339, 'def': 'a wooden or metal staff that fits under the armpit and reaches to the ground', 'name': 'crutch'}, {'frequency': 'c', 'synset': 'cub.n.03', 'synonyms': ['cub_(animal)'], 'id': 340, 'def': 'the young of certain carnivorous mammals such as the bear or wolf or lion', 'name': 'cub_(animal)'}, {'frequency': 'c', 'synset': 'cube.n.05', 'synonyms': ['cube', 'square_block'], 'id': 341, 'def': 'a block in the (approximate) shape of a cube', 'name': 'cube'}, {'frequency': 'f', 'synset': 'cucumber.n.02', 'synonyms': ['cucumber', 'cuke'], 'id': 342, 'def': 'cylindrical green fruit with thin green rind and white flesh eaten as a vegetable', 'name': 'cucumber'}, {'frequency': 'c', 'synset': 'cufflink.n.01', 'synonyms': ['cufflink'], 'id': 343, 'def': 'jewelry consisting of linked buttons used to fasten the cuffs of a shirt', 'name': 'cufflink'}, {'frequency': 'f', 'synset': 'cup.n.01', 'synonyms': ['cup'], 'id': 344, 'def': 'a small open container usually used for drinking; usually has a handle', 'name': 'cup'}, {'frequency': 'c', 'synset': 'cup.n.08', 'synonyms': ['trophy_cup'], 'id': 345, 'def': 'a metal award or cup-shaped vessel with handles that is awarded as a trophy to a competition winner', 'name': 'trophy_cup'}, {'frequency': 'f', 'synset': 'cupboard.n.01', 'synonyms': ['cupboard', 'closet'], 'id': 346, 'def': 'a small room (or recess) or cabinet used for storage space', 'name': 'cupboard'}, {'frequency': 'f', 'synset': 'cupcake.n.01', 'synonyms': ['cupcake'], 'id': 347, 'def': 'small cake baked in a muffin tin', 'name': 'cupcake'}, {'frequency': 'r', 'synset': 'curler.n.01', 'synonyms': ['hair_curler', 'hair_roller', 'hair_crimper'], 'id': 348, 'def': 'a cylindrical tube around which the hair is wound to curl it', 'name': 'hair_curler'}, {'frequency': 'r', 'synset': 'curling_iron.n.01', 'synonyms': ['curling_iron'], 'id': 349, 'def': 'a cylindrical home appliance that heats hair that has been curled around it', 'name': 'curling_iron'}, {'frequency': 'f', 'synset': 'curtain.n.01', 'synonyms': ['curtain', 'drapery'], 'id': 350, 'def': 'hanging cloth used as a blind (especially for a window)', 'name': 'curtain'}, {'frequency': 'f', 'synset': 'cushion.n.03', 'synonyms': ['cushion'], 'id': 351, 'def': 'a soft bag filled with air or padding such as feathers or foam rubber', 'name': 'cushion'}, {'frequency': 'r', 'synset': 'cylinder.n.04', 'synonyms': ['cylinder'], 'id': 352, 'def': 'a cylindrical container', 'name': 'cylinder'}, {'frequency': 'r', 'synset': 'cymbal.n.01', 'synonyms': ['cymbal'], 'id': 353, 'def': 'a percussion instrument consisting of a concave brass disk', 'name': 'cymbal'}, {'frequency': 'r', 'synset': 'dagger.n.01', 'synonyms': ['dagger'], 'id': 354, 'def': 'a short knife with a pointed blade used for piercing or stabbing', 'name': 'dagger'}, {'frequency': 'r', 'synset': 'dalmatian.n.02', 'synonyms': ['dalmatian'], 'id': 355, 'def': 'a large breed having a smooth white coat with black or brown spots', 'name': 'dalmatian'}, {'frequency': 'c', 'synset': 'dartboard.n.01', 'synonyms': ['dartboard'], 'id': 356, 'def': 'a circular board of wood or cork used as the target in the game of darts', 'name': 'dartboard'}, {'frequency': 'r', 'synset': 'date.n.08', 'synonyms': ['date_(fruit)'], 'id': 357, 'def': 'sweet edible fruit of the date palm with a single long woody seed', 'name': 'date_(fruit)'}, {'frequency': 'f', 'synset': 'deck_chair.n.01', 'synonyms': ['deck_chair', 'beach_chair'], 'id': 358, 'def': 'a folding chair for use outdoors; a wooden frame supports a length of canvas', 'name': 'deck_chair'}, {'frequency': 'c', 'synset': 'deer.n.01', 'synonyms': ['deer', 'cervid'], 'id': 359, 'def': "distinguished from Bovidae by the male's having solid deciduous antlers", 'name': 'deer'}, {'frequency': 'c', 'synset': 'dental_floss.n.01', 'synonyms': ['dental_floss', 'floss'], 'id': 360, 'def': 'a soft thread for cleaning the spaces between the teeth', 'name': 'dental_floss'}, {'frequency': 'f', 'synset': 'desk.n.01', 'synonyms': ['desk'], 'id': 361, 'def': 'a piece of furniture with a writing surface and usually drawers or other compartments', 'name': 'desk'}, {'frequency': 'r', 'synset': 'detergent.n.01', 'synonyms': ['detergent'], 'id': 362, 'def': 'a surface-active chemical widely used in industry and laundering', 'name': 'detergent'}, {'frequency': 'c', 'synset': 'diaper.n.01', 'synonyms': ['diaper'], 'id': 363, 'def': 'garment consisting of a folded cloth drawn up between the legs and fastened at the waist', 'name': 'diaper'}, {'frequency': 'r', 'synset': 'diary.n.01', 'synonyms': ['diary', 'journal'], 'id': 364, 'def': 'yearly planner book', 'name': 'diary'}, {'frequency': 'r', 'synset': 'die.n.01', 'synonyms': ['die', 'dice'], 'id': 365, 'def': 'a small cube with 1 to 6 spots on the six faces; used in gambling', 'name': 'die'}, {'frequency': 'r', 'synset': 'dinghy.n.01', 'synonyms': ['dinghy', 'dory', 'rowboat'], 'id': 366, 'def': 'a small boat of shallow draft with seats and oars with which it is propelled', 'name': 'dinghy'}, {'frequency': 'f', 'synset': 'dining_table.n.01', 'synonyms': ['dining_table'], 'id': 367, 'def': 'a table at which meals are served', 'name': 'dining_table'}, {'frequency': 'r', 'synset': 'dinner_jacket.n.01', 'synonyms': ['tux', 'tuxedo'], 'id': 368, 'def': 'semiformal evening dress for men', 'name': 'tux'}, {'frequency': 'f', 'synset': 'dish.n.01', 'synonyms': ['dish'], 'id': 369, 'def': 'a piece of dishware normally used as a container for holding or serving food', 'name': 'dish'}, {'frequency': 'c', 'synset': 'dish.n.05', 'synonyms': ['dish_antenna'], 'id': 370, 'def': 'directional antenna consisting of a parabolic reflector', 'name': 'dish_antenna'}, {'frequency': 'c', 'synset': 'dishrag.n.01', 'synonyms': ['dishrag', 'dishcloth'], 'id': 371, 'def': 'a cloth for washing dishes or cleaning in general', 'name': 'dishrag'}, {'frequency': 'f', 'synset': 'dishtowel.n.01', 'synonyms': ['dishtowel', 'tea_towel'], 'id': 372, 'def': 'a towel for drying dishes', 'name': 'dishtowel'}, {'frequency': 'f', 'synset': 'dishwasher.n.01', 'synonyms': ['dishwasher', 'dishwashing_machine'], 'id': 373, 'def': 'a machine for washing dishes', 'name': 'dishwasher'}, {'frequency': 'r', 'synset': 'dishwasher_detergent.n.01', 'synonyms': ['dishwasher_detergent', 'dishwashing_detergent', 'dishwashing_liquid', 'dishsoap'], 'id': 374, 'def': 'dishsoap or dish detergent designed for use in dishwashers', 'name': 'dishwasher_detergent'}, {'frequency': 'f', 'synset': 'dispenser.n.01', 'synonyms': ['dispenser'], 'id': 375, 'def': 'a container so designed that the contents can be used in prescribed amounts', 'name': 'dispenser'}, {'frequency': 'r', 'synset': 'diving_board.n.01', 'synonyms': ['diving_board'], 'id': 376, 'def': 'a springboard from which swimmers can dive', 'name': 'diving_board'}, {'frequency': 'f', 'synset': 'dixie_cup.n.01', 'synonyms': ['Dixie_cup', 'paper_cup'], 'id': 377, 'def': 'a disposable cup made of paper; for holding drinks', 'name': 'Dixie_cup'}, {'frequency': 'f', 'synset': 'dog.n.01', 'synonyms': ['dog'], 'id': 378, 'def': 'a common domesticated dog', 'name': 'dog'}, {'frequency': 'f', 'synset': 'dog_collar.n.01', 'synonyms': ['dog_collar'], 'id': 379, 'def': 'a collar for a dog', 'name': 'dog_collar'}, {'frequency': 'f', 'synset': 'doll.n.01', 'synonyms': ['doll'], 'id': 380, 'def': 'a toy replica of a HUMAN (NOT AN ANIMAL)', 'name': 'doll'}, {'frequency': 'r', 'synset': 'dollar.n.02', 'synonyms': ['dollar', 'dollar_bill', 'one_dollar_bill'], 'id': 381, 'def': 'a piece of paper money worth one dollar', 'name': 'dollar'}, {'frequency': 'r', 'synset': 'dollhouse.n.01', 'synonyms': ['dollhouse', "doll's_house"], 'id': 382, 'def': "a house so small that it is likened to a child's plaything", 'name': 'dollhouse'}, {'frequency': 'c', 'synset': 'dolphin.n.02', 'synonyms': ['dolphin'], 'id': 383, 'def': 'any of various small toothed whales with a beaklike snout; larger than porpoises', 'name': 'dolphin'}, {'frequency': 'c', 'synset': 'domestic_ass.n.01', 'synonyms': ['domestic_ass', 'donkey'], 'id': 384, 'def': 'domestic beast of burden descended from the African wild ass; patient but stubborn', 'name': 'domestic_ass'}, {'frequency': 'f', 'synset': 'doorknob.n.01', 'synonyms': ['doorknob', 'doorhandle'], 'id': 385, 'def': "a knob used to open a door (often called `doorhandle' in Great Britain)", 'name': 'doorknob'}, {'frequency': 'c', 'synset': 'doormat.n.02', 'synonyms': ['doormat', 'welcome_mat'], 'id': 386, 'def': 'a mat placed outside an exterior door for wiping the shoes before entering', 'name': 'doormat'}, {'frequency': 'f', 'synset': 'doughnut.n.02', 'synonyms': ['doughnut', 'donut'], 'id': 387, 'def': 'a small ring-shaped friedcake', 'name': 'doughnut'}, {'frequency': 'r', 'synset': 'dove.n.01', 'synonyms': ['dove'], 'id': 388, 'def': 'any of numerous small pigeons', 'name': 'dove'}, {'frequency': 'r', 'synset': 'dragonfly.n.01', 'synonyms': ['dragonfly'], 'id': 389, 'def': 'slender-bodied non-stinging insect having iridescent wings that are outspread at rest', 'name': 'dragonfly'}, {'frequency': 'f', 'synset': 'drawer.n.01', 'synonyms': ['drawer'], 'id': 390, 'def': 'a boxlike container in a piece of furniture; made so as to slide in and out', 'name': 'drawer'}, {'frequency': 'c', 'synset': 'drawers.n.01', 'synonyms': ['underdrawers', 'boxers', 'boxershorts'], 'id': 391, 'def': 'underpants worn by men', 'name': 'underdrawers'}, {'frequency': 'f', 'synset': 'dress.n.01', 'synonyms': ['dress', 'frock'], 'id': 392, 'def': 'a one-piece garment for a woman; has skirt and bodice', 'name': 'dress'}, {'frequency': 'c', 'synset': 'dress_hat.n.01', 'synonyms': ['dress_hat', 'high_hat', 'opera_hat', 'silk_hat', 'top_hat'], 'id': 393, 'def': "a man's hat with a tall crown; usually covered with silk or with beaver fur", 'name': 'dress_hat'}, {'frequency': 'f', 'synset': 'dress_suit.n.01', 'synonyms': ['dress_suit'], 'id': 394, 'def': 'formalwear consisting of full evening dress for men', 'name': 'dress_suit'}, {'frequency': 'f', 'synset': 'dresser.n.05', 'synonyms': ['dresser'], 'id': 395, 'def': 'a cabinet with shelves', 'name': 'dresser'}, {'frequency': 'c', 'synset': 'drill.n.01', 'synonyms': ['drill'], 'id': 396, 'def': 'a tool with a sharp rotating point for making holes in hard materials', 'name': 'drill'}, {'frequency': 'r', 'synset': 'drone.n.04', 'synonyms': ['drone'], 'id': 397, 'def': 'an aircraft without a pilot that is operated by remote control', 'name': 'drone'}, {'frequency': 'r', 'synset': 'dropper.n.01', 'synonyms': ['dropper', 'eye_dropper'], 'id': 398, 'def': 'pipet consisting of a small tube with a vacuum bulb at one end for drawing liquid in and releasing it a drop at a time', 'name': 'dropper'}, {'frequency': 'c', 'synset': 'drum.n.01', 'synonyms': ['drum_(musical_instrument)'], 'id': 399, 'def': 'a musical percussion instrument; usually consists of a hollow cylinder with a membrane stretched across each end', 'name': 'drum_(musical_instrument)'}, {'frequency': 'r', 'synset': 'drumstick.n.02', 'synonyms': ['drumstick'], 'id': 400, 'def': 'a stick used for playing a drum', 'name': 'drumstick'}, {'frequency': 'f', 'synset': 'duck.n.01', 'synonyms': ['duck'], 'id': 401, 'def': 'small web-footed broad-billed swimming bird', 'name': 'duck'}, {'frequency': 'c', 'synset': 'duckling.n.02', 'synonyms': ['duckling'], 'id': 402, 'def': 'young duck', 'name': 'duckling'}, {'frequency': 'c', 'synset': 'duct_tape.n.01', 'synonyms': ['duct_tape'], 'id': 403, 'def': 'a wide silvery adhesive tape', 'name': 'duct_tape'}, {'frequency': 'f', 'synset': 'duffel_bag.n.01', 'synonyms': ['duffel_bag', 'duffle_bag', 'duffel', 'duffle'], 'id': 404, 'def': 'a large cylindrical bag of heavy cloth (does not include suitcases)', 'name': 'duffel_bag'}, {'frequency': 'r', 'synset': 'dumbbell.n.01', 'synonyms': ['dumbbell'], 'id': 405, 'def': 'an exercising weight with two ball-like ends connected by a short handle', 'name': 'dumbbell'}, {'frequency': 'c', 'synset': 'dumpster.n.01', 'synonyms': ['dumpster'], 'id': 406, 'def': 'a container designed to receive and transport and dump waste', 'name': 'dumpster'}, {'frequency': 'r', 'synset': 'dustpan.n.02', 'synonyms': ['dustpan'], 'id': 407, 'def': 'a short-handled receptacle into which dust can be swept', 'name': 'dustpan'}, {'frequency': 'c', 'synset': 'eagle.n.01', 'synonyms': ['eagle'], 'id': 408, 'def': 'large birds of prey noted for their broad wings and strong soaring flight', 'name': 'eagle'}, {'frequency': 'f', 'synset': 'earphone.n.01', 'synonyms': ['earphone', 'earpiece', 'headphone'], 'id': 409, 'def': 'device for listening to audio that is held over or inserted into the ear', 'name': 'earphone'}, {'frequency': 'r', 'synset': 'earplug.n.01', 'synonyms': ['earplug'], 'id': 410, 'def': 'a soft plug that is inserted into the ear canal to block sound', 'name': 'earplug'}, {'frequency': 'f', 'synset': 'earring.n.01', 'synonyms': ['earring'], 'id': 411, 'def': 'jewelry to ornament the ear', 'name': 'earring'}, {'frequency': 'c', 'synset': 'easel.n.01', 'synonyms': ['easel'], 'id': 412, 'def': "an upright tripod for displaying something (usually an artist's canvas)", 'name': 'easel'}, {'frequency': 'r', 'synset': 'eclair.n.01', 'synonyms': ['eclair'], 'id': 413, 'def': 'oblong cream puff', 'name': 'eclair'}, {'frequency': 'r', 'synset': 'eel.n.01', 'synonyms': ['eel'], 'id': 414, 'def': 'an elongate fish with fatty flesh', 'name': 'eel'}, {'frequency': 'f', 'synset': 'egg.n.02', 'synonyms': ['egg', 'eggs'], 'id': 415, 'def': 'oval reproductive body of a fowl (especially a hen) used as food', 'name': 'egg'}, {'frequency': 'r', 'synset': 'egg_roll.n.01', 'synonyms': ['egg_roll', 'spring_roll'], 'id': 416, 'def': 'minced vegetables and meat wrapped in a pancake and fried', 'name': 'egg_roll'}, {'frequency': 'c', 'synset': 'egg_yolk.n.01', 'synonyms': ['egg_yolk', 'yolk_(egg)'], 'id': 417, 'def': 'the yellow spherical part of an egg', 'name': 'egg_yolk'}, {'frequency': 'c', 'synset': 'eggbeater.n.02', 'synonyms': ['eggbeater', 'eggwhisk'], 'id': 418, 'def': 'a mixer for beating eggs or whipping cream', 'name': 'eggbeater'}, {'frequency': 'c', 'synset': 'eggplant.n.01', 'synonyms': ['eggplant', 'aubergine'], 'id': 419, 'def': 'egg-shaped vegetable having a shiny skin typically dark purple', 'name': 'eggplant'}, {'frequency': 'r', 'synset': 'electric_chair.n.01', 'synonyms': ['electric_chair'], 'id': 420, 'def': 'a chair-shaped instrument of execution by electrocution', 'name': 'electric_chair'}, {'frequency': 'f', 'synset': 'electric_refrigerator.n.01', 'synonyms': ['refrigerator'], 'id': 421, 'def': 'a refrigerator in which the coolant is pumped around by an electric motor', 'name': 'refrigerator'}, {'frequency': 'f', 'synset': 'elephant.n.01', 'synonyms': ['elephant'], 'id': 422, 'def': 'a common elephant', 'name': 'elephant'}, {'frequency': 'c', 'synset': 'elk.n.01', 'synonyms': ['elk', 'moose'], 'id': 423, 'def': 'large northern deer with enormous flattened antlers in the male', 'name': 'elk'}, {'frequency': 'c', 'synset': 'envelope.n.01', 'synonyms': ['envelope'], 'id': 424, 'def': 'a flat (usually rectangular) container for a letter, thin package, etc.', 'name': 'envelope'}, {'frequency': 'c', 'synset': 'eraser.n.01', 'synonyms': ['eraser'], 'id': 425, 'def': 'an implement used to erase something', 'name': 'eraser'}, {'frequency': 'r', 'synset': 'escargot.n.01', 'synonyms': ['escargot'], 'id': 426, 'def': 'edible snail usually served in the shell with a sauce of melted butter and garlic', 'name': 'escargot'}, {'frequency': 'r', 'synset': 'eyepatch.n.01', 'synonyms': ['eyepatch'], 'id': 427, 'def': 'a protective cloth covering for an injured eye', 'name': 'eyepatch'}, {'frequency': 'r', 'synset': 'falcon.n.01', 'synonyms': ['falcon'], 'id': 428, 'def': 'birds of prey having long pointed powerful wings adapted for swift flight', 'name': 'falcon'}, {'frequency': 'f', 'synset': 'fan.n.01', 'synonyms': ['fan'], 'id': 429, 'def': 'a device for creating a current of air by movement of a surface or surfaces', 'name': 'fan'}, {'frequency': 'f', 'synset': 'faucet.n.01', 'synonyms': ['faucet', 'spigot', 'tap'], 'id': 430, 'def': 'a regulator for controlling the flow of a liquid from a reservoir', 'name': 'faucet'}, {'frequency': 'r', 'synset': 'fedora.n.01', 'synonyms': ['fedora'], 'id': 431, 'def': 'a hat made of felt with a creased crown', 'name': 'fedora'}, {'frequency': 'r', 'synset': 'ferret.n.02', 'synonyms': ['ferret'], 'id': 432, 'def': 'domesticated albino variety of the European polecat bred for hunting rats and rabbits', 'name': 'ferret'}, {'frequency': 'c', 'synset': 'ferris_wheel.n.01', 'synonyms': ['Ferris_wheel'], 'id': 433, 'def': 'a large wheel with suspended seats that remain upright as the wheel rotates', 'name': 'Ferris_wheel'}, {'frequency': 'c', 'synset': 'ferry.n.01', 'synonyms': ['ferry', 'ferryboat'], 'id': 434, 'def': 'a boat that transports people or vehicles across a body of water and operates on a regular schedule', 'name': 'ferry'}, {'frequency': 'r', 'synset': 'fig.n.04', 'synonyms': ['fig_(fruit)'], 'id': 435, 'def': 'fleshy sweet pear-shaped yellowish or purple fruit eaten fresh or preserved or dried', 'name': 'fig_(fruit)'}, {'frequency': 'c', 'synset': 'fighter.n.02', 'synonyms': ['fighter_jet', 'fighter_aircraft', 'attack_aircraft'], 'id': 436, 'def': 'a high-speed military or naval airplane designed to destroy enemy targets', 'name': 'fighter_jet'}, {'frequency': 'f', 'synset': 'figurine.n.01', 'synonyms': ['figurine'], 'id': 437, 'def': 'a small carved or molded figure', 'name': 'figurine'}, {'frequency': 'c', 'synset': 'file.n.03', 'synonyms': ['file_cabinet', 'filing_cabinet'], 'id': 438, 'def': 'office furniture consisting of a container for keeping papers in order', 'name': 'file_cabinet'}, {'frequency': 'r', 'synset': 'file.n.04', 'synonyms': ['file_(tool)'], 'id': 439, 'def': 'a steel hand tool with small sharp teeth on some or all of its surfaces; used for smoothing wood or metal', 'name': 'file_(tool)'}, {'frequency': 'f', 'synset': 'fire_alarm.n.02', 'synonyms': ['fire_alarm', 'smoke_alarm'], 'id': 440, 'def': 'an alarm that is tripped off by fire or smoke', 'name': 'fire_alarm'}, {'frequency': 'f', 'synset': 'fire_engine.n.01', 'synonyms': ['fire_engine', 'fire_truck'], 'id': 441, 'def': 'large trucks that carry firefighters and equipment to the site of a fire', 'name': 'fire_engine'}, {'frequency': 'f', 'synset': 'fire_extinguisher.n.01', 'synonyms': ['fire_extinguisher', 'extinguisher'], 'id': 442, 'def': 'a manually operated device for extinguishing small fires', 'name': 'fire_extinguisher'}, {'frequency': 'c', 'synset': 'fire_hose.n.01', 'synonyms': ['fire_hose'], 'id': 443, 'def': 'a large hose that carries water from a fire hydrant to the site of the fire', 'name': 'fire_hose'}, {'frequency': 'f', 'synset': 'fireplace.n.01', 'synonyms': ['fireplace'], 'id': 444, 'def': 'an open recess in a wall at the base of a chimney where a fire can be built', 'name': 'fireplace'}, {'frequency': 'f', 'synset': 'fireplug.n.01', 'synonyms': ['fireplug', 'fire_hydrant', 'hydrant'], 'id': 445, 'def': 'an upright hydrant for drawing water to use in fighting a fire', 'name': 'fireplug'}, {'frequency': 'r', 'synset': 'first-aid_kit.n.01', 'synonyms': ['first-aid_kit'], 'id': 446, 'def': 'kit consisting of a set of bandages and medicines for giving first aid', 'name': 'first-aid_kit'}, {'frequency': 'f', 'synset': 'fish.n.01', 'synonyms': ['fish'], 'id': 447, 'def': 'any of various mostly cold-blooded aquatic vertebrates usually having scales and breathing through gills', 'name': 'fish'}, {'frequency': 'c', 'synset': 'fish.n.02', 'synonyms': ['fish_(food)'], 'id': 448, 'def': 'the flesh of fish used as food', 'name': 'fish_(food)'}, {'frequency': 'r', 'synset': 'fishbowl.n.02', 'synonyms': ['fishbowl', 'goldfish_bowl'], 'id': 449, 'def': 'a transparent bowl in which small fish are kept', 'name': 'fishbowl'}, {'frequency': 'c', 'synset': 'fishing_rod.n.01', 'synonyms': ['fishing_rod', 'fishing_pole'], 'id': 450, 'def': 'a rod that is used in fishing to extend the fishing line', 'name': 'fishing_rod'}, {'frequency': 'f', 'synset': 'flag.n.01', 'synonyms': ['flag'], 'id': 451, 'def': 'emblem usually consisting of a rectangular piece of cloth of distinctive design (do not include pole)', 'name': 'flag'}, {'frequency': 'f', 'synset': 'flagpole.n.02', 'synonyms': ['flagpole', 'flagstaff'], 'id': 452, 'def': 'a tall staff or pole on which a flag is raised', 'name': 'flagpole'}, {'frequency': 'c', 'synset': 'flamingo.n.01', 'synonyms': ['flamingo'], 'id': 453, 'def': 'large pink web-footed bird with down-bent bill', 'name': 'flamingo'}, {'frequency': 'c', 'synset': 'flannel.n.01', 'synonyms': ['flannel'], 'id': 454, 'def': 'a soft light woolen fabric; used for clothing', 'name': 'flannel'}, {'frequency': 'c', 'synset': 'flap.n.01', 'synonyms': ['flap'], 'id': 455, 'def': 'any broad thin covering attached at one edge, such as a mud flap next to a wheel or a flap on an airplane wing', 'name': 'flap'}, {'frequency': 'r', 'synset': 'flash.n.10', 'synonyms': ['flash', 'flashbulb'], 'id': 456, 'def': 'a lamp for providing momentary light to take a photograph', 'name': 'flash'}, {'frequency': 'c', 'synset': 'flashlight.n.01', 'synonyms': ['flashlight', 'torch'], 'id': 457, 'def': 'a small portable battery-powered electric lamp', 'name': 'flashlight'}, {'frequency': 'r', 'synset': 'fleece.n.03', 'synonyms': ['fleece'], 'id': 458, 'def': 'a soft bulky fabric with deep pile; used chiefly for clothing', 'name': 'fleece'}, {'frequency': 'f', 'synset': 'flip-flop.n.02', 'synonyms': ['flip-flop_(sandal)'], 'id': 459, 'def': 'a backless sandal held to the foot by a thong between two toes', 'name': 'flip-flop_(sandal)'}, {'frequency': 'c', 'synset': 'flipper.n.01', 'synonyms': ['flipper_(footwear)', 'fin_(footwear)'], 'id': 460, 'def': 'a shoe to aid a person in swimming', 'name': 'flipper_(footwear)'}, {'frequency': 'f', 'synset': 'flower_arrangement.n.01', 'synonyms': ['flower_arrangement', 'floral_arrangement'], 'id': 461, 'def': 'a decorative arrangement of flowers', 'name': 'flower_arrangement'}, {'frequency': 'c', 'synset': 'flute.n.02', 'synonyms': ['flute_glass', 'champagne_flute'], 'id': 462, 'def': 'a tall narrow wineglass', 'name': 'flute_glass'}, {'frequency': 'c', 'synset': 'foal.n.01', 'synonyms': ['foal'], 'id': 463, 'def': 'a young horse', 'name': 'foal'}, {'frequency': 'c', 'synset': 'folding_chair.n.01', 'synonyms': ['folding_chair'], 'id': 464, 'def': 'a chair that can be folded flat for storage', 'name': 'folding_chair'}, {'frequency': 'c', 'synset': 'food_processor.n.01', 'synonyms': ['food_processor'], 'id': 465, 'def': 'a kitchen appliance for shredding, blending, chopping, or slicing food', 'name': 'food_processor'}, {'frequency': 'c', 'synset': 'football.n.02', 'synonyms': ['football_(American)'], 'id': 466, 'def': 'the inflated oblong ball used in playing American football', 'name': 'football_(American)'}, {'frequency': 'r', 'synset': 'football_helmet.n.01', 'synonyms': ['football_helmet'], 'id': 467, 'def': 'a padded helmet with a face mask to protect the head of football players', 'name': 'football_helmet'}, {'frequency': 'c', 'synset': 'footstool.n.01', 'synonyms': ['footstool', 'footrest'], 'id': 468, 'def': 'a low seat or a stool to rest the feet of a seated person', 'name': 'footstool'}, {'frequency': 'f', 'synset': 'fork.n.01', 'synonyms': ['fork'], 'id': 469, 'def': 'cutlery used for serving and eating food', 'name': 'fork'}, {'frequency': 'c', 'synset': 'forklift.n.01', 'synonyms': ['forklift'], 'id': 470, 'def': 'an industrial vehicle with a power operated fork in front that can be inserted under loads to lift and move them', 'name': 'forklift'}, {'frequency': 'c', 'synset': 'freight_car.n.01', 'synonyms': ['freight_car'], 'id': 471, 'def': 'a railway car that carries freight', 'name': 'freight_car'}, {'frequency': 'c', 'synset': 'french_toast.n.01', 'synonyms': ['French_toast'], 'id': 472, 'def': 'bread slice dipped in egg and milk and fried', 'name': 'French_toast'}, {'frequency': 'c', 'synset': 'freshener.n.01', 'synonyms': ['freshener', 'air_freshener'], 'id': 473, 'def': 'anything that freshens air by removing or covering odor', 'name': 'freshener'}, {'frequency': 'f', 'synset': 'frisbee.n.01', 'synonyms': ['frisbee'], 'id': 474, 'def': 'a light, plastic disk propelled with a flip of the wrist for recreation or competition', 'name': 'frisbee'}, {'frequency': 'c', 'synset': 'frog.n.01', 'synonyms': ['frog', 'toad', 'toad_frog'], 'id': 475, 'def': 'a tailless stout-bodied amphibians with long hind limbs for leaping', 'name': 'frog'}, {'frequency': 'c', 'synset': 'fruit_juice.n.01', 'synonyms': ['fruit_juice'], 'id': 476, 'def': 'drink produced by squeezing or crushing fruit', 'name': 'fruit_juice'}, {'frequency': 'f', 'synset': 'frying_pan.n.01', 'synonyms': ['frying_pan', 'frypan', 'skillet'], 'id': 477, 'def': 'a pan used for frying foods', 'name': 'frying_pan'}, {'frequency': 'r', 'synset': 'fudge.n.01', 'synonyms': ['fudge'], 'id': 478, 'def': 'soft creamy candy', 'name': 'fudge'}, {'frequency': 'r', 'synset': 'funnel.n.02', 'synonyms': ['funnel'], 'id': 479, 'def': 'a cone-shaped utensil used to channel a substance into a container with a small mouth', 'name': 'funnel'}, {'frequency': 'r', 'synset': 'futon.n.01', 'synonyms': ['futon'], 'id': 480, 'def': 'a pad that is used for sleeping on the floor or on a raised frame', 'name': 'futon'}, {'frequency': 'r', 'synset': 'gag.n.02', 'synonyms': ['gag', 'muzzle'], 'id': 481, 'def': "restraint put into a person's mouth to prevent speaking or shouting", 'name': 'gag'}, {'frequency': 'r', 'synset': 'garbage.n.03', 'synonyms': ['garbage'], 'id': 482, 'def': 'a receptacle where waste can be discarded', 'name': 'garbage'}, {'frequency': 'c', 'synset': 'garbage_truck.n.01', 'synonyms': ['garbage_truck'], 'id': 483, 'def': 'a truck for collecting domestic refuse', 'name': 'garbage_truck'}, {'frequency': 'c', 'synset': 'garden_hose.n.01', 'synonyms': ['garden_hose'], 'id': 484, 'def': 'a hose used for watering a lawn or garden', 'name': 'garden_hose'}, {'frequency': 'c', 'synset': 'gargle.n.01', 'synonyms': ['gargle', 'mouthwash'], 'id': 485, 'def': 'a medicated solution used for gargling and rinsing the mouth', 'name': 'gargle'}, {'frequency': 'r', 'synset': 'gargoyle.n.02', 'synonyms': ['gargoyle'], 'id': 486, 'def': 'an ornament consisting of a grotesquely carved figure of a person or animal', 'name': 'gargoyle'}, {'frequency': 'c', 'synset': 'garlic.n.02', 'synonyms': ['garlic', 'ail'], 'id': 487, 'def': 'aromatic bulb used as seasoning', 'name': 'garlic'}, {'frequency': 'r', 'synset': 'gasmask.n.01', 'synonyms': ['gasmask', 'respirator', 'gas_helmet'], 'id': 488, 'def': 'a protective face mask with a filter', 'name': 'gasmask'}, {'frequency': 'c', 'synset': 'gazelle.n.01', 'synonyms': ['gazelle'], 'id': 489, 'def': 'small swift graceful antelope of Africa and Asia having lustrous eyes', 'name': 'gazelle'}, {'frequency': 'c', 'synset': 'gelatin.n.02', 'synonyms': ['gelatin', 'jelly'], 'id': 490, 'def': 'an edible jelly made with gelatin and used as a dessert or salad base or a coating for foods', 'name': 'gelatin'}, {'frequency': 'r', 'synset': 'gem.n.02', 'synonyms': ['gemstone'], 'id': 491, 'def': 'a crystalline rock that can be cut and polished for jewelry', 'name': 'gemstone'}, {'frequency': 'r', 'synset': 'generator.n.02', 'synonyms': ['generator'], 'id': 492, 'def': 'engine that converts mechanical energy into electrical energy by electromagnetic induction', 'name': 'generator'}, {'frequency': 'c', 'synset': 'giant_panda.n.01', 'synonyms': ['giant_panda', 'panda', 'panda_bear'], 'id': 493, 'def': 'large black-and-white herbivorous mammal of bamboo forests of China and Tibet', 'name': 'giant_panda'}, {'frequency': 'c', 'synset': 'gift_wrap.n.01', 'synonyms': ['gift_wrap'], 'id': 494, 'def': 'attractive wrapping paper suitable for wrapping gifts', 'name': 'gift_wrap'}, {'frequency': 'c', 'synset': 'ginger.n.03', 'synonyms': ['ginger', 'gingerroot'], 'id': 495, 'def': 'the root of the common ginger plant; used fresh as a seasoning', 'name': 'ginger'}, {'frequency': 'f', 'synset': 'giraffe.n.01', 'synonyms': ['giraffe'], 'id': 496, 'def': 'tall animal having a spotted coat and small horns and very long neck and legs', 'name': 'giraffe'}, {'frequency': 'c', 'synset': 'girdle.n.02', 'synonyms': ['cincture', 'sash', 'waistband', 'waistcloth'], 'id': 497, 'def': 'a band of material around the waist that strengthens a skirt or trousers', 'name': 'cincture'}, {'frequency': 'f', 'synset': 'glass.n.02', 'synonyms': ['glass_(drink_container)', 'drinking_glass'], 'id': 498, 'def': 'a container for holding liquids while drinking', 'name': 'glass_(drink_container)'}, {'frequency': 'c', 'synset': 'globe.n.03', 'synonyms': ['globe'], 'id': 499, 'def': 'a sphere on which a map (especially of the earth) is represented', 'name': 'globe'}, {'frequency': 'f', 'synset': 'glove.n.02', 'synonyms': ['glove'], 'id': 500, 'def': 'handwear covering the hand', 'name': 'glove'}, {'frequency': 'c', 'synset': 'goat.n.01', 'synonyms': ['goat'], 'id': 501, 'def': 'a common goat', 'name': 'goat'}, {'frequency': 'f', 'synset': 'goggles.n.01', 'synonyms': ['goggles'], 'id': 502, 'def': 'tight-fitting spectacles worn to protect the eyes', 'name': 'goggles'}, {'frequency': 'r', 'synset': 'goldfish.n.01', 'synonyms': ['goldfish'], 'id': 503, 'def': 'small golden or orange-red freshwater fishes used as pond or aquarium pets', 'name': 'goldfish'}, {'frequency': 'c', 'synset': 'golf_club.n.02', 'synonyms': ['golf_club', 'golf-club'], 'id': 504, 'def': 'golf equipment used by a golfer to hit a golf ball', 'name': 'golf_club'}, {'frequency': 'c', 'synset': 'golfcart.n.01', 'synonyms': ['golfcart'], 'id': 505, 'def': 'a small motor vehicle in which golfers can ride between shots', 'name': 'golfcart'}, {'frequency': 'r', 'synset': 'gondola.n.02', 'synonyms': ['gondola_(boat)'], 'id': 506, 'def': 'long narrow flat-bottomed boat propelled by sculling; traditionally used on canals of Venice', 'name': 'gondola_(boat)'}, {'frequency': 'c', 'synset': 'goose.n.01', 'synonyms': ['goose'], 'id': 507, 'def': 'loud, web-footed long-necked aquatic birds usually larger than ducks', 'name': 'goose'}, {'frequency': 'r', 'synset': 'gorilla.n.01', 'synonyms': ['gorilla'], 'id': 508, 'def': 'largest ape', 'name': 'gorilla'}, {'frequency': 'r', 'synset': 'gourd.n.02', 'synonyms': ['gourd'], 'id': 509, 'def': 'any of numerous inedible fruits with hard rinds', 'name': 'gourd'}, {'frequency': 'f', 'synset': 'grape.n.01', 'synonyms': ['grape'], 'id': 510, 'def': 'any of various juicy fruit with green or purple skins; grow in clusters', 'name': 'grape'}, {'frequency': 'c', 'synset': 'grater.n.01', 'synonyms': ['grater'], 'id': 511, 'def': 'utensil with sharp perforations for shredding foods (as vegetables or cheese)', 'name': 'grater'}, {'frequency': 'c', 'synset': 'gravestone.n.01', 'synonyms': ['gravestone', 'headstone', 'tombstone'], 'id': 512, 'def': 'a stone that is used to mark a grave', 'name': 'gravestone'}, {'frequency': 'r', 'synset': 'gravy_boat.n.01', 'synonyms': ['gravy_boat', 'gravy_holder'], 'id': 513, 'def': 'a dish (often boat-shaped) for serving gravy or sauce', 'name': 'gravy_boat'}, {'frequency': 'f', 'synset': 'green_bean.n.02', 'synonyms': ['green_bean'], 'id': 514, 'def': 'a common bean plant cultivated for its slender green edible pods', 'name': 'green_bean'}, {'frequency': 'f', 'synset': 'green_onion.n.01', 'synonyms': ['green_onion', 'spring_onion', 'scallion'], 'id': 515, 'def': 'a young onion before the bulb has enlarged', 'name': 'green_onion'}, {'frequency': 'r', 'synset': 'griddle.n.01', 'synonyms': ['griddle'], 'id': 516, 'def': 'cooking utensil consisting of a flat heated surface on which food is cooked', 'name': 'griddle'}, {'frequency': 'f', 'synset': 'grill.n.02', 'synonyms': ['grill', 'grille', 'grillwork', 'radiator_grille'], 'id': 517, 'def': 'a framework of metal bars used as a partition or a grate', 'name': 'grill'}, {'frequency': 'r', 'synset': 'grits.n.01', 'synonyms': ['grits', 'hominy_grits'], 'id': 518, 'def': 'coarsely ground corn boiled as a breakfast dish', 'name': 'grits'}, {'frequency': 'c', 'synset': 'grizzly.n.01', 'synonyms': ['grizzly', 'grizzly_bear'], 'id': 519, 'def': 'powerful brownish-yellow bear of the uplands of western North America', 'name': 'grizzly'}, {'frequency': 'c', 'synset': 'grocery_bag.n.01', 'synonyms': ['grocery_bag'], 'id': 520, 'def': "a sack for holding customer's groceries", 'name': 'grocery_bag'}, {'frequency': 'f', 'synset': 'guitar.n.01', 'synonyms': ['guitar'], 'id': 521, 'def': 'a stringed instrument usually having six strings; played by strumming or plucking', 'name': 'guitar'}, {'frequency': 'c', 'synset': 'gull.n.02', 'synonyms': ['gull', 'seagull'], 'id': 522, 'def': 'mostly white aquatic bird having long pointed wings and short legs', 'name': 'gull'}, {'frequency': 'c', 'synset': 'gun.n.01', 'synonyms': ['gun'], 'id': 523, 'def': 'a weapon that discharges a bullet at high velocity from a metal tube', 'name': 'gun'}, {'frequency': 'f', 'synset': 'hairbrush.n.01', 'synonyms': ['hairbrush'], 'id': 524, 'def': "a brush used to groom a person's hair", 'name': 'hairbrush'}, {'frequency': 'c', 'synset': 'hairnet.n.01', 'synonyms': ['hairnet'], 'id': 525, 'def': 'a small net that someone wears over their hair to keep it in place', 'name': 'hairnet'}, {'frequency': 'c', 'synset': 'hairpin.n.01', 'synonyms': ['hairpin'], 'id': 526, 'def': "a double pronged pin used to hold women's hair in place", 'name': 'hairpin'}, {'frequency': 'r', 'synset': 'halter.n.03', 'synonyms': ['halter_top'], 'id': 527, 'def': "a woman's top that fastens behind the back and neck leaving the back and arms uncovered", 'name': 'halter_top'}, {'frequency': 'f', 'synset': 'ham.n.01', 'synonyms': ['ham', 'jambon', 'gammon'], 'id': 528, 'def': 'meat cut from the thigh of a hog (usually smoked)', 'name': 'ham'}, {'frequency': 'c', 'synset': 'hamburger.n.01', 'synonyms': ['hamburger', 'beefburger', 'burger'], 'id': 529, 'def': 'a sandwich consisting of a patty of minced beef served on a bun', 'name': 'hamburger'}, {'frequency': 'c', 'synset': 'hammer.n.02', 'synonyms': ['hammer'], 'id': 530, 'def': 'a hand tool with a heavy head and a handle; used to deliver an impulsive force by striking', 'name': 'hammer'}, {'frequency': 'c', 'synset': 'hammock.n.02', 'synonyms': ['hammock'], 'id': 531, 'def': 'a hanging bed of canvas or rope netting (usually suspended between two trees)', 'name': 'hammock'}, {'frequency': 'r', 'synset': 'hamper.n.02', 'synonyms': ['hamper'], 'id': 532, 'def': 'a basket usually with a cover', 'name': 'hamper'}, {'frequency': 'c', 'synset': 'hamster.n.01', 'synonyms': ['hamster'], 'id': 533, 'def': 'short-tailed burrowing rodent with large cheek pouches', 'name': 'hamster'}, {'frequency': 'f', 'synset': 'hand_blower.n.01', 'synonyms': ['hair_dryer'], 'id': 534, 'def': 'a hand-held electric blower that can blow warm air onto the hair', 'name': 'hair_dryer'}, {'frequency': 'r', 'synset': 'hand_glass.n.01', 'synonyms': ['hand_glass', 'hand_mirror'], 'id': 535, 'def': 'a mirror intended to be held in the hand', 'name': 'hand_glass'}, {'frequency': 'f', 'synset': 'hand_towel.n.01', 'synonyms': ['hand_towel', 'face_towel'], 'id': 536, 'def': 'a small towel used to dry the hands or face', 'name': 'hand_towel'}, {'frequency': 'c', 'synset': 'handcart.n.01', 'synonyms': ['handcart', 'pushcart', 'hand_truck'], 'id': 537, 'def': 'wheeled vehicle that can be pushed by a person', 'name': 'handcart'}, {'frequency': 'r', 'synset': 'handcuff.n.01', 'synonyms': ['handcuff'], 'id': 538, 'def': 'shackle that consists of a metal loop that can be locked around the wrist', 'name': 'handcuff'}, {'frequency': 'c', 'synset': 'handkerchief.n.01', 'synonyms': ['handkerchief'], 'id': 539, 'def': 'a square piece of cloth used for wiping the eyes or nose or as a costume accessory', 'name': 'handkerchief'}, {'frequency': 'f', 'synset': 'handle.n.01', 'synonyms': ['handle', 'grip', 'handgrip'], 'id': 540, 'def': 'the appendage to an object that is designed to be held in order to use or move it', 'name': 'handle'}, {'frequency': 'r', 'synset': 'handsaw.n.01', 'synonyms': ['handsaw', "carpenter's_saw"], 'id': 541, 'def': 'a saw used with one hand for cutting wood', 'name': 'handsaw'}, {'frequency': 'r', 'synset': 'hardback.n.01', 'synonyms': ['hardback_book', 'hardcover_book'], 'id': 542, 'def': 'a book with cardboard or cloth or leather covers', 'name': 'hardback_book'}, {'frequency': 'r', 'synset': 'harmonium.n.01', 'synonyms': ['harmonium', 'organ_(musical_instrument)', 'reed_organ_(musical_instrument)'], 'id': 543, 'def': 'a free-reed instrument in which air is forced through the reeds by bellows', 'name': 'harmonium'}, {'frequency': 'f', 'synset': 'hat.n.01', 'synonyms': ['hat'], 'id': 544, 'def': 'headwear that protects the head from bad weather, sun, or worn for fashion', 'name': 'hat'}, {'frequency': 'r', 'synset': 'hatbox.n.01', 'synonyms': ['hatbox'], 'id': 545, 'def': 'a round piece of luggage for carrying hats', 'name': 'hatbox'}, {'frequency': 'c', 'synset': 'head_covering.n.01', 'synonyms': ['veil'], 'id': 546, 'def': 'a garment that covers the head OR face', 'name': 'veil'}, {'frequency': 'f', 'synset': 'headband.n.01', 'synonyms': ['headband'], 'id': 547, 'def': 'a band worn around or over the head', 'name': 'headband'}, {'frequency': 'f', 'synset': 'headboard.n.01', 'synonyms': ['headboard'], 'id': 548, 'def': 'a vertical board or panel forming the head of a bedstead', 'name': 'headboard'}, {'frequency': 'f', 'synset': 'headlight.n.01', 'synonyms': ['headlight', 'headlamp'], 'id': 549, 'def': 'a powerful light with reflector; attached to the front of an automobile or locomotive', 'name': 'headlight'}, {'frequency': 'c', 'synset': 'headscarf.n.01', 'synonyms': ['headscarf'], 'id': 550, 'def': 'a kerchief worn over the head and tied under the chin', 'name': 'headscarf'}, {'frequency': 'r', 'synset': 'headset.n.01', 'synonyms': ['headset'], 'id': 551, 'def': 'receiver consisting of a pair of headphones', 'name': 'headset'}, {'frequency': 'c', 'synset': 'headstall.n.01', 'synonyms': ['headstall_(for_horses)', 'headpiece_(for_horses)'], 'id': 552, 'def': "the band that is the part of a bridle that fits around a horse's head", 'name': 'headstall_(for_horses)'}, {'frequency': 'c', 'synset': 'heart.n.02', 'synonyms': ['heart'], 'id': 553, 'def': 'a muscular organ; its contractions move the blood through the body', 'name': 'heart'}, {'frequency': 'c', 'synset': 'heater.n.01', 'synonyms': ['heater', 'warmer'], 'id': 554, 'def': 'device that heats water or supplies warmth to a room', 'name': 'heater'}, {'frequency': 'c', 'synset': 'helicopter.n.01', 'synonyms': ['helicopter'], 'id': 555, 'def': 'an aircraft without wings that obtains its lift from the rotation of overhead blades', 'name': 'helicopter'}, {'frequency': 'f', 'synset': 'helmet.n.02', 'synonyms': ['helmet'], 'id': 556, 'def': 'a protective headgear made of hard material to resist blows', 'name': 'helmet'}, {'frequency': 'r', 'synset': 'heron.n.02', 'synonyms': ['heron'], 'id': 557, 'def': 'grey or white wading bird with long neck and long legs and (usually) long bill', 'name': 'heron'}, {'frequency': 'c', 'synset': 'highchair.n.01', 'synonyms': ['highchair', 'feeding_chair'], 'id': 558, 'def': 'a chair for feeding a very young child', 'name': 'highchair'}, {'frequency': 'f', 'synset': 'hinge.n.01', 'synonyms': ['hinge'], 'id': 559, 'def': 'a joint that holds two parts together so that one can swing relative to the other', 'name': 'hinge'}, {'frequency': 'r', 'synset': 'hippopotamus.n.01', 'synonyms': ['hippopotamus'], 'id': 560, 'def': 'massive thick-skinned animal living in or around rivers of tropical Africa', 'name': 'hippopotamus'}, {'frequency': 'r', 'synset': 'hockey_stick.n.01', 'synonyms': ['hockey_stick'], 'id': 561, 'def': 'sports implement consisting of a stick used by hockey players to move the puck', 'name': 'hockey_stick'}, {'frequency': 'c', 'synset': 'hog.n.03', 'synonyms': ['hog', 'pig'], 'id': 562, 'def': 'domestic swine', 'name': 'hog'}, {'frequency': 'f', 'synset': 'home_plate.n.01', 'synonyms': ['home_plate_(baseball)', 'home_base_(baseball)'], 'id': 563, 'def': '(baseball) a rubber slab where the batter stands; it must be touched by a base runner in order to score', 'name': 'home_plate_(baseball)'}, {'frequency': 'c', 'synset': 'honey.n.01', 'synonyms': ['honey'], 'id': 564, 'def': 'a sweet yellow liquid produced by bees', 'name': 'honey'}, {'frequency': 'f', 'synset': 'hood.n.06', 'synonyms': ['fume_hood', 'exhaust_hood'], 'id': 565, 'def': 'metal covering leading to a vent that exhausts smoke or fumes', 'name': 'fume_hood'}, {'frequency': 'f', 'synset': 'hook.n.05', 'synonyms': ['hook'], 'id': 566, 'def': 'a curved or bent implement for suspending or pulling something', 'name': 'hook'}, {'frequency': 'r', 'synset': 'hookah.n.01', 'synonyms': ['hookah', 'narghile', 'nargileh', 'sheesha', 'shisha', 'water_pipe'], 'id': 567, 'def': 'a tobacco pipe with a long flexible tube connected to a container where the smoke is cooled by passing through water', 'name': 'hookah'}, {'frequency': 'r', 'synset': 'hornet.n.01', 'synonyms': ['hornet'], 'id': 568, 'def': 'large stinging wasp', 'name': 'hornet'}, {'frequency': 'f', 'synset': 'horse.n.01', 'synonyms': ['horse'], 'id': 569, 'def': 'a common horse', 'name': 'horse'}, {'frequency': 'f', 'synset': 'hose.n.03', 'synonyms': ['hose', 'hosepipe'], 'id': 570, 'def': 'a flexible pipe for conveying a liquid or gas', 'name': 'hose'}, {'frequency': 'r', 'synset': 'hot-air_balloon.n.01', 'synonyms': ['hot-air_balloon'], 'id': 571, 'def': 'balloon for travel through the air in a basket suspended below a large bag of heated air', 'name': 'hot-air_balloon'}, {'frequency': 'r', 'synset': 'hot_plate.n.01', 'synonyms': ['hotplate'], 'id': 572, 'def': 'a portable electric appliance for heating or cooking or keeping food warm', 'name': 'hotplate'}, {'frequency': 'c', 'synset': 'hot_sauce.n.01', 'synonyms': ['hot_sauce'], 'id': 573, 'def': 'a pungent peppery sauce', 'name': 'hot_sauce'}, {'frequency': 'r', 'synset': 'hourglass.n.01', 'synonyms': ['hourglass'], 'id': 574, 'def': 'a sandglass timer that runs for sixty minutes', 'name': 'hourglass'}, {'frequency': 'r', 'synset': 'houseboat.n.01', 'synonyms': ['houseboat'], 'id': 575, 'def': 'a barge that is designed and equipped for use as a dwelling', 'name': 'houseboat'}, {'frequency': 'c', 'synset': 'hummingbird.n.01', 'synonyms': ['hummingbird'], 'id': 576, 'def': 'tiny American bird having brilliant iridescent plumage and long slender bills', 'name': 'hummingbird'}, {'frequency': 'r', 'synset': 'hummus.n.01', 'synonyms': ['hummus', 'humus', 'hommos', 'hoummos', 'humous'], 'id': 577, 'def': 'a thick spread made from mashed chickpeas', 'name': 'hummus'}, {'frequency': 'f', 'synset': 'ice_bear.n.01', 'synonyms': ['polar_bear'], 'id': 578, 'def': 'white bear of Arctic regions', 'name': 'polar_bear'}, {'frequency': 'c', 'synset': 'ice_cream.n.01', 'synonyms': ['icecream'], 'id': 579, 'def': 'frozen dessert containing cream and sugar and flavoring', 'name': 'icecream'}, {'frequency': 'r', 'synset': 'ice_lolly.n.01', 'synonyms': ['popsicle'], 'id': 580, 'def': 'ice cream or water ice on a small wooden stick', 'name': 'popsicle'}, {'frequency': 'c', 'synset': 'ice_maker.n.01', 'synonyms': ['ice_maker'], 'id': 581, 'def': 'an appliance included in some electric refrigerators for making ice cubes', 'name': 'ice_maker'}, {'frequency': 'r', 'synset': 'ice_pack.n.01', 'synonyms': ['ice_pack', 'ice_bag'], 'id': 582, 'def': 'a waterproof bag filled with ice: applied to the body (especially the head) to cool or reduce swelling', 'name': 'ice_pack'}, {'frequency': 'r', 'synset': 'ice_skate.n.01', 'synonyms': ['ice_skate'], 'id': 583, 'def': 'skate consisting of a boot with a steel blade fitted to the sole', 'name': 'ice_skate'}, {'frequency': 'c', 'synset': 'igniter.n.01', 'synonyms': ['igniter', 'ignitor', 'lighter'], 'id': 584, 'def': 'a substance or device used to start a fire', 'name': 'igniter'}, {'frequency': 'r', 'synset': 'inhaler.n.01', 'synonyms': ['inhaler', 'inhalator'], 'id': 585, 'def': 'a dispenser that produces a chemical vapor to be inhaled through mouth or nose', 'name': 'inhaler'}, {'frequency': 'f', 'synset': 'ipod.n.01', 'synonyms': ['iPod'], 'id': 586, 'def': 'a pocket-sized device used to play music files', 'name': 'iPod'}, {'frequency': 'c', 'synset': 'iron.n.04', 'synonyms': ['iron_(for_clothing)', 'smoothing_iron_(for_clothing)'], 'id': 587, 'def': 'home appliance consisting of a flat metal base that is heated and used to smooth cloth', 'name': 'iron_(for_clothing)'}, {'frequency': 'c', 'synset': 'ironing_board.n.01', 'synonyms': ['ironing_board'], 'id': 588, 'def': 'narrow padded board on collapsible supports; used for ironing clothes', 'name': 'ironing_board'}, {'frequency': 'f', 'synset': 'jacket.n.01', 'synonyms': ['jacket'], 'id': 589, 'def': 'a waist-length coat', 'name': 'jacket'}, {'frequency': 'c', 'synset': 'jam.n.01', 'synonyms': ['jam'], 'id': 590, 'def': 'preserve of crushed fruit', 'name': 'jam'}, {'frequency': 'f', 'synset': 'jar.n.01', 'synonyms': ['jar'], 'id': 591, 'def': 'a vessel (usually cylindrical) with a wide mouth and without handles', 'name': 'jar'}, {'frequency': 'f', 'synset': 'jean.n.01', 'synonyms': ['jean', 'blue_jean', 'denim'], 'id': 592, 'def': '(usually plural) close-fitting trousers of heavy denim for manual work or casual wear', 'name': 'jean'}, {'frequency': 'c', 'synset': 'jeep.n.01', 'synonyms': ['jeep', 'landrover'], 'id': 593, 'def': 'a car suitable for traveling over rough terrain', 'name': 'jeep'}, {'frequency': 'r', 'synset': 'jelly_bean.n.01', 'synonyms': ['jelly_bean', 'jelly_egg'], 'id': 594, 'def': 'sugar-glazed jellied candy', 'name': 'jelly_bean'}, {'frequency': 'f', 'synset': 'jersey.n.03', 'synonyms': ['jersey', 'T-shirt', 'tee_shirt'], 'id': 595, 'def': 'a close-fitting pullover shirt', 'name': 'jersey'}, {'frequency': 'c', 'synset': 'jet.n.01', 'synonyms': ['jet_plane', 'jet-propelled_plane'], 'id': 596, 'def': 'an airplane powered by one or more jet engines', 'name': 'jet_plane'}, {'frequency': 'r', 'synset': 'jewel.n.01', 'synonyms': ['jewel', 'gem', 'precious_stone'], 'id': 597, 'def': 'a precious or semiprecious stone incorporated into a piece of jewelry', 'name': 'jewel'}, {'frequency': 'c', 'synset': 'jewelry.n.01', 'synonyms': ['jewelry', 'jewellery'], 'id': 598, 'def': 'an adornment (as a bracelet or ring or necklace) made of precious metals and set with gems (or imitation gems)', 'name': 'jewelry'}, {'frequency': 'r', 'synset': 'joystick.n.02', 'synonyms': ['joystick'], 'id': 599, 'def': 'a control device for computers consisting of a vertical handle that can move freely in two directions', 'name': 'joystick'}, {'frequency': 'c', 'synset': 'jump_suit.n.01', 'synonyms': ['jumpsuit'], 'id': 600, 'def': "one-piece garment fashioned after a parachutist's uniform", 'name': 'jumpsuit'}, {'frequency': 'c', 'synset': 'kayak.n.01', 'synonyms': ['kayak'], 'id': 601, 'def': 'a small canoe consisting of a light frame made watertight with animal skins', 'name': 'kayak'}, {'frequency': 'r', 'synset': 'keg.n.02', 'synonyms': ['keg'], 'id': 602, 'def': 'small cask or barrel', 'name': 'keg'}, {'frequency': 'r', 'synset': 'kennel.n.01', 'synonyms': ['kennel', 'doghouse'], 'id': 603, 'def': 'outbuilding that serves as a shelter for a dog', 'name': 'kennel'}, {'frequency': 'c', 'synset': 'kettle.n.01', 'synonyms': ['kettle', 'boiler'], 'id': 604, 'def': 'a metal pot for stewing or boiling; usually has a lid', 'name': 'kettle'}, {'frequency': 'f', 'synset': 'key.n.01', 'synonyms': ['key'], 'id': 605, 'def': 'metal instrument used to unlock a lock', 'name': 'key'}, {'frequency': 'r', 'synset': 'keycard.n.01', 'synonyms': ['keycard'], 'id': 606, 'def': 'a plastic card used to gain access typically to a door', 'name': 'keycard'}, {'frequency': 'c', 'synset': 'kilt.n.01', 'synonyms': ['kilt'], 'id': 607, 'def': 'a knee-length pleated tartan skirt worn by men as part of the traditional dress in the Highlands of northern Scotland', 'name': 'kilt'}, {'frequency': 'c', 'synset': 'kimono.n.01', 'synonyms': ['kimono'], 'id': 608, 'def': 'a loose robe; imitated from robes originally worn by Japanese', 'name': 'kimono'}, {'frequency': 'f', 'synset': 'kitchen_sink.n.01', 'synonyms': ['kitchen_sink'], 'id': 609, 'def': 'a sink in a kitchen', 'name': 'kitchen_sink'}, {'frequency': 'r', 'synset': 'kitchen_table.n.01', 'synonyms': ['kitchen_table'], 'id': 610, 'def': 'a table in the kitchen', 'name': 'kitchen_table'}, {'frequency': 'f', 'synset': 'kite.n.03', 'synonyms': ['kite'], 'id': 611, 'def': 'plaything consisting of a light frame covered with tissue paper; flown in wind at end of a string', 'name': 'kite'}, {'frequency': 'c', 'synset': 'kitten.n.01', 'synonyms': ['kitten', 'kitty'], 'id': 612, 'def': 'young domestic cat', 'name': 'kitten'}, {'frequency': 'c', 'synset': 'kiwi.n.03', 'synonyms': ['kiwi_fruit'], 'id': 613, 'def': 'fuzzy brown egg-shaped fruit with slightly tart green flesh', 'name': 'kiwi_fruit'}, {'frequency': 'f', 'synset': 'knee_pad.n.01', 'synonyms': ['knee_pad'], 'id': 614, 'def': 'protective garment consisting of a pad worn by football or baseball or hockey players', 'name': 'knee_pad'}, {'frequency': 'f', 'synset': 'knife.n.01', 'synonyms': ['knife'], 'id': 615, 'def': 'tool with a blade and point used as a cutting instrument', 'name': 'knife'}, {'frequency': 'r', 'synset': 'knitting_needle.n.01', 'synonyms': ['knitting_needle'], 'id': 616, 'def': 'needle consisting of a slender rod with pointed ends; usually used in pairs', 'name': 'knitting_needle'}, {'frequency': 'f', 'synset': 'knob.n.02', 'synonyms': ['knob'], 'id': 617, 'def': 'a round handle often found on a door', 'name': 'knob'}, {'frequency': 'r', 'synset': 'knocker.n.05', 'synonyms': ['knocker_(on_a_door)', 'doorknocker'], 'id': 618, 'def': 'a device (usually metal and ornamental) attached by a hinge to a door', 'name': 'knocker_(on_a_door)'}, {'frequency': 'r', 'synset': 'koala.n.01', 'synonyms': ['koala', 'koala_bear'], 'id': 619, 'def': 'sluggish tailless Australian marsupial with grey furry ears and coat', 'name': 'koala'}, {'frequency': 'r', 'synset': 'lab_coat.n.01', 'synonyms': ['lab_coat', 'laboratory_coat'], 'id': 620, 'def': 'a light coat worn to protect clothing from substances used while working in a laboratory', 'name': 'lab_coat'}, {'frequency': 'f', 'synset': 'ladder.n.01', 'synonyms': ['ladder'], 'id': 621, 'def': 'steps consisting of two parallel members connected by rungs', 'name': 'ladder'}, {'frequency': 'c', 'synset': 'ladle.n.01', 'synonyms': ['ladle'], 'id': 622, 'def': 'a spoon-shaped vessel with a long handle frequently used to transfer liquids', 'name': 'ladle'}, {'frequency': 'c', 'synset': 'ladybug.n.01', 'synonyms': ['ladybug', 'ladybeetle', 'ladybird_beetle'], 'id': 623, 'def': 'small round bright-colored and spotted beetle, typically red and black', 'name': 'ladybug'}, {'frequency': 'f', 'synset': 'lamb.n.01', 'synonyms': ['lamb_(animal)'], 'id': 624, 'def': 'young sheep', 'name': 'lamb_(animal)'}, {'frequency': 'r', 'synset': 'lamb_chop.n.01', 'synonyms': ['lamb-chop', 'lambchop'], 'id': 625, 'def': 'chop cut from a lamb', 'name': 'lamb-chop'}, {'frequency': 'f', 'synset': 'lamp.n.02', 'synonyms': ['lamp'], 'id': 626, 'def': 'a piece of furniture holding one or more electric light bulbs', 'name': 'lamp'}, {'frequency': 'f', 'synset': 'lamppost.n.01', 'synonyms': ['lamppost'], 'id': 627, 'def': 'a metal post supporting an outdoor lamp (such as a streetlight)', 'name': 'lamppost'}, {'frequency': 'f', 'synset': 'lampshade.n.01', 'synonyms': ['lampshade'], 'id': 628, 'def': 'a protective ornamental shade used to screen a light bulb from direct view', 'name': 'lampshade'}, {'frequency': 'c', 'synset': 'lantern.n.01', 'synonyms': ['lantern'], 'id': 629, 'def': 'light in a transparent protective case', 'name': 'lantern'}, {'frequency': 'f', 'synset': 'lanyard.n.02', 'synonyms': ['lanyard', 'laniard'], 'id': 630, 'def': 'a cord worn around the neck to hold a knife or whistle, etc.', 'name': 'lanyard'}, {'frequency': 'f', 'synset': 'laptop.n.01', 'synonyms': ['laptop_computer', 'notebook_computer'], 'id': 631, 'def': 'a portable computer small enough to use in your lap', 'name': 'laptop_computer'}, {'frequency': 'r', 'synset': 'lasagna.n.01', 'synonyms': ['lasagna', 'lasagne'], 'id': 632, 'def': 'baked dish of layers of lasagna pasta with sauce and cheese and meat or vegetables', 'name': 'lasagna'}, {'frequency': 'f', 'synset': 'latch.n.02', 'synonyms': ['latch'], 'id': 633, 'def': 'a bar that can be lowered or slid into a groove to fasten a door or gate', 'name': 'latch'}, {'frequency': 'r', 'synset': 'lawn_mower.n.01', 'synonyms': ['lawn_mower'], 'id': 634, 'def': 'garden tool for mowing grass on lawns', 'name': 'lawn_mower'}, {'frequency': 'r', 'synset': 'leather.n.01', 'synonyms': ['leather'], 'id': 635, 'def': 'an animal skin made smooth and flexible by removing the hair and then tanning', 'name': 'leather'}, {'frequency': 'c', 'synset': 'legging.n.01', 'synonyms': ['legging_(clothing)', 'leging_(clothing)', 'leg_covering'], 'id': 636, 'def': 'a garment covering the leg (usually extending from the knee to the ankle)', 'name': 'legging_(clothing)'}, {'frequency': 'c', 'synset': 'lego.n.01', 'synonyms': ['Lego', 'Lego_set'], 'id': 637, 'def': "a child's plastic construction set for making models from blocks", 'name': 'Lego'}, {'frequency': 'r', 'synset': 'legume.n.02', 'synonyms': ['legume'], 'id': 638, 'def': 'the fruit or seed of bean or pea plants', 'name': 'legume'}, {'frequency': 'f', 'synset': 'lemon.n.01', 'synonyms': ['lemon'], 'id': 639, 'def': 'yellow oval fruit with juicy acidic flesh', 'name': 'lemon'}, {'frequency': 'r', 'synset': 'lemonade.n.01', 'synonyms': ['lemonade'], 'id': 640, 'def': 'sweetened beverage of diluted lemon juice', 'name': 'lemonade'}, {'frequency': 'f', 'synset': 'lettuce.n.02', 'synonyms': ['lettuce'], 'id': 641, 'def': 'leafy plant commonly eaten in salad or on sandwiches', 'name': 'lettuce'}, {'frequency': 'f', 'synset': 'license_plate.n.01', 'synonyms': ['license_plate', 'numberplate'], 'id': 642, 'def': "a plate mounted on the front and back of car and bearing the car's registration number", 'name': 'license_plate'}, {'frequency': 'f', 'synset': 'life_buoy.n.01', 'synonyms': ['life_buoy', 'lifesaver', 'life_belt', 'life_ring'], 'id': 643, 'def': 'a ring-shaped life preserver used to prevent drowning (NOT a life-jacket or vest)', 'name': 'life_buoy'}, {'frequency': 'f', 'synset': 'life_jacket.n.01', 'synonyms': ['life_jacket', 'life_vest'], 'id': 644, 'def': 'life preserver consisting of a sleeveless jacket of buoyant or inflatable design', 'name': 'life_jacket'}, {'frequency': 'f', 'synset': 'light_bulb.n.01', 'synonyms': ['lightbulb'], 'id': 645, 'def': 'lightblub/source of light', 'name': 'lightbulb'}, {'frequency': 'r', 'synset': 'lightning_rod.n.02', 'synonyms': ['lightning_rod', 'lightning_conductor'], 'id': 646, 'def': 'a metallic conductor that is attached to a high point and leads to the ground', 'name': 'lightning_rod'}, {'frequency': 'f', 'synset': 'lime.n.06', 'synonyms': ['lime'], 'id': 647, 'def': 'the green acidic fruit of any of various lime trees', 'name': 'lime'}, {'frequency': 'r', 'synset': 'limousine.n.01', 'synonyms': ['limousine'], 'id': 648, 'def': 'long luxurious car; usually driven by a chauffeur', 'name': 'limousine'}, {'frequency': 'c', 'synset': 'lion.n.01', 'synonyms': ['lion'], 'id': 649, 'def': 'large gregarious predatory cat of Africa and India', 'name': 'lion'}, {'frequency': 'c', 'synset': 'lip_balm.n.01', 'synonyms': ['lip_balm'], 'id': 650, 'def': 'a balm applied to the lips', 'name': 'lip_balm'}, {'frequency': 'r', 'synset': 'liquor.n.01', 'synonyms': ['liquor', 'spirits', 'hard_liquor', 'liqueur', 'cordial'], 'id': 651, 'def': 'liquor or beer', 'name': 'liquor'}, {'frequency': 'c', 'synset': 'lizard.n.01', 'synonyms': ['lizard'], 'id': 652, 'def': 'a reptile with usually two pairs of legs and a tapering tail', 'name': 'lizard'}, {'frequency': 'f', 'synset': 'log.n.01', 'synonyms': ['log'], 'id': 653, 'def': 'a segment of the trunk of a tree when stripped of branches', 'name': 'log'}, {'frequency': 'c', 'synset': 'lollipop.n.02', 'synonyms': ['lollipop'], 'id': 654, 'def': 'hard candy on a stick', 'name': 'lollipop'}, {'frequency': 'f', 'synset': 'loudspeaker.n.01', 'synonyms': ['speaker_(stero_equipment)'], 'id': 655, 'def': 'electronic device that produces sound often as part of a stereo system', 'name': 'speaker_(stero_equipment)'}, {'frequency': 'c', 'synset': 'love_seat.n.01', 'synonyms': ['loveseat'], 'id': 656, 'def': 'small sofa that seats two people', 'name': 'loveseat'}, {'frequency': 'r', 'synset': 'machine_gun.n.01', 'synonyms': ['machine_gun'], 'id': 657, 'def': 'a rapidly firing automatic gun', 'name': 'machine_gun'}, {'frequency': 'f', 'synset': 'magazine.n.02', 'synonyms': ['magazine'], 'id': 658, 'def': 'a paperback periodic publication', 'name': 'magazine'}, {'frequency': 'f', 'synset': 'magnet.n.01', 'synonyms': ['magnet'], 'id': 659, 'def': 'a device that attracts iron and produces a magnetic field', 'name': 'magnet'}, {'frequency': 'c', 'synset': 'mail_slot.n.01', 'synonyms': ['mail_slot'], 'id': 660, 'def': 'a slot (usually in a door) through which mail can be delivered', 'name': 'mail_slot'}, {'frequency': 'f', 'synset': 'mailbox.n.01', 'synonyms': ['mailbox_(at_home)', 'letter_box_(at_home)'], 'id': 661, 'def': 'a private box for delivery of mail', 'name': 'mailbox_(at_home)'}, {'frequency': 'r', 'synset': 'mallard.n.01', 'synonyms': ['mallard'], 'id': 662, 'def': 'wild dabbling duck from which domestic ducks are descended', 'name': 'mallard'}, {'frequency': 'r', 'synset': 'mallet.n.01', 'synonyms': ['mallet'], 'id': 663, 'def': 'a sports implement with a long handle and a hammer-like head used to hit a ball', 'name': 'mallet'}, {'frequency': 'r', 'synset': 'mammoth.n.01', 'synonyms': ['mammoth'], 'id': 664, 'def': 'any of numerous extinct elephants widely distributed in the Pleistocene', 'name': 'mammoth'}, {'frequency': 'r', 'synset': 'manatee.n.01', 'synonyms': ['manatee'], 'id': 665, 'def': 'sirenian mammal of tropical coastal waters of America', 'name': 'manatee'}, {'frequency': 'c', 'synset': 'mandarin.n.05', 'synonyms': ['mandarin_orange'], 'id': 666, 'def': 'a somewhat flat reddish-orange loose skinned citrus of China', 'name': 'mandarin_orange'}, {'frequency': 'c', 'synset': 'manger.n.01', 'synonyms': ['manger', 'trough'], 'id': 667, 'def': 'a container (usually in a barn or stable) from which cattle or horses feed', 'name': 'manger'}, {'frequency': 'f', 'synset': 'manhole.n.01', 'synonyms': ['manhole'], 'id': 668, 'def': 'a hole (usually with a flush cover) through which a person can gain access to an underground structure', 'name': 'manhole'}, {'frequency': 'f', 'synset': 'map.n.01', 'synonyms': ['map'], 'id': 669, 'def': "a diagrammatic representation of the earth's surface (or part of it)", 'name': 'map'}, {'frequency': 'f', 'synset': 'marker.n.03', 'synonyms': ['marker'], 'id': 670, 'def': 'a writing implement for making a mark', 'name': 'marker'}, {'frequency': 'r', 'synset': 'martini.n.01', 'synonyms': ['martini'], 'id': 671, 'def': 'a cocktail made of gin (or vodka) with dry vermouth', 'name': 'martini'}, {'frequency': 'r', 'synset': 'mascot.n.01', 'synonyms': ['mascot'], 'id': 672, 'def': 'a person or animal that is adopted by a team or other group as a symbolic figure', 'name': 'mascot'}, {'frequency': 'c', 'synset': 'mashed_potato.n.01', 'synonyms': ['mashed_potato'], 'id': 673, 'def': 'potato that has been peeled and boiled and then mashed', 'name': 'mashed_potato'}, {'frequency': 'r', 'synset': 'masher.n.02', 'synonyms': ['masher'], 'id': 674, 'def': 'a kitchen utensil used for mashing (e.g. potatoes)', 'name': 'masher'}, {'frequency': 'f', 'synset': 'mask.n.04', 'synonyms': ['mask', 'facemask'], 'id': 675, 'def': 'a protective covering worn over the face', 'name': 'mask'}, {'frequency': 'f', 'synset': 'mast.n.01', 'synonyms': ['mast'], 'id': 676, 'def': 'a vertical spar for supporting sails', 'name': 'mast'}, {'frequency': 'c', 'synset': 'mat.n.03', 'synonyms': ['mat_(gym_equipment)', 'gym_mat'], 'id': 677, 'def': 'sports equipment consisting of a piece of thick padding on the floor for gymnastics', 'name': 'mat_(gym_equipment)'}, {'frequency': 'r', 'synset': 'matchbox.n.01', 'synonyms': ['matchbox'], 'id': 678, 'def': 'a box for holding matches', 'name': 'matchbox'}, {'frequency': 'f', 'synset': 'mattress.n.01', 'synonyms': ['mattress'], 'id': 679, 'def': 'a thick pad filled with resilient material used as a bed or part of a bed', 'name': 'mattress'}, {'frequency': 'c', 'synset': 'measuring_cup.n.01', 'synonyms': ['measuring_cup'], 'id': 680, 'def': 'graduated cup used to measure liquid or granular ingredients', 'name': 'measuring_cup'}, {'frequency': 'c', 'synset': 'measuring_stick.n.01', 'synonyms': ['measuring_stick', 'ruler_(measuring_stick)', 'measuring_rod'], 'id': 681, 'def': 'measuring instrument having a sequence of marks at regular intervals', 'name': 'measuring_stick'}, {'frequency': 'c', 'synset': 'meatball.n.01', 'synonyms': ['meatball'], 'id': 682, 'def': 'ground meat formed into a ball and fried or simmered in broth', 'name': 'meatball'}, {'frequency': 'c', 'synset': 'medicine.n.02', 'synonyms': ['medicine'], 'id': 683, 'def': 'something that treats or prevents or alleviates the symptoms of disease', 'name': 'medicine'}, {'frequency': 'c', 'synset': 'melon.n.01', 'synonyms': ['melon'], 'id': 684, 'def': 'fruit of the gourd family having a hard rind and sweet juicy flesh', 'name': 'melon'}, {'frequency': 'f', 'synset': 'microphone.n.01', 'synonyms': ['microphone'], 'id': 685, 'def': 'device for converting sound waves into electrical energy', 'name': 'microphone'}, {'frequency': 'r', 'synset': 'microscope.n.01', 'synonyms': ['microscope'], 'id': 686, 'def': 'magnifier of the image of small objects', 'name': 'microscope'}, {'frequency': 'f', 'synset': 'microwave.n.02', 'synonyms': ['microwave_oven'], 'id': 687, 'def': 'kitchen appliance that cooks food by passing an electromagnetic wave through it', 'name': 'microwave_oven'}, {'frequency': 'r', 'synset': 'milestone.n.01', 'synonyms': ['milestone', 'milepost'], 'id': 688, 'def': 'stone post at side of a road to show distances', 'name': 'milestone'}, {'frequency': 'f', 'synset': 'milk.n.01', 'synonyms': ['milk'], 'id': 689, 'def': 'a white nutritious liquid secreted by mammals and used as food by human beings', 'name': 'milk'}, {'frequency': 'r', 'synset': 'milk_can.n.01', 'synonyms': ['milk_can'], 'id': 690, 'def': 'can for transporting milk', 'name': 'milk_can'}, {'frequency': 'r', 'synset': 'milkshake.n.01', 'synonyms': ['milkshake'], 'id': 691, 'def': 'frothy drink of milk and flavoring and sometimes fruit or ice cream', 'name': 'milkshake'}, {'frequency': 'f', 'synset': 'minivan.n.01', 'synonyms': ['minivan'], 'id': 692, 'def': 'a small box-shaped passenger van', 'name': 'minivan'}, {'frequency': 'r', 'synset': 'mint.n.05', 'synonyms': ['mint_candy'], 'id': 693, 'def': 'a candy that is flavored with a mint oil', 'name': 'mint_candy'}, {'frequency': 'f', 'synset': 'mirror.n.01', 'synonyms': ['mirror'], 'id': 694, 'def': 'polished surface that forms images by reflecting light', 'name': 'mirror'}, {'frequency': 'c', 'synset': 'mitten.n.01', 'synonyms': ['mitten'], 'id': 695, 'def': 'glove that encases the thumb separately and the other four fingers together', 'name': 'mitten'}, {'frequency': 'c', 'synset': 'mixer.n.04', 'synonyms': ['mixer_(kitchen_tool)', 'stand_mixer'], 'id': 696, 'def': 'a kitchen utensil that is used for mixing foods', 'name': 'mixer_(kitchen_tool)'}, {'frequency': 'c', 'synset': 'money.n.03', 'synonyms': ['money'], 'id': 697, 'def': 'the official currency issued by a government or national bank', 'name': 'money'}, {'frequency': 'f', 'synset': 'monitor.n.04', 'synonyms': ['monitor_(computer_equipment) computer_monitor'], 'id': 698, 'def': 'a computer monitor', 'name': 'monitor_(computer_equipment) computer_monitor'}, {'frequency': 'c', 'synset': 'monkey.n.01', 'synonyms': ['monkey'], 'id': 699, 'def': 'any of various long-tailed primates', 'name': 'monkey'}, {'frequency': 'f', 'synset': 'motor.n.01', 'synonyms': ['motor'], 'id': 700, 'def': 'machine that converts other forms of energy into mechanical energy and so imparts motion', 'name': 'motor'}, {'frequency': 'f', 'synset': 'motor_scooter.n.01', 'synonyms': ['motor_scooter', 'scooter'], 'id': 701, 'def': 'a wheeled vehicle with small wheels and a low-powered engine', 'name': 'motor_scooter'}, {'frequency': 'r', 'synset': 'motor_vehicle.n.01', 'synonyms': ['motor_vehicle', 'automotive_vehicle'], 'id': 702, 'def': 'a self-propelled wheeled vehicle that does not run on rails', 'name': 'motor_vehicle'}, {'frequency': 'f', 'synset': 'motorcycle.n.01', 'synonyms': ['motorcycle'], 'id': 703, 'def': 'a motor vehicle with two wheels and a strong frame', 'name': 'motorcycle'}, {'frequency': 'f', 'synset': 'mound.n.01', 'synonyms': ['mound_(baseball)', "pitcher's_mound"], 'id': 704, 'def': '(baseball) the slight elevation on which the pitcher stands', 'name': 'mound_(baseball)'}, {'frequency': 'f', 'synset': 'mouse.n.04', 'synonyms': ['mouse_(computer_equipment)', 'computer_mouse'], 'id': 705, 'def': 'a computer input device that controls an on-screen pointer (does not include trackpads / touchpads)', 'name': 'mouse_(computer_equipment)'}, {'frequency': 'f', 'synset': 'mousepad.n.01', 'synonyms': ['mousepad'], 'id': 706, 'def': 'a small portable pad that provides an operating surface for a computer mouse', 'name': 'mousepad'}, {'frequency': 'c', 'synset': 'muffin.n.01', 'synonyms': ['muffin'], 'id': 707, 'def': 'a sweet quick bread baked in a cup-shaped pan', 'name': 'muffin'}, {'frequency': 'f', 'synset': 'mug.n.04', 'synonyms': ['mug'], 'id': 708, 'def': 'with handle and usually cylindrical', 'name': 'mug'}, {'frequency': 'f', 'synset': 'mushroom.n.02', 'synonyms': ['mushroom'], 'id': 709, 'def': 'a common mushroom', 'name': 'mushroom'}, {'frequency': 'r', 'synset': 'music_stool.n.01', 'synonyms': ['music_stool', 'piano_stool'], 'id': 710, 'def': 'a stool for piano players; usually adjustable in height', 'name': 'music_stool'}, {'frequency': 'c', 'synset': 'musical_instrument.n.01', 'synonyms': ['musical_instrument', 'instrument_(musical)'], 'id': 711, 'def': 'any of various devices or contrivances that can be used to produce musical tones or sounds', 'name': 'musical_instrument'}, {'frequency': 'r', 'synset': 'nailfile.n.01', 'synonyms': ['nailfile'], 'id': 712, 'def': 'a small flat file for shaping the nails', 'name': 'nailfile'}, {'frequency': 'f', 'synset': 'napkin.n.01', 'synonyms': ['napkin', 'table_napkin', 'serviette'], 'id': 713, 'def': 'a small piece of table linen or paper that is used to wipe the mouth and to cover the lap in order to protect clothing', 'name': 'napkin'}, {'frequency': 'r', 'synset': 'neckerchief.n.01', 'synonyms': ['neckerchief'], 'id': 714, 'def': 'a kerchief worn around the neck', 'name': 'neckerchief'}, {'frequency': 'f', 'synset': 'necklace.n.01', 'synonyms': ['necklace'], 'id': 715, 'def': 'jewelry consisting of a cord or chain (often bearing gems) worn about the neck as an ornament', 'name': 'necklace'}, {'frequency': 'f', 'synset': 'necktie.n.01', 'synonyms': ['necktie', 'tie_(necktie)'], 'id': 716, 'def': 'neckwear consisting of a long narrow piece of material worn under a collar and tied in knot at the front', 'name': 'necktie'}, {'frequency': 'c', 'synset': 'needle.n.03', 'synonyms': ['needle'], 'id': 717, 'def': 'a sharp pointed implement (usually metal)', 'name': 'needle'}, {'frequency': 'c', 'synset': 'nest.n.01', 'synonyms': ['nest'], 'id': 718, 'def': 'a structure in which animals lay eggs or give birth to their young', 'name': 'nest'}, {'frequency': 'f', 'synset': 'newspaper.n.01', 'synonyms': ['newspaper', 'paper_(newspaper)'], 'id': 719, 'def': 'a daily or weekly publication on folded sheets containing news, articles, and advertisements', 'name': 'newspaper'}, {'frequency': 'c', 'synset': 'newsstand.n.01', 'synonyms': ['newsstand'], 'id': 720, 'def': 'a stall where newspapers and other periodicals are sold', 'name': 'newsstand'}, {'frequency': 'c', 'synset': 'nightwear.n.01', 'synonyms': ['nightshirt', 'nightwear', 'sleepwear', 'nightclothes'], 'id': 721, 'def': 'garments designed to be worn in bed', 'name': 'nightshirt'}, {'frequency': 'r', 'synset': 'nosebag.n.01', 'synonyms': ['nosebag_(for_animals)', 'feedbag'], 'id': 722, 'def': 'a canvas bag that is used to feed an animal (such as a horse); covers the muzzle and fastens at the top of the head', 'name': 'nosebag_(for_animals)'}, {'frequency': 'c', 'synset': 'noseband.n.01', 'synonyms': ['noseband_(for_animals)', 'nosepiece_(for_animals)'], 'id': 723, 'def': "a strap that is the part of a bridle that goes over the animal's nose", 'name': 'noseband_(for_animals)'}, {'frequency': 'f', 'synset': 'notebook.n.01', 'synonyms': ['notebook'], 'id': 724, 'def': 'a book with blank pages for recording notes or memoranda', 'name': 'notebook'}, {'frequency': 'c', 'synset': 'notepad.n.01', 'synonyms': ['notepad'], 'id': 725, 'def': 'a pad of paper for keeping notes', 'name': 'notepad'}, {'frequency': 'f', 'synset': 'nut.n.03', 'synonyms': ['nut'], 'id': 726, 'def': 'a small metal block (usually square or hexagonal) with internal screw thread to be fitted onto a bolt', 'name': 'nut'}, {'frequency': 'r', 'synset': 'nutcracker.n.01', 'synonyms': ['nutcracker'], 'id': 727, 'def': 'a hand tool used to crack nuts open', 'name': 'nutcracker'}, {'frequency': 'f', 'synset': 'oar.n.01', 'synonyms': ['oar'], 'id': 728, 'def': 'an implement used to propel or steer a boat', 'name': 'oar'}, {'frequency': 'r', 'synset': 'octopus.n.01', 'synonyms': ['octopus_(food)'], 'id': 729, 'def': 'tentacles of octopus prepared as food', 'name': 'octopus_(food)'}, {'frequency': 'r', 'synset': 'octopus.n.02', 'synonyms': ['octopus_(animal)'], 'id': 730, 'def': 'bottom-living cephalopod having a soft oval body with eight long tentacles', 'name': 'octopus_(animal)'}, {'frequency': 'c', 'synset': 'oil_lamp.n.01', 'synonyms': ['oil_lamp', 'kerosene_lamp', 'kerosine_lamp'], 'id': 731, 'def': 'a lamp that burns oil (as kerosine) for light', 'name': 'oil_lamp'}, {'frequency': 'c', 'synset': 'olive_oil.n.01', 'synonyms': ['olive_oil'], 'id': 732, 'def': 'oil from olives', 'name': 'olive_oil'}, {'frequency': 'r', 'synset': 'omelet.n.01', 'synonyms': ['omelet', 'omelette'], 'id': 733, 'def': 'beaten eggs cooked until just set; may be folded around e.g. ham or cheese or jelly', 'name': 'omelet'}, {'frequency': 'f', 'synset': 'onion.n.01', 'synonyms': ['onion'], 'id': 734, 'def': 'the bulb of an onion plant', 'name': 'onion'}, {'frequency': 'f', 'synset': 'orange.n.01', 'synonyms': ['orange_(fruit)'], 'id': 735, 'def': 'orange (FRUIT of an orange tree)', 'name': 'orange_(fruit)'}, {'frequency': 'c', 'synset': 'orange_juice.n.01', 'synonyms': ['orange_juice'], 'id': 736, 'def': 'bottled or freshly squeezed juice of oranges', 'name': 'orange_juice'}, {'frequency': 'c', 'synset': 'ostrich.n.02', 'synonyms': ['ostrich'], 'id': 737, 'def': 'fast-running African flightless bird with two-toed feet; largest living bird', 'name': 'ostrich'}, {'frequency': 'f', 'synset': 'ottoman.n.03', 'synonyms': ['ottoman', 'pouf', 'pouffe', 'hassock'], 'id': 738, 'def': 'a thick standalone cushion used as a seat or footrest, often next to a chair', 'name': 'ottoman'}, {'frequency': 'f', 'synset': 'oven.n.01', 'synonyms': ['oven'], 'id': 739, 'def': 'kitchen appliance used for baking or roasting', 'name': 'oven'}, {'frequency': 'c', 'synset': 'overall.n.01', 'synonyms': ['overalls_(clothing)'], 'id': 740, 'def': 'work clothing consisting of denim trousers usually with a bib and shoulder straps', 'name': 'overalls_(clothing)'}, {'frequency': 'c', 'synset': 'owl.n.01', 'synonyms': ['owl'], 'id': 741, 'def': 'nocturnal bird of prey with hawk-like beak and claws and large head with front-facing eyes', 'name': 'owl'}, {'frequency': 'c', 'synset': 'packet.n.03', 'synonyms': ['packet'], 'id': 742, 'def': 'a small package or bundle', 'name': 'packet'}, {'frequency': 'r', 'synset': 'pad.n.03', 'synonyms': ['inkpad', 'inking_pad', 'stamp_pad'], 'id': 743, 'def': 'absorbent material saturated with ink used to transfer ink evenly to a rubber stamp', 'name': 'inkpad'}, {'frequency': 'c', 'synset': 'pad.n.04', 'synonyms': ['pad'], 'id': 744, 'def': 'mostly arm/knee pads labeled', 'name': 'pad'}, {'frequency': 'f', 'synset': 'paddle.n.04', 'synonyms': ['paddle', 'boat_paddle'], 'id': 745, 'def': 'a short light oar used without an oarlock to propel a canoe or small boat', 'name': 'paddle'}, {'frequency': 'c', 'synset': 'padlock.n.01', 'synonyms': ['padlock'], 'id': 746, 'def': 'a detachable, portable lock', 'name': 'padlock'}, {'frequency': 'c', 'synset': 'paintbrush.n.01', 'synonyms': ['paintbrush'], 'id': 747, 'def': 'a brush used as an applicator to apply paint', 'name': 'paintbrush'}, {'frequency': 'f', 'synset': 'painting.n.01', 'synonyms': ['painting'], 'id': 748, 'def': 'graphic art consisting of an artistic composition made by applying paints to a surface', 'name': 'painting'}, {'frequency': 'f', 'synset': 'pajama.n.02', 'synonyms': ['pajamas', 'pyjamas'], 'id': 749, 'def': 'loose-fitting nightclothes worn for sleeping or lounging', 'name': 'pajamas'}, {'frequency': 'c', 'synset': 'palette.n.02', 'synonyms': ['palette', 'pallet'], 'id': 750, 'def': 'board that provides a flat surface on which artists mix paints and the range of colors used', 'name': 'palette'}, {'frequency': 'f', 'synset': 'pan.n.01', 'synonyms': ['pan_(for_cooking)', 'cooking_pan'], 'id': 751, 'def': 'cooking utensil consisting of a wide metal vessel', 'name': 'pan_(for_cooking)'}, {'frequency': 'r', 'synset': 'pan.n.03', 'synonyms': ['pan_(metal_container)'], 'id': 752, 'def': 'shallow container made of metal', 'name': 'pan_(metal_container)'}, {'frequency': 'c', 'synset': 'pancake.n.01', 'synonyms': ['pancake'], 'id': 753, 'def': 'a flat cake of thin batter fried on both sides on a griddle', 'name': 'pancake'}, {'frequency': 'r', 'synset': 'pantyhose.n.01', 'synonyms': ['pantyhose'], 'id': 754, 'def': "a woman's tights consisting of underpants and stockings", 'name': 'pantyhose'}, {'frequency': 'r', 'synset': 'papaya.n.02', 'synonyms': ['papaya'], 'id': 755, 'def': 'large oval melon-like tropical fruit with yellowish flesh', 'name': 'papaya'}, {'frequency': 'f', 'synset': 'paper_plate.n.01', 'synonyms': ['paper_plate'], 'id': 756, 'def': 'a disposable plate made of cardboard', 'name': 'paper_plate'}, {'frequency': 'f', 'synset': 'paper_towel.n.01', 'synonyms': ['paper_towel'], 'id': 757, 'def': 'a disposable towel made of absorbent paper', 'name': 'paper_towel'}, {'frequency': 'r', 'synset': 'paperback_book.n.01', 'synonyms': ['paperback_book', 'paper-back_book', 'softback_book', 'soft-cover_book'], 'id': 758, 'def': 'a book with paper covers', 'name': 'paperback_book'}, {'frequency': 'r', 'synset': 'paperweight.n.01', 'synonyms': ['paperweight'], 'id': 759, 'def': 'a weight used to hold down a stack of papers', 'name': 'paperweight'}, {'frequency': 'c', 'synset': 'parachute.n.01', 'synonyms': ['parachute'], 'id': 760, 'def': 'rescue equipment consisting of a device that fills with air and retards your fall', 'name': 'parachute'}, {'frequency': 'c', 'synset': 'parakeet.n.01', 'synonyms': ['parakeet', 'parrakeet', 'parroket', 'paraquet', 'paroquet', 'parroquet'], 'id': 761, 'def': 'any of numerous small slender long-tailed parrots', 'name': 'parakeet'}, {'frequency': 'c', 'synset': 'parasail.n.01', 'synonyms': ['parasail_(sports)'], 'id': 762, 'def': 'parachute that will lift a person up into the air when it is towed by a motorboat or a car', 'name': 'parasail_(sports)'}, {'frequency': 'c', 'synset': 'parasol.n.01', 'synonyms': ['parasol', 'sunshade'], 'id': 763, 'def': 'a handheld collapsible source of shade', 'name': 'parasol'}, {'frequency': 'r', 'synset': 'parchment.n.01', 'synonyms': ['parchment'], 'id': 764, 'def': 'a superior paper resembling sheepskin', 'name': 'parchment'}, {'frequency': 'c', 'synset': 'parka.n.01', 'synonyms': ['parka', 'anorak'], 'id': 765, 'def': "a kind of heavy jacket (`windcheater' is a British term)", 'name': 'parka'}, {'frequency': 'f', 'synset': 'parking_meter.n.01', 'synonyms': ['parking_meter'], 'id': 766, 'def': 'a coin-operated timer located next to a parking space', 'name': 'parking_meter'}, {'frequency': 'c', 'synset': 'parrot.n.01', 'synonyms': ['parrot'], 'id': 767, 'def': 'usually brightly colored tropical birds with short hooked beaks and the ability to mimic sounds', 'name': 'parrot'}, {'frequency': 'c', 'synset': 'passenger_car.n.01', 'synonyms': ['passenger_car_(part_of_a_train)', 'coach_(part_of_a_train)'], 'id': 768, 'def': 'a railcar where passengers ride', 'name': 'passenger_car_(part_of_a_train)'}, {'frequency': 'r', 'synset': 'passenger_ship.n.01', 'synonyms': ['passenger_ship'], 'id': 769, 'def': 'a ship built to carry passengers', 'name': 'passenger_ship'}, {'frequency': 'c', 'synset': 'passport.n.02', 'synonyms': ['passport'], 'id': 770, 'def': 'a document issued by a country to a citizen allowing that person to travel abroad and re-enter the home country', 'name': 'passport'}, {'frequency': 'f', 'synset': 'pastry.n.02', 'synonyms': ['pastry'], 'id': 771, 'def': 'any of various baked foods made of dough or batter', 'name': 'pastry'}, {'frequency': 'r', 'synset': 'patty.n.01', 'synonyms': ['patty_(food)'], 'id': 772, 'def': 'small flat mass of chopped food', 'name': 'patty_(food)'}, {'frequency': 'c', 'synset': 'pea.n.01', 'synonyms': ['pea_(food)'], 'id': 773, 'def': 'seed of a pea plant used for food', 'name': 'pea_(food)'}, {'frequency': 'c', 'synset': 'peach.n.03', 'synonyms': ['peach'], 'id': 774, 'def': 'downy juicy fruit with sweet yellowish or whitish flesh', 'name': 'peach'}, {'frequency': 'c', 'synset': 'peanut_butter.n.01', 'synonyms': ['peanut_butter'], 'id': 775, 'def': 'a spread made from ground peanuts', 'name': 'peanut_butter'}, {'frequency': 'f', 'synset': 'pear.n.01', 'synonyms': ['pear'], 'id': 776, 'def': 'sweet juicy gritty-textured fruit available in many varieties', 'name': 'pear'}, {'frequency': 'c', 'synset': 'peeler.n.03', 'synonyms': ['peeler_(tool_for_fruit_and_vegetables)'], 'id': 777, 'def': 'a device for peeling vegetables or fruits', 'name': 'peeler_(tool_for_fruit_and_vegetables)'}, {'frequency': 'r', 'synset': 'peg.n.04', 'synonyms': ['wooden_leg', 'pegleg'], 'id': 778, 'def': 'a prosthesis that replaces a missing leg', 'name': 'wooden_leg'}, {'frequency': 'r', 'synset': 'pegboard.n.01', 'synonyms': ['pegboard'], 'id': 779, 'def': 'a board perforated with regularly spaced holes into which pegs can be fitted', 'name': 'pegboard'}, {'frequency': 'c', 'synset': 'pelican.n.01', 'synonyms': ['pelican'], 'id': 780, 'def': 'large long-winged warm-water seabird having a large bill with a distensible pouch for fish', 'name': 'pelican'}, {'frequency': 'f', 'synset': 'pen.n.01', 'synonyms': ['pen'], 'id': 781, 'def': 'a writing implement with a point from which ink flows', 'name': 'pen'}, {'frequency': 'f', 'synset': 'pencil.n.01', 'synonyms': ['pencil'], 'id': 782, 'def': 'a thin cylindrical pointed writing implement made of wood and graphite', 'name': 'pencil'}, {'frequency': 'r', 'synset': 'pencil_box.n.01', 'synonyms': ['pencil_box', 'pencil_case'], 'id': 783, 'def': 'a box for holding pencils', 'name': 'pencil_box'}, {'frequency': 'r', 'synset': 'pencil_sharpener.n.01', 'synonyms': ['pencil_sharpener'], 'id': 784, 'def': 'a rotary implement for sharpening the point on pencils', 'name': 'pencil_sharpener'}, {'frequency': 'r', 'synset': 'pendulum.n.01', 'synonyms': ['pendulum'], 'id': 785, 'def': 'an apparatus consisting of an object mounted so that it swings freely under the influence of gravity', 'name': 'pendulum'}, {'frequency': 'c', 'synset': 'penguin.n.01', 'synonyms': ['penguin'], 'id': 786, 'def': 'short-legged flightless birds of cold southern regions having webbed feet and wings modified as flippers', 'name': 'penguin'}, {'frequency': 'r', 'synset': 'pennant.n.02', 'synonyms': ['pennant'], 'id': 787, 'def': 'a flag longer than it is wide (and often tapering)', 'name': 'pennant'}, {'frequency': 'r', 'synset': 'penny.n.02', 'synonyms': ['penny_(coin)'], 'id': 788, 'def': 'a coin worth one-hundredth of the value of the basic unit', 'name': 'penny_(coin)'}, {'frequency': 'f', 'synset': 'pepper.n.03', 'synonyms': ['pepper', 'peppercorn'], 'id': 789, 'def': 'pungent seasoning from the berry of the common pepper plant; whole or ground', 'name': 'pepper'}, {'frequency': 'c', 'synset': 'pepper_mill.n.01', 'synonyms': ['pepper_mill', 'pepper_grinder'], 'id': 790, 'def': 'a mill for grinding pepper', 'name': 'pepper_mill'}, {'frequency': 'c', 'synset': 'perfume.n.02', 'synonyms': ['perfume'], 'id': 791, 'def': 'a toiletry that emits and diffuses a fragrant odor', 'name': 'perfume'}, {'frequency': 'r', 'synset': 'persimmon.n.02', 'synonyms': ['persimmon'], 'id': 792, 'def': 'orange fruit resembling a plum; edible when fully ripe', 'name': 'persimmon'}, {'frequency': 'f', 'synset': 'person.n.01', 'synonyms': ['person', 'baby', 'child', 'boy', 'girl', 'man', 'woman', 'human'], 'id': 793, 'def': 'a human being', 'name': 'person'}, {'frequency': 'c', 'synset': 'pet.n.01', 'synonyms': ['pet'], 'id': 794, 'def': 'a domesticated animal kept for companionship or amusement', 'name': 'pet'}, {'frequency': 'c', 'synset': 'pew.n.01', 'synonyms': ['pew_(church_bench)', 'church_bench'], 'id': 795, 'def': 'long bench with backs; used in church by the congregation', 'name': 'pew_(church_bench)'}, {'frequency': 'r', 'synset': 'phonebook.n.01', 'synonyms': ['phonebook', 'telephone_book', 'telephone_directory'], 'id': 796, 'def': 'a directory containing an alphabetical list of telephone subscribers and their telephone numbers', 'name': 'phonebook'}, {'frequency': 'c', 'synset': 'phonograph_record.n.01', 'synonyms': ['phonograph_record', 'phonograph_recording', 'record_(phonograph_recording)'], 'id': 797, 'def': 'sound recording consisting of a typically black disk with a continuous groove', 'name': 'phonograph_record'}, {'frequency': 'f', 'synset': 'piano.n.01', 'synonyms': ['piano'], 'id': 798, 'def': 'a keyboard instrument that is played by depressing keys that cause hammers to strike tuned strings and produce sounds', 'name': 'piano'}, {'frequency': 'f', 'synset': 'pickle.n.01', 'synonyms': ['pickle'], 'id': 799, 'def': 'vegetables (especially cucumbers) preserved in brine or vinegar', 'name': 'pickle'}, {'frequency': 'f', 'synset': 'pickup.n.01', 'synonyms': ['pickup_truck'], 'id': 800, 'def': 'a light truck with an open body and low sides and a tailboard', 'name': 'pickup_truck'}, {'frequency': 'c', 'synset': 'pie.n.01', 'synonyms': ['pie'], 'id': 801, 'def': 'dish baked in pastry-lined pan often with a pastry top', 'name': 'pie'}, {'frequency': 'c', 'synset': 'pigeon.n.01', 'synonyms': ['pigeon'], 'id': 802, 'def': 'wild and domesticated birds having a heavy body and short legs', 'name': 'pigeon'}, {'frequency': 'r', 'synset': 'piggy_bank.n.01', 'synonyms': ['piggy_bank', 'penny_bank'], 'id': 803, 'def': "a child's coin bank (often shaped like a pig)", 'name': 'piggy_bank'}, {'frequency': 'f', 'synset': 'pillow.n.01', 'synonyms': ['pillow'], 'id': 804, 'def': 'a cushion to support the head of a sleeping person', 'name': 'pillow'}, {'frequency': 'r', 'synset': 'pin.n.09', 'synonyms': ['pin_(non_jewelry)'], 'id': 805, 'def': 'a small slender (often pointed) piece of wood or metal used to support or fasten or attach things', 'name': 'pin_(non_jewelry)'}, {'frequency': 'f', 'synset': 'pineapple.n.02', 'synonyms': ['pineapple'], 'id': 806, 'def': 'large sweet fleshy tropical fruit with a tuft of stiff leaves', 'name': 'pineapple'}, {'frequency': 'c', 'synset': 'pinecone.n.01', 'synonyms': ['pinecone'], 'id': 807, 'def': 'the seed-producing cone of a pine tree', 'name': 'pinecone'}, {'frequency': 'r', 'synset': 'ping-pong_ball.n.01', 'synonyms': ['ping-pong_ball'], 'id': 808, 'def': 'light hollow ball used in playing table tennis', 'name': 'ping-pong_ball'}, {'frequency': 'r', 'synset': 'pinwheel.n.03', 'synonyms': ['pinwheel'], 'id': 809, 'def': 'a toy consisting of vanes of colored paper or plastic that is pinned to a stick and spins when it is pointed into the wind', 'name': 'pinwheel'}, {'frequency': 'r', 'synset': 'pipe.n.01', 'synonyms': ['tobacco_pipe'], 'id': 810, 'def': 'a tube with a small bowl at one end; used for smoking tobacco', 'name': 'tobacco_pipe'}, {'frequency': 'f', 'synset': 'pipe.n.02', 'synonyms': ['pipe', 'piping'], 'id': 811, 'def': 'a long tube made of metal or plastic that is used to carry water or oil or gas etc.', 'name': 'pipe'}, {'frequency': 'r', 'synset': 'pistol.n.01', 'synonyms': ['pistol', 'handgun'], 'id': 812, 'def': 'a firearm that is held and fired with one hand', 'name': 'pistol'}, {'frequency': 'c', 'synset': 'pita.n.01', 'synonyms': ['pita_(bread)', 'pocket_bread'], 'id': 813, 'def': 'usually small round bread that can open into a pocket for filling', 'name': 'pita_(bread)'}, {'frequency': 'f', 'synset': 'pitcher.n.02', 'synonyms': ['pitcher_(vessel_for_liquid)', 'ewer'], 'id': 814, 'def': 'an open vessel with a handle and a spout for pouring', 'name': 'pitcher_(vessel_for_liquid)'}, {'frequency': 'r', 'synset': 'pitchfork.n.01', 'synonyms': ['pitchfork'], 'id': 815, 'def': 'a long-handled hand tool with sharp widely spaced prongs for lifting and pitching hay', 'name': 'pitchfork'}, {'frequency': 'f', 'synset': 'pizza.n.01', 'synonyms': ['pizza'], 'id': 816, 'def': 'Italian open pie made of thin bread dough spread with a spiced mixture of e.g. tomato sauce and cheese', 'name': 'pizza'}, {'frequency': 'f', 'synset': 'place_mat.n.01', 'synonyms': ['place_mat'], 'id': 817, 'def': 'a mat placed on a table for an individual place setting', 'name': 'place_mat'}, {'frequency': 'f', 'synset': 'plate.n.04', 'synonyms': ['plate'], 'id': 818, 'def': 'dish on which food is served or from which food is eaten', 'name': 'plate'}, {'frequency': 'c', 'synset': 'platter.n.01', 'synonyms': ['platter'], 'id': 819, 'def': 'a large shallow dish used for serving food', 'name': 'platter'}, {'frequency': 'r', 'synset': 'playpen.n.01', 'synonyms': ['playpen'], 'id': 820, 'def': 'a portable enclosure in which babies may be left to play', 'name': 'playpen'}, {'frequency': 'c', 'synset': 'pliers.n.01', 'synonyms': ['pliers', 'plyers'], 'id': 821, 'def': 'a gripping hand tool with two hinged arms and (usually) serrated jaws', 'name': 'pliers'}, {'frequency': 'r', 'synset': 'plow.n.01', 'synonyms': ['plow_(farm_equipment)', 'plough_(farm_equipment)'], 'id': 822, 'def': 'a farm tool having one or more heavy blades to break the soil and cut a furrow prior to sowing', 'name': 'plow_(farm_equipment)'}, {'frequency': 'r', 'synset': 'plume.n.02', 'synonyms': ['plume'], 'id': 823, 'def': 'a feather or cluster of feathers worn as an ornament', 'name': 'plume'}, {'frequency': 'r', 'synset': 'pocket_watch.n.01', 'synonyms': ['pocket_watch'], 'id': 824, 'def': 'a watch that is carried in a small watch pocket', 'name': 'pocket_watch'}, {'frequency': 'c', 'synset': 'pocketknife.n.01', 'synonyms': ['pocketknife'], 'id': 825, 'def': 'a knife with a blade that folds into the handle; suitable for carrying in the pocket', 'name': 'pocketknife'}, {'frequency': 'c', 'synset': 'poker.n.01', 'synonyms': ['poker_(fire_stirring_tool)', 'stove_poker', 'fire_hook'], 'id': 826, 'def': 'fire iron consisting of a metal rod with a handle; used to stir a fire', 'name': 'poker_(fire_stirring_tool)'}, {'frequency': 'f', 'synset': 'pole.n.01', 'synonyms': ['pole', 'post'], 'id': 827, 'def': 'a long (usually round) rod of wood or metal or plastic', 'name': 'pole'}, {'frequency': 'f', 'synset': 'polo_shirt.n.01', 'synonyms': ['polo_shirt', 'sport_shirt'], 'id': 828, 'def': 'a shirt with short sleeves designed for comfort and casual wear', 'name': 'polo_shirt'}, {'frequency': 'r', 'synset': 'poncho.n.01', 'synonyms': ['poncho'], 'id': 829, 'def': 'a blanket-like cloak with a hole in the center for the head', 'name': 'poncho'}, {'frequency': 'c', 'synset': 'pony.n.05', 'synonyms': ['pony'], 'id': 830, 'def': 'any of various breeds of small gentle horses usually less than five feet high at the shoulder', 'name': 'pony'}, {'frequency': 'r', 'synset': 'pool_table.n.01', 'synonyms': ['pool_table', 'billiard_table', 'snooker_table'], 'id': 831, 'def': 'game equipment consisting of a heavy table on which pool is played', 'name': 'pool_table'}, {'frequency': 'f', 'synset': 'pop.n.02', 'synonyms': ['pop_(soda)', 'soda_(pop)', 'tonic', 'soft_drink'], 'id': 832, 'def': 'a sweet drink containing carbonated water and flavoring', 'name': 'pop_(soda)'}, {'frequency': 'c', 'synset': 'postbox.n.01', 'synonyms': ['postbox_(public)', 'mailbox_(public)'], 'id': 833, 'def': 'public box for deposit of mail', 'name': 'postbox_(public)'}, {'frequency': 'c', 'synset': 'postcard.n.01', 'synonyms': ['postcard', 'postal_card', 'mailing-card'], 'id': 834, 'def': 'a card for sending messages by post without an envelope', 'name': 'postcard'}, {'frequency': 'f', 'synset': 'poster.n.01', 'synonyms': ['poster', 'placard'], 'id': 835, 'def': 'a sign posted in a public place as an advertisement', 'name': 'poster'}, {'frequency': 'f', 'synset': 'pot.n.01', 'synonyms': ['pot'], 'id': 836, 'def': 'metal or earthenware cooking vessel that is usually round and deep; often has a handle and lid', 'name': 'pot'}, {'frequency': 'f', 'synset': 'pot.n.04', 'synonyms': ['flowerpot'], 'id': 837, 'def': 'a container in which plants are cultivated', 'name': 'flowerpot'}, {'frequency': 'f', 'synset': 'potato.n.01', 'synonyms': ['potato'], 'id': 838, 'def': 'an edible tuber native to South America', 'name': 'potato'}, {'frequency': 'c', 'synset': 'potholder.n.01', 'synonyms': ['potholder'], 'id': 839, 'def': 'an insulated pad for holding hot pots', 'name': 'potholder'}, {'frequency': 'c', 'synset': 'pottery.n.01', 'synonyms': ['pottery', 'clayware'], 'id': 840, 'def': 'ceramic ware made from clay and baked in a kiln', 'name': 'pottery'}, {'frequency': 'c', 'synset': 'pouch.n.01', 'synonyms': ['pouch'], 'id': 841, 'def': 'a small or medium size container for holding or carrying things', 'name': 'pouch'}, {'frequency': 'c', 'synset': 'power_shovel.n.01', 'synonyms': ['power_shovel', 'excavator', 'digger'], 'id': 842, 'def': 'a machine for excavating', 'name': 'power_shovel'}, {'frequency': 'c', 'synset': 'prawn.n.01', 'synonyms': ['prawn', 'shrimp'], 'id': 843, 'def': 'any of various edible decapod crustaceans', 'name': 'prawn'}, {'frequency': 'c', 'synset': 'pretzel.n.01', 'synonyms': ['pretzel'], 'id': 844, 'def': 'glazed and salted cracker typically in the shape of a loose knot', 'name': 'pretzel'}, {'frequency': 'f', 'synset': 'printer.n.03', 'synonyms': ['printer', 'printing_machine'], 'id': 845, 'def': 'a machine that prints', 'name': 'printer'}, {'frequency': 'c', 'synset': 'projectile.n.01', 'synonyms': ['projectile_(weapon)', 'missile'], 'id': 846, 'def': 'a weapon that is forcibly thrown or projected at a targets', 'name': 'projectile_(weapon)'}, {'frequency': 'c', 'synset': 'projector.n.02', 'synonyms': ['projector'], 'id': 847, 'def': 'an optical instrument that projects an enlarged image onto a screen', 'name': 'projector'}, {'frequency': 'f', 'synset': 'propeller.n.01', 'synonyms': ['propeller', 'propellor'], 'id': 848, 'def': 'a mechanical device that rotates to push against air or water', 'name': 'propeller'}, {'frequency': 'r', 'synset': 'prune.n.01', 'synonyms': ['prune'], 'id': 849, 'def': 'dried plum', 'name': 'prune'}, {'frequency': 'r', 'synset': 'pudding.n.01', 'synonyms': ['pudding'], 'id': 850, 'def': 'any of various soft thick unsweetened baked dishes', 'name': 'pudding'}, {'frequency': 'r', 'synset': 'puffer.n.02', 'synonyms': ['puffer_(fish)', 'pufferfish', 'blowfish', 'globefish'], 'id': 851, 'def': 'fishes whose elongated spiny body can inflate itself with water or air to form a globe', 'name': 'puffer_(fish)'}, {'frequency': 'r', 'synset': 'puffin.n.01', 'synonyms': ['puffin'], 'id': 852, 'def': 'seabirds having short necks and brightly colored compressed bills', 'name': 'puffin'}, {'frequency': 'r', 'synset': 'pug.n.01', 'synonyms': ['pug-dog'], 'id': 853, 'def': 'small compact smooth-coated breed of Asiatic origin having a tightly curled tail and broad flat wrinkled muzzle', 'name': 'pug-dog'}, {'frequency': 'c', 'synset': 'pumpkin.n.02', 'synonyms': ['pumpkin'], 'id': 854, 'def': 'usually large pulpy deep-yellow round fruit of the squash family maturing in late summer or early autumn', 'name': 'pumpkin'}, {'frequency': 'r', 'synset': 'punch.n.03', 'synonyms': ['puncher'], 'id': 855, 'def': 'a tool for making holes or indentations', 'name': 'puncher'}, {'frequency': 'r', 'synset': 'puppet.n.01', 'synonyms': ['puppet', 'marionette'], 'id': 856, 'def': 'a small figure of a person operated from above with strings by a puppeteer', 'name': 'puppet'}, {'frequency': 'c', 'synset': 'puppy.n.01', 'synonyms': ['puppy'], 'id': 857, 'def': 'a young dog', 'name': 'puppy'}, {'frequency': 'r', 'synset': 'quesadilla.n.01', 'synonyms': ['quesadilla'], 'id': 858, 'def': 'a tortilla that is filled with cheese and heated', 'name': 'quesadilla'}, {'frequency': 'r', 'synset': 'quiche.n.02', 'synonyms': ['quiche'], 'id': 859, 'def': 'a tart filled with rich unsweetened custard; often contains other ingredients (as cheese or ham or seafood or vegetables)', 'name': 'quiche'}, {'frequency': 'f', 'synset': 'quilt.n.01', 'synonyms': ['quilt', 'comforter'], 'id': 860, 'def': 'bedding made of two layers of cloth filled with stuffing and stitched together', 'name': 'quilt'}, {'frequency': 'c', 'synset': 'rabbit.n.01', 'synonyms': ['rabbit'], 'id': 861, 'def': 'any of various burrowing animals of the family Leporidae having long ears and short tails', 'name': 'rabbit'}, {'frequency': 'r', 'synset': 'racer.n.02', 'synonyms': ['race_car', 'racing_car'], 'id': 862, 'def': 'a fast car that competes in races', 'name': 'race_car'}, {'frequency': 'c', 'synset': 'racket.n.04', 'synonyms': ['racket', 'racquet'], 'id': 863, 'def': 'a sports implement used to strike a ball in various games', 'name': 'racket'}, {'frequency': 'r', 'synset': 'radar.n.01', 'synonyms': ['radar'], 'id': 864, 'def': 'measuring instrument in which the echo of a pulse of microwave radiation is used to detect and locate distant objects', 'name': 'radar'}, {'frequency': 'f', 'synset': 'radiator.n.03', 'synonyms': ['radiator'], 'id': 865, 'def': 'a mechanism consisting of a metal honeycomb through which hot fluids circulate', 'name': 'radiator'}, {'frequency': 'c', 'synset': 'radio_receiver.n.01', 'synonyms': ['radio_receiver', 'radio_set', 'radio', 'tuner_(radio)'], 'id': 866, 'def': 'an electronic receiver that detects and demodulates and amplifies transmitted radio signals', 'name': 'radio_receiver'}, {'frequency': 'c', 'synset': 'radish.n.03', 'synonyms': ['radish', 'daikon'], 'id': 867, 'def': 'pungent edible root of any of various cultivated radish plants', 'name': 'radish'}, {'frequency': 'c', 'synset': 'raft.n.01', 'synonyms': ['raft'], 'id': 868, 'def': 'a flat float (usually made of logs or planks) that can be used for transport or as a platform for swimmers', 'name': 'raft'}, {'frequency': 'r', 'synset': 'rag_doll.n.01', 'synonyms': ['rag_doll'], 'id': 869, 'def': 'a cloth doll that is stuffed and (usually) painted', 'name': 'rag_doll'}, {'frequency': 'c', 'synset': 'raincoat.n.01', 'synonyms': ['raincoat', 'waterproof_jacket'], 'id': 870, 'def': 'a water-resistant coat', 'name': 'raincoat'}, {'frequency': 'c', 'synset': 'ram.n.05', 'synonyms': ['ram_(animal)'], 'id': 871, 'def': 'uncastrated adult male sheep', 'name': 'ram_(animal)'}, {'frequency': 'c', 'synset': 'raspberry.n.02', 'synonyms': ['raspberry'], 'id': 872, 'def': 'red or black edible aggregate berries usually smaller than the related blackberries', 'name': 'raspberry'}, {'frequency': 'r', 'synset': 'rat.n.01', 'synonyms': ['rat'], 'id': 873, 'def': 'any of various long-tailed rodents similar to but larger than a mouse', 'name': 'rat'}, {'frequency': 'c', 'synset': 'razorblade.n.01', 'synonyms': ['razorblade'], 'id': 874, 'def': 'a blade that has very sharp edge', 'name': 'razorblade'}, {'frequency': 'c', 'synset': 'reamer.n.01', 'synonyms': ['reamer_(juicer)', 'juicer', 'juice_reamer'], 'id': 875, 'def': 'a squeezer with a conical ridged center that is used for squeezing juice from citrus fruit', 'name': 'reamer_(juicer)'}, {'frequency': 'f', 'synset': 'rearview_mirror.n.01', 'synonyms': ['rearview_mirror'], 'id': 876, 'def': 'vehicle mirror (side or rearview)', 'name': 'rearview_mirror'}, {'frequency': 'c', 'synset': 'receipt.n.02', 'synonyms': ['receipt'], 'id': 877, 'def': 'an acknowledgment (usually tangible) that payment has been made', 'name': 'receipt'}, {'frequency': 'c', 'synset': 'recliner.n.01', 'synonyms': ['recliner', 'reclining_chair', 'lounger_(chair)'], 'id': 878, 'def': 'an armchair whose back can be lowered and foot can be raised to allow the sitter to recline in it', 'name': 'recliner'}, {'frequency': 'c', 'synset': 'record_player.n.01', 'synonyms': ['record_player', 'phonograph_(record_player)', 'turntable'], 'id': 879, 'def': 'machine in which rotating records cause a stylus to vibrate and the vibrations are amplified acoustically or electronically', 'name': 'record_player'}, {'frequency': 'f', 'synset': 'reflector.n.01', 'synonyms': ['reflector'], 'id': 880, 'def': 'device that reflects light, radiation, etc.', 'name': 'reflector'}, {'frequency': 'f', 'synset': 'remote_control.n.01', 'synonyms': ['remote_control'], 'id': 881, 'def': 'a device that can be used to control a machine or apparatus from a distance', 'name': 'remote_control'}, {'frequency': 'c', 'synset': 'rhinoceros.n.01', 'synonyms': ['rhinoceros'], 'id': 882, 'def': 'massive powerful herbivorous odd-toed ungulate of southeast Asia and Africa having very thick skin and one or two horns on the snout', 'name': 'rhinoceros'}, {'frequency': 'r', 'synset': 'rib.n.03', 'synonyms': ['rib_(food)'], 'id': 883, 'def': 'cut of meat including one or more ribs', 'name': 'rib_(food)'}, {'frequency': 'c', 'synset': 'rifle.n.01', 'synonyms': ['rifle'], 'id': 884, 'def': 'a shoulder firearm with a long barrel', 'name': 'rifle'}, {'frequency': 'f', 'synset': 'ring.n.08', 'synonyms': ['ring'], 'id': 885, 'def': 'jewelry consisting of a circlet of precious metal (often set with jewels) worn on the finger', 'name': 'ring'}, {'frequency': 'r', 'synset': 'river_boat.n.01', 'synonyms': ['river_boat'], 'id': 886, 'def': 'a boat used on rivers or to ply a river', 'name': 'river_boat'}, {'frequency': 'r', 'synset': 'road_map.n.02', 'synonyms': ['road_map'], 'id': 887, 'def': '(NOT A ROAD) a MAP showing roads (for automobile travel)', 'name': 'road_map'}, {'frequency': 'c', 'synset': 'robe.n.01', 'synonyms': ['robe'], 'id': 888, 'def': 'any loose flowing garment', 'name': 'robe'}, {'frequency': 'c', 'synset': 'rocking_chair.n.01', 'synonyms': ['rocking_chair'], 'id': 889, 'def': 'a chair mounted on rockers', 'name': 'rocking_chair'}, {'frequency': 'r', 'synset': 'rodent.n.01', 'synonyms': ['rodent'], 'id': 890, 'def': 'relatively small placental mammals having a single pair of constantly growing incisor teeth specialized for gnawing', 'name': 'rodent'}, {'frequency': 'r', 'synset': 'roller_skate.n.01', 'synonyms': ['roller_skate'], 'id': 891, 'def': 'a shoe with pairs of rollers (small hard wheels) fixed to the sole', 'name': 'roller_skate'}, {'frequency': 'r', 'synset': 'rollerblade.n.01', 'synonyms': ['Rollerblade'], 'id': 892, 'def': 'an in-line variant of a roller skate', 'name': 'Rollerblade'}, {'frequency': 'c', 'synset': 'rolling_pin.n.01', 'synonyms': ['rolling_pin'], 'id': 893, 'def': 'utensil consisting of a cylinder (usually of wood) with a handle at each end; used to roll out dough', 'name': 'rolling_pin'}, {'frequency': 'r', 'synset': 'root_beer.n.01', 'synonyms': ['root_beer'], 'id': 894, 'def': 'carbonated drink containing extracts of roots and herbs', 'name': 'root_beer'}, {'frequency': 'c', 'synset': 'router.n.02', 'synonyms': ['router_(computer_equipment)'], 'id': 895, 'def': 'a device that forwards data packets between computer networks', 'name': 'router_(computer_equipment)'}, {'frequency': 'f', 'synset': 'rubber_band.n.01', 'synonyms': ['rubber_band', 'elastic_band'], 'id': 896, 'def': 'a narrow band of elastic rubber used to hold things (such as papers) together', 'name': 'rubber_band'}, {'frequency': 'c', 'synset': 'runner.n.08', 'synonyms': ['runner_(carpet)'], 'id': 897, 'def': 'a long narrow carpet', 'name': 'runner_(carpet)'}, {'frequency': 'f', 'synset': 'sack.n.01', 'synonyms': ['plastic_bag', 'paper_bag'], 'id': 898, 'def': "a bag made of paper or plastic for holding customer's purchases", 'name': 'plastic_bag'}, {'frequency': 'f', 'synset': 'saddle.n.01', 'synonyms': ['saddle_(on_an_animal)'], 'id': 899, 'def': 'a seat for the rider of a horse or camel', 'name': 'saddle_(on_an_animal)'}, {'frequency': 'f', 'synset': 'saddle_blanket.n.01', 'synonyms': ['saddle_blanket', 'saddlecloth', 'horse_blanket'], 'id': 900, 'def': 'stable gear consisting of a blanket placed under the saddle', 'name': 'saddle_blanket'}, {'frequency': 'c', 'synset': 'saddlebag.n.01', 'synonyms': ['saddlebag'], 'id': 901, 'def': 'a large bag (or pair of bags) hung over a saddle', 'name': 'saddlebag'}, {'frequency': 'r', 'synset': 'safety_pin.n.01', 'synonyms': ['safety_pin'], 'id': 902, 'def': 'a pin in the form of a clasp; has a guard so the point of the pin will not stick the user', 'name': 'safety_pin'}, {'frequency': 'f', 'synset': 'sail.n.01', 'synonyms': ['sail'], 'id': 903, 'def': 'a large piece of fabric by means of which wind is used to propel a sailing vessel', 'name': 'sail'}, {'frequency': 'f', 'synset': 'salad.n.01', 'synonyms': ['salad'], 'id': 904, 'def': 'food mixtures either arranged on a plate or tossed and served with a moist dressing; usually consisting of or including greens', 'name': 'salad'}, {'frequency': 'r', 'synset': 'salad_plate.n.01', 'synonyms': ['salad_plate', 'salad_bowl'], 'id': 905, 'def': 'a plate or bowl for individual servings of salad', 'name': 'salad_plate'}, {'frequency': 'c', 'synset': 'salami.n.01', 'synonyms': ['salami'], 'id': 906, 'def': 'highly seasoned fatty sausage of pork and beef usually dried', 'name': 'salami'}, {'frequency': 'c', 'synset': 'salmon.n.01', 'synonyms': ['salmon_(fish)'], 'id': 907, 'def': 'any of various large food and game fishes of northern waters', 'name': 'salmon_(fish)'}, {'frequency': 'r', 'synset': 'salmon.n.03', 'synonyms': ['salmon_(food)'], 'id': 908, 'def': 'flesh of any of various marine or freshwater fish of the family Salmonidae', 'name': 'salmon_(food)'}, {'frequency': 'c', 'synset': 'salsa.n.01', 'synonyms': ['salsa'], 'id': 909, 'def': 'spicy sauce of tomatoes and onions and chili peppers to accompany Mexican foods', 'name': 'salsa'}, {'frequency': 'f', 'synset': 'saltshaker.n.01', 'synonyms': ['saltshaker'], 'id': 910, 'def': 'a shaker with a perforated top for sprinkling salt', 'name': 'saltshaker'}, {'frequency': 'f', 'synset': 'sandal.n.01', 'synonyms': ['sandal_(type_of_shoe)'], 'id': 911, 'def': 'a shoe consisting of a sole fastened by straps to the foot', 'name': 'sandal_(type_of_shoe)'}, {'frequency': 'f', 'synset': 'sandwich.n.01', 'synonyms': ['sandwich'], 'id': 912, 'def': 'two (or more) slices of bread with a filling between them', 'name': 'sandwich'}, {'frequency': 'r', 'synset': 'satchel.n.01', 'synonyms': ['satchel'], 'id': 913, 'def': 'luggage consisting of a small case with a flat bottom and (usually) a shoulder strap', 'name': 'satchel'}, {'frequency': 'r', 'synset': 'saucepan.n.01', 'synonyms': ['saucepan'], 'id': 914, 'def': 'a deep pan with a handle; used for stewing or boiling', 'name': 'saucepan'}, {'frequency': 'f', 'synset': 'saucer.n.02', 'synonyms': ['saucer'], 'id': 915, 'def': 'a small shallow dish for holding a cup at the table', 'name': 'saucer'}, {'frequency': 'f', 'synset': 'sausage.n.01', 'synonyms': ['sausage'], 'id': 916, 'def': 'highly seasoned minced meat stuffed in casings', 'name': 'sausage'}, {'frequency': 'r', 'synset': 'sawhorse.n.01', 'synonyms': ['sawhorse', 'sawbuck'], 'id': 917, 'def': 'a framework for holding wood that is being sawed', 'name': 'sawhorse'}, {'frequency': 'r', 'synset': 'sax.n.02', 'synonyms': ['saxophone'], 'id': 918, 'def': "a wind instrument with a `J'-shaped form typically made of brass", 'name': 'saxophone'}, {'frequency': 'f', 'synset': 'scale.n.07', 'synonyms': ['scale_(measuring_instrument)'], 'id': 919, 'def': 'a measuring instrument for weighing; shows amount of mass', 'name': 'scale_(measuring_instrument)'}, {'frequency': 'r', 'synset': 'scarecrow.n.01', 'synonyms': ['scarecrow', 'strawman'], 'id': 920, 'def': 'an effigy in the shape of a man to frighten birds away from seeds', 'name': 'scarecrow'}, {'frequency': 'f', 'synset': 'scarf.n.01', 'synonyms': ['scarf'], 'id': 921, 'def': 'a garment worn around the head or neck or shoulders for warmth or decoration', 'name': 'scarf'}, {'frequency': 'c', 'synset': 'school_bus.n.01', 'synonyms': ['school_bus'], 'id': 922, 'def': 'a bus used to transport children to or from school', 'name': 'school_bus'}, {'frequency': 'f', 'synset': 'scissors.n.01', 'synonyms': ['scissors'], 'id': 923, 'def': 'a tool having two crossed pivoting blades with looped handles', 'name': 'scissors'}, {'frequency': 'f', 'synset': 'scoreboard.n.01', 'synonyms': ['scoreboard'], 'id': 924, 'def': 'a large board for displaying the score of a contest (and some other information)', 'name': 'scoreboard'}, {'frequency': 'r', 'synset': 'scraper.n.01', 'synonyms': ['scraper'], 'id': 925, 'def': 'any of various hand tools for scraping', 'name': 'scraper'}, {'frequency': 'c', 'synset': 'screwdriver.n.01', 'synonyms': ['screwdriver'], 'id': 926, 'def': 'a hand tool for driving screws; has a tip that fits into the head of a screw', 'name': 'screwdriver'}, {'frequency': 'f', 'synset': 'scrub_brush.n.01', 'synonyms': ['scrubbing_brush'], 'id': 927, 'def': 'a brush with short stiff bristles for heavy cleaning', 'name': 'scrubbing_brush'}, {'frequency': 'c', 'synset': 'sculpture.n.01', 'synonyms': ['sculpture'], 'id': 928, 'def': 'a three-dimensional work of art', 'name': 'sculpture'}, {'frequency': 'c', 'synset': 'seabird.n.01', 'synonyms': ['seabird', 'seafowl'], 'id': 929, 'def': 'a bird that frequents coastal waters and the open ocean: gulls; pelicans; gannets; cormorants; albatrosses; petrels; etc.', 'name': 'seabird'}, {'frequency': 'c', 'synset': 'seahorse.n.02', 'synonyms': ['seahorse'], 'id': 930, 'def': 'small fish with horse-like heads bent sharply downward and curled tails', 'name': 'seahorse'}, {'frequency': 'r', 'synset': 'seaplane.n.01', 'synonyms': ['seaplane', 'hydroplane'], 'id': 931, 'def': 'an airplane that can land on or take off from water', 'name': 'seaplane'}, {'frequency': 'c', 'synset': 'seashell.n.01', 'synonyms': ['seashell'], 'id': 932, 'def': 'the shell of a marine organism', 'name': 'seashell'}, {'frequency': 'c', 'synset': 'sewing_machine.n.01', 'synonyms': ['sewing_machine'], 'id': 933, 'def': 'a textile machine used as a home appliance for sewing', 'name': 'sewing_machine'}, {'frequency': 'c', 'synset': 'shaker.n.03', 'synonyms': ['shaker'], 'id': 934, 'def': 'a container in which something can be shaken', 'name': 'shaker'}, {'frequency': 'c', 'synset': 'shampoo.n.01', 'synonyms': ['shampoo'], 'id': 935, 'def': 'cleansing agent consisting of soaps or detergents used for washing the hair', 'name': 'shampoo'}, {'frequency': 'c', 'synset': 'shark.n.01', 'synonyms': ['shark'], 'id': 936, 'def': 'typically large carnivorous fishes with sharpe teeth', 'name': 'shark'}, {'frequency': 'r', 'synset': 'sharpener.n.01', 'synonyms': ['sharpener'], 'id': 937, 'def': 'any implement that is used to make something (an edge or a point) sharper', 'name': 'sharpener'}, {'frequency': 'r', 'synset': 'sharpie.n.03', 'synonyms': ['Sharpie'], 'id': 938, 'def': 'a pen with indelible ink that will write on any surface', 'name': 'Sharpie'}, {'frequency': 'r', 'synset': 'shaver.n.03', 'synonyms': ['shaver_(electric)', 'electric_shaver', 'electric_razor'], 'id': 939, 'def': 'a razor powered by an electric motor', 'name': 'shaver_(electric)'}, {'frequency': 'c', 'synset': 'shaving_cream.n.01', 'synonyms': ['shaving_cream', 'shaving_soap'], 'id': 940, 'def': 'toiletry consisting that forms a rich lather for softening the beard before shaving', 'name': 'shaving_cream'}, {'frequency': 'r', 'synset': 'shawl.n.01', 'synonyms': ['shawl'], 'id': 941, 'def': 'cloak consisting of an oblong piece of cloth used to cover the head and shoulders', 'name': 'shawl'}, {'frequency': 'r', 'synset': 'shears.n.01', 'synonyms': ['shears'], 'id': 942, 'def': 'large scissors with strong blades', 'name': 'shears'}, {'frequency': 'f', 'synset': 'sheep.n.01', 'synonyms': ['sheep'], 'id': 943, 'def': 'woolly usually horned ruminant mammal related to the goat', 'name': 'sheep'}, {'frequency': 'r', 'synset': 'shepherd_dog.n.01', 'synonyms': ['shepherd_dog', 'sheepdog'], 'id': 944, 'def': 'any of various usually long-haired breeds of dog reared to herd and guard sheep', 'name': 'shepherd_dog'}, {'frequency': 'r', 'synset': 'sherbert.n.01', 'synonyms': ['sherbert', 'sherbet'], 'id': 945, 'def': 'a frozen dessert made primarily of fruit juice and sugar', 'name': 'sherbert'}, {'frequency': 'c', 'synset': 'shield.n.02', 'synonyms': ['shield'], 'id': 946, 'def': 'armor carried on the arm to intercept blows', 'name': 'shield'}, {'frequency': 'f', 'synset': 'shirt.n.01', 'synonyms': ['shirt'], 'id': 947, 'def': 'a garment worn on the upper half of the body', 'name': 'shirt'}, {'frequency': 'f', 'synset': 'shoe.n.01', 'synonyms': ['shoe', 'sneaker_(type_of_shoe)', 'tennis_shoe'], 'id': 948, 'def': 'common footwear covering the foot', 'name': 'shoe'}, {'frequency': 'f', 'synset': 'shopping_bag.n.01', 'synonyms': ['shopping_bag'], 'id': 949, 'def': 'a bag made of plastic or strong paper (often with handles); used to transport goods after shopping', 'name': 'shopping_bag'}, {'frequency': 'c', 'synset': 'shopping_cart.n.01', 'synonyms': ['shopping_cart'], 'id': 950, 'def': 'a handcart that holds groceries or other goods while shopping', 'name': 'shopping_cart'}, {'frequency': 'f', 'synset': 'short_pants.n.01', 'synonyms': ['short_pants', 'shorts_(clothing)', 'trunks_(clothing)'], 'id': 951, 'def': 'trousers that end at or above the knee', 'name': 'short_pants'}, {'frequency': 'r', 'synset': 'shot_glass.n.01', 'synonyms': ['shot_glass'], 'id': 952, 'def': 'a small glass adequate to hold a single swallow of whiskey', 'name': 'shot_glass'}, {'frequency': 'f', 'synset': 'shoulder_bag.n.01', 'synonyms': ['shoulder_bag'], 'id': 953, 'def': 'a large handbag that can be carried by a strap looped over the shoulder', 'name': 'shoulder_bag'}, {'frequency': 'c', 'synset': 'shovel.n.01', 'synonyms': ['shovel'], 'id': 954, 'def': 'a hand tool for lifting loose material such as snow, dirt, etc.', 'name': 'shovel'}, {'frequency': 'f', 'synset': 'shower.n.01', 'synonyms': ['shower_head'], 'id': 955, 'def': 'a plumbing fixture that sprays water over you', 'name': 'shower_head'}, {'frequency': 'r', 'synset': 'shower_cap.n.01', 'synonyms': ['shower_cap'], 'id': 956, 'def': 'a tight cap worn to keep hair dry while showering', 'name': 'shower_cap'}, {'frequency': 'f', 'synset': 'shower_curtain.n.01', 'synonyms': ['shower_curtain'], 'id': 957, 'def': 'a curtain that keeps water from splashing out of the shower area', 'name': 'shower_curtain'}, {'frequency': 'r', 'synset': 'shredder.n.01', 'synonyms': ['shredder_(for_paper)'], 'id': 958, 'def': 'a device that shreds documents', 'name': 'shredder_(for_paper)'}, {'frequency': 'f', 'synset': 'signboard.n.01', 'synonyms': ['signboard'], 'id': 959, 'def': 'structure displaying a board on which advertisements can be posted', 'name': 'signboard'}, {'frequency': 'c', 'synset': 'silo.n.01', 'synonyms': ['silo'], 'id': 960, 'def': 'a cylindrical tower used for storing goods', 'name': 'silo'}, {'frequency': 'f', 'synset': 'sink.n.01', 'synonyms': ['sink'], 'id': 961, 'def': 'plumbing fixture consisting of a water basin fixed to a wall or floor and having a drainpipe', 'name': 'sink'}, {'frequency': 'f', 'synset': 'skateboard.n.01', 'synonyms': ['skateboard'], 'id': 962, 'def': 'a board with wheels that is ridden in a standing or crouching position and propelled by foot', 'name': 'skateboard'}, {'frequency': 'c', 'synset': 'skewer.n.01', 'synonyms': ['skewer'], 'id': 963, 'def': 'a long pin for holding meat in position while it is being roasted', 'name': 'skewer'}, {'frequency': 'f', 'synset': 'ski.n.01', 'synonyms': ['ski'], 'id': 964, 'def': 'sports equipment for skiing on snow', 'name': 'ski'}, {'frequency': 'f', 'synset': 'ski_boot.n.01', 'synonyms': ['ski_boot'], 'id': 965, 'def': 'a stiff boot that is fastened to a ski with a ski binding', 'name': 'ski_boot'}, {'frequency': 'f', 'synset': 'ski_parka.n.01', 'synonyms': ['ski_parka', 'ski_jacket'], 'id': 966, 'def': 'a parka to be worn while skiing', 'name': 'ski_parka'}, {'frequency': 'f', 'synset': 'ski_pole.n.01', 'synonyms': ['ski_pole'], 'id': 967, 'def': 'a pole with metal points used as an aid in skiing', 'name': 'ski_pole'}, {'frequency': 'f', 'synset': 'skirt.n.02', 'synonyms': ['skirt'], 'id': 968, 'def': 'a garment hanging from the waist; worn mainly by girls and women', 'name': 'skirt'}, {'frequency': 'r', 'synset': 'skullcap.n.01', 'synonyms': ['skullcap'], 'id': 969, 'def': 'rounded brimless cap fitting the crown of the head', 'name': 'skullcap'}, {'frequency': 'c', 'synset': 'sled.n.01', 'synonyms': ['sled', 'sledge', 'sleigh'], 'id': 970, 'def': 'a vehicle or flat object for transportation over snow by sliding or pulled by dogs, etc.', 'name': 'sled'}, {'frequency': 'c', 'synset': 'sleeping_bag.n.01', 'synonyms': ['sleeping_bag'], 'id': 971, 'def': 'large padded bag designed to be slept in outdoors', 'name': 'sleeping_bag'}, {'frequency': 'r', 'synset': 'sling.n.05', 'synonyms': ['sling_(bandage)', 'triangular_bandage'], 'id': 972, 'def': 'bandage to support an injured forearm; slung over the shoulder or neck', 'name': 'sling_(bandage)'}, {'frequency': 'c', 'synset': 'slipper.n.01', 'synonyms': ['slipper_(footwear)', 'carpet_slipper_(footwear)'], 'id': 973, 'def': 'low footwear that can be slipped on and off easily; usually worn indoors', 'name': 'slipper_(footwear)'}, {'frequency': 'r', 'synset': 'smoothie.n.02', 'synonyms': ['smoothie'], 'id': 974, 'def': 'a thick smooth drink consisting of fresh fruit pureed with ice cream or yoghurt or milk', 'name': 'smoothie'}, {'frequency': 'r', 'synset': 'snake.n.01', 'synonyms': ['snake', 'serpent'], 'id': 975, 'def': 'limbless scaly elongate reptile; some are venomous', 'name': 'snake'}, {'frequency': 'f', 'synset': 'snowboard.n.01', 'synonyms': ['snowboard'], 'id': 976, 'def': 'a board that resembles a broad ski or a small surfboard; used in a standing position to slide down snow-covered slopes', 'name': 'snowboard'}, {'frequency': 'c', 'synset': 'snowman.n.01', 'synonyms': ['snowman'], 'id': 977, 'def': 'a figure of a person made of packed snow', 'name': 'snowman'}, {'frequency': 'c', 'synset': 'snowmobile.n.01', 'synonyms': ['snowmobile'], 'id': 978, 'def': 'tracked vehicle for travel on snow having skis in front', 'name': 'snowmobile'}, {'frequency': 'f', 'synset': 'soap.n.01', 'synonyms': ['soap'], 'id': 979, 'def': 'a cleansing agent made from the salts of vegetable or animal fats', 'name': 'soap'}, {'frequency': 'f', 'synset': 'soccer_ball.n.01', 'synonyms': ['soccer_ball'], 'id': 980, 'def': "an inflated ball used in playing soccer (called `football' outside of the United States)", 'name': 'soccer_ball'}, {'frequency': 'f', 'synset': 'sock.n.01', 'synonyms': ['sock'], 'id': 981, 'def': 'cloth covering for the foot; worn inside the shoe; reaches to between the ankle and the knee', 'name': 'sock'}, {'frequency': 'f', 'synset': 'sofa.n.01', 'synonyms': ['sofa', 'couch', 'lounge'], 'id': 982, 'def': 'an upholstered seat for more than one person', 'name': 'sofa'}, {'frequency': 'r', 'synset': 'softball.n.01', 'synonyms': ['softball'], 'id': 983, 'def': 'ball used in playing softball', 'name': 'softball'}, {'frequency': 'c', 'synset': 'solar_array.n.01', 'synonyms': ['solar_array', 'solar_battery', 'solar_panel'], 'id': 984, 'def': 'electrical device consisting of a large array of connected solar cells', 'name': 'solar_array'}, {'frequency': 'r', 'synset': 'sombrero.n.02', 'synonyms': ['sombrero'], 'id': 985, 'def': 'a straw hat with a tall crown and broad brim; worn in American southwest and in Mexico', 'name': 'sombrero'}, {'frequency': 'f', 'synset': 'soup.n.01', 'synonyms': ['soup'], 'id': 986, 'def': 'liquid food especially of meat or fish or vegetable stock often containing pieces of solid food', 'name': 'soup'}, {'frequency': 'r', 'synset': 'soup_bowl.n.01', 'synonyms': ['soup_bowl'], 'id': 987, 'def': 'a bowl for serving soup', 'name': 'soup_bowl'}, {'frequency': 'c', 'synset': 'soupspoon.n.01', 'synonyms': ['soupspoon'], 'id': 988, 'def': 'a spoon with a rounded bowl for eating soup', 'name': 'soupspoon'}, {'frequency': 'c', 'synset': 'sour_cream.n.01', 'synonyms': ['sour_cream', 'soured_cream'], 'id': 989, 'def': 'soured light cream', 'name': 'sour_cream'}, {'frequency': 'r', 'synset': 'soya_milk.n.01', 'synonyms': ['soya_milk', 'soybean_milk', 'soymilk'], 'id': 990, 'def': 'a milk substitute containing soybean flour and water; used in some infant formulas and in making tofu', 'name': 'soya_milk'}, {'frequency': 'r', 'synset': 'space_shuttle.n.01', 'synonyms': ['space_shuttle'], 'id': 991, 'def': "a reusable spacecraft with wings for a controlled descent through the Earth's atmosphere", 'name': 'space_shuttle'}, {'frequency': 'r', 'synset': 'sparkler.n.02', 'synonyms': ['sparkler_(fireworks)'], 'id': 992, 'def': 'a firework that burns slowly and throws out a shower of sparks', 'name': 'sparkler_(fireworks)'}, {'frequency': 'f', 'synset': 'spatula.n.02', 'synonyms': ['spatula'], 'id': 993, 'def': 'a hand tool with a thin flexible blade used to mix or spread soft substances', 'name': 'spatula'}, {'frequency': 'r', 'synset': 'spear.n.01', 'synonyms': ['spear', 'lance'], 'id': 994, 'def': 'a long pointed rod used as a tool or weapon', 'name': 'spear'}, {'frequency': 'f', 'synset': 'spectacles.n.01', 'synonyms': ['spectacles', 'specs', 'eyeglasses', 'glasses'], 'id': 995, 'def': 'optical instrument consisting of a frame that holds a pair of lenses for correcting defective vision', 'name': 'spectacles'}, {'frequency': 'c', 'synset': 'spice_rack.n.01', 'synonyms': ['spice_rack'], 'id': 996, 'def': 'a rack for displaying containers filled with spices', 'name': 'spice_rack'}, {'frequency': 'c', 'synset': 'spider.n.01', 'synonyms': ['spider'], 'id': 997, 'def': 'predatory arachnid with eight legs, two poison fangs, two feelers, and usually two silk-spinning organs at the back end of the body', 'name': 'spider'}, {'frequency': 'r', 'synset': 'spiny_lobster.n.02', 'synonyms': ['crawfish', 'crayfish'], 'id': 998, 'def': 'large edible marine crustacean having a spiny carapace but lacking the large pincers of true lobsters', 'name': 'crawfish'}, {'frequency': 'c', 'synset': 'sponge.n.01', 'synonyms': ['sponge'], 'id': 999, 'def': 'a porous mass usable to absorb water typically used for cleaning', 'name': 'sponge'}, {'frequency': 'f', 'synset': 'spoon.n.01', 'synonyms': ['spoon'], 'id': 1000, 'def': 'a piece of cutlery with a shallow bowl-shaped container and a handle', 'name': 'spoon'}, {'frequency': 'c', 'synset': 'sportswear.n.01', 'synonyms': ['sportswear', 'athletic_wear', 'activewear'], 'id': 1001, 'def': 'attire worn for sport or for casual wear', 'name': 'sportswear'}, {'frequency': 'c', 'synset': 'spotlight.n.02', 'synonyms': ['spotlight'], 'id': 1002, 'def': 'a lamp that produces a strong beam of light to illuminate a restricted area; used to focus attention of a stage performer', 'name': 'spotlight'}, {'frequency': 'r', 'synset': 'squid.n.01', 'synonyms': ['squid_(food)', 'calamari', 'calamary'], 'id': 1003, 'def': '(Italian cuisine) squid prepared as food', 'name': 'squid_(food)'}, {'frequency': 'c', 'synset': 'squirrel.n.01', 'synonyms': ['squirrel'], 'id': 1004, 'def': 'a kind of arboreal rodent having a long bushy tail', 'name': 'squirrel'}, {'frequency': 'r', 'synset': 'stagecoach.n.01', 'synonyms': ['stagecoach'], 'id': 1005, 'def': 'a large coach-and-four formerly used to carry passengers and mail on regular routes between towns', 'name': 'stagecoach'}, {'frequency': 'c', 'synset': 'stapler.n.01', 'synonyms': ['stapler_(stapling_machine)'], 'id': 1006, 'def': 'a machine that inserts staples into sheets of paper in order to fasten them together', 'name': 'stapler_(stapling_machine)'}, {'frequency': 'c', 'synset': 'starfish.n.01', 'synonyms': ['starfish', 'sea_star'], 'id': 1007, 'def': 'echinoderms characterized by five arms extending from a central disk', 'name': 'starfish'}, {'frequency': 'f', 'synset': 'statue.n.01', 'synonyms': ['statue_(sculpture)'], 'id': 1008, 'def': 'a sculpture representing a human or animal', 'name': 'statue_(sculpture)'}, {'frequency': 'c', 'synset': 'steak.n.01', 'synonyms': ['steak_(food)'], 'id': 1009, 'def': 'a slice of meat cut from the fleshy part of an animal or large fish', 'name': 'steak_(food)'}, {'frequency': 'r', 'synset': 'steak_knife.n.01', 'synonyms': ['steak_knife'], 'id': 1010, 'def': 'a sharp table knife used in eating steak', 'name': 'steak_knife'}, {'frequency': 'f', 'synset': 'steering_wheel.n.01', 'synonyms': ['steering_wheel'], 'id': 1011, 'def': 'a handwheel that is used for steering', 'name': 'steering_wheel'}, {'frequency': 'r', 'synset': 'step_ladder.n.01', 'synonyms': ['stepladder'], 'id': 1012, 'def': 'a folding portable ladder hinged at the top', 'name': 'stepladder'}, {'frequency': 'c', 'synset': 'step_stool.n.01', 'synonyms': ['step_stool'], 'id': 1013, 'def': 'a stool that has one or two steps that fold under the seat', 'name': 'step_stool'}, {'frequency': 'c', 'synset': 'stereo.n.01', 'synonyms': ['stereo_(sound_system)'], 'id': 1014, 'def': 'electronic device for playing audio', 'name': 'stereo_(sound_system)'}, {'frequency': 'r', 'synset': 'stew.n.02', 'synonyms': ['stew'], 'id': 1015, 'def': 'food prepared by stewing especially meat or fish with vegetables', 'name': 'stew'}, {'frequency': 'r', 'synset': 'stirrer.n.02', 'synonyms': ['stirrer'], 'id': 1016, 'def': 'an implement used for stirring', 'name': 'stirrer'}, {'frequency': 'f', 'synset': 'stirrup.n.01', 'synonyms': ['stirrup'], 'id': 1017, 'def': "support consisting of metal loops into which rider's feet go", 'name': 'stirrup'}, {'frequency': 'f', 'synset': 'stool.n.01', 'synonyms': ['stool'], 'id': 1018, 'def': 'a simple seat without a back or arms', 'name': 'stool'}, {'frequency': 'f', 'synset': 'stop_sign.n.01', 'synonyms': ['stop_sign'], 'id': 1019, 'def': 'a traffic sign to notify drivers that they must come to a complete stop', 'name': 'stop_sign'}, {'frequency': 'f', 'synset': 'stoplight.n.01', 'synonyms': ['brake_light'], 'id': 1020, 'def': 'a red light on the rear of a motor vehicle that signals when the brakes are applied', 'name': 'brake_light'}, {'frequency': 'f', 'synset': 'stove.n.01', 'synonyms': ['stove', 'kitchen_stove', 'range_(kitchen_appliance)', 'kitchen_range', 'cooking_stove'], 'id': 1021, 'def': 'a kitchen appliance used for cooking food', 'name': 'stove'}, {'frequency': 'c', 'synset': 'strainer.n.01', 'synonyms': ['strainer'], 'id': 1022, 'def': 'a filter to retain larger pieces while smaller pieces and liquids pass through', 'name': 'strainer'}, {'frequency': 'f', 'synset': 'strap.n.01', 'synonyms': ['strap'], 'id': 1023, 'def': 'an elongated strip of material for binding things together or holding', 'name': 'strap'}, {'frequency': 'f', 'synset': 'straw.n.04', 'synonyms': ['straw_(for_drinking)', 'drinking_straw'], 'id': 1024, 'def': 'a thin paper or plastic tube used to suck liquids into the mouth', 'name': 'straw_(for_drinking)'}, {'frequency': 'f', 'synset': 'strawberry.n.01', 'synonyms': ['strawberry'], 'id': 1025, 'def': 'sweet fleshy red fruit', 'name': 'strawberry'}, {'frequency': 'f', 'synset': 'street_sign.n.01', 'synonyms': ['street_sign'], 'id': 1026, 'def': 'a sign visible from the street', 'name': 'street_sign'}, {'frequency': 'f', 'synset': 'streetlight.n.01', 'synonyms': ['streetlight', 'street_lamp'], 'id': 1027, 'def': 'a lamp supported on a lamppost; for illuminating a street', 'name': 'streetlight'}, {'frequency': 'r', 'synset': 'string_cheese.n.01', 'synonyms': ['string_cheese'], 'id': 1028, 'def': 'cheese formed in long strings twisted together', 'name': 'string_cheese'}, {'frequency': 'r', 'synset': 'stylus.n.02', 'synonyms': ['stylus'], 'id': 1029, 'def': 'a pointed tool for writing or drawing or engraving, including pens', 'name': 'stylus'}, {'frequency': 'r', 'synset': 'subwoofer.n.01', 'synonyms': ['subwoofer'], 'id': 1030, 'def': 'a loudspeaker that is designed to reproduce very low bass frequencies', 'name': 'subwoofer'}, {'frequency': 'r', 'synset': 'sugar_bowl.n.01', 'synonyms': ['sugar_bowl'], 'id': 1031, 'def': 'a dish in which sugar is served', 'name': 'sugar_bowl'}, {'frequency': 'r', 'synset': 'sugarcane.n.01', 'synonyms': ['sugarcane_(plant)'], 'id': 1032, 'def': 'juicy canes whose sap is a source of molasses and commercial sugar; fresh canes are sometimes chewed for the juice', 'name': 'sugarcane_(plant)'}, {'frequency': 'f', 'synset': 'suit.n.01', 'synonyms': ['suit_(clothing)'], 'id': 1033, 'def': 'a set of garments (usually including a jacket and trousers or skirt) for outerwear all of the same fabric and color', 'name': 'suit_(clothing)'}, {'frequency': 'c', 'synset': 'sunflower.n.01', 'synonyms': ['sunflower'], 'id': 1034, 'def': 'any plant of the genus Helianthus having large flower heads with dark disk florets and showy yellow rays', 'name': 'sunflower'}, {'frequency': 'f', 'synset': 'sunglasses.n.01', 'synonyms': ['sunglasses'], 'id': 1035, 'def': 'spectacles that are darkened or polarized to protect the eyes from the glare of the sun', 'name': 'sunglasses'}, {'frequency': 'c', 'synset': 'sunhat.n.01', 'synonyms': ['sunhat'], 'id': 1036, 'def': 'a hat with a broad brim that protects the face from direct exposure to the sun', 'name': 'sunhat'}, {'frequency': 'f', 'synset': 'surfboard.n.01', 'synonyms': ['surfboard'], 'id': 1037, 'def': 'a narrow buoyant board for riding surf', 'name': 'surfboard'}, {'frequency': 'c', 'synset': 'sushi.n.01', 'synonyms': ['sushi'], 'id': 1038, 'def': 'rice (with raw fish) wrapped in seaweed', 'name': 'sushi'}, {'frequency': 'c', 'synset': 'swab.n.02', 'synonyms': ['mop'], 'id': 1039, 'def': 'cleaning implement consisting of absorbent material fastened to a handle; for cleaning floors', 'name': 'mop'}, {'frequency': 'c', 'synset': 'sweat_pants.n.01', 'synonyms': ['sweat_pants'], 'id': 1040, 'def': 'loose-fitting trousers with elastic cuffs; worn by athletes', 'name': 'sweat_pants'}, {'frequency': 'c', 'synset': 'sweatband.n.02', 'synonyms': ['sweatband'], 'id': 1041, 'def': 'a band of material tied around the forehead or wrist to absorb sweat', 'name': 'sweatband'}, {'frequency': 'f', 'synset': 'sweater.n.01', 'synonyms': ['sweater'], 'id': 1042, 'def': 'a crocheted or knitted garment covering the upper part of the body', 'name': 'sweater'}, {'frequency': 'f', 'synset': 'sweatshirt.n.01', 'synonyms': ['sweatshirt'], 'id': 1043, 'def': 'cotton knit pullover with long sleeves worn during athletic activity', 'name': 'sweatshirt'}, {'frequency': 'c', 'synset': 'sweet_potato.n.02', 'synonyms': ['sweet_potato'], 'id': 1044, 'def': 'the edible tuberous root of the sweet potato vine', 'name': 'sweet_potato'}, {'frequency': 'f', 'synset': 'swimsuit.n.01', 'synonyms': ['swimsuit', 'swimwear', 'bathing_suit', 'swimming_costume', 'bathing_costume', 'swimming_trunks', 'bathing_trunks'], 'id': 1045, 'def': 'garment worn for swimming', 'name': 'swimsuit'}, {'frequency': 'c', 'synset': 'sword.n.01', 'synonyms': ['sword'], 'id': 1046, 'def': 'a cutting or thrusting weapon that has a long metal blade', 'name': 'sword'}, {'frequency': 'r', 'synset': 'syringe.n.01', 'synonyms': ['syringe'], 'id': 1047, 'def': 'a medical instrument used to inject or withdraw fluids', 'name': 'syringe'}, {'frequency': 'r', 'synset': 'tabasco.n.02', 'synonyms': ['Tabasco_sauce'], 'id': 1048, 'def': 'very spicy sauce (trade name Tabasco) made from fully-aged red peppers', 'name': 'Tabasco_sauce'}, {'frequency': 'r', 'synset': 'table-tennis_table.n.01', 'synonyms': ['table-tennis_table', 'ping-pong_table'], 'id': 1049, 'def': 'a table used for playing table tennis', 'name': 'table-tennis_table'}, {'frequency': 'f', 'synset': 'table.n.02', 'synonyms': ['table'], 'id': 1050, 'def': 'a piece of furniture having a smooth flat top that is usually supported by one or more vertical legs', 'name': 'table'}, {'frequency': 'c', 'synset': 'table_lamp.n.01', 'synonyms': ['table_lamp'], 'id': 1051, 'def': 'a lamp that sits on a table', 'name': 'table_lamp'}, {'frequency': 'f', 'synset': 'tablecloth.n.01', 'synonyms': ['tablecloth'], 'id': 1052, 'def': 'a covering spread over a dining table', 'name': 'tablecloth'}, {'frequency': 'r', 'synset': 'tachometer.n.01', 'synonyms': ['tachometer'], 'id': 1053, 'def': 'measuring instrument for indicating speed of rotation', 'name': 'tachometer'}, {'frequency': 'r', 'synset': 'taco.n.02', 'synonyms': ['taco'], 'id': 1054, 'def': 'a small tortilla cupped around a filling', 'name': 'taco'}, {'frequency': 'f', 'synset': 'tag.n.02', 'synonyms': ['tag'], 'id': 1055, 'def': 'a label associated with something for the purpose of identification or information', 'name': 'tag'}, {'frequency': 'f', 'synset': 'taillight.n.01', 'synonyms': ['taillight', 'rear_light'], 'id': 1056, 'def': 'lamp (usually red) mounted at the rear of a motor vehicle', 'name': 'taillight'}, {'frequency': 'r', 'synset': 'tambourine.n.01', 'synonyms': ['tambourine'], 'id': 1057, 'def': 'a shallow drum with a single drumhead and with metallic disks in the sides', 'name': 'tambourine'}, {'frequency': 'r', 'synset': 'tank.n.01', 'synonyms': ['army_tank', 'armored_combat_vehicle', 'armoured_combat_vehicle'], 'id': 1058, 'def': 'an enclosed armored military vehicle; has a cannon and moves on caterpillar treads', 'name': 'army_tank'}, {'frequency': 'f', 'synset': 'tank.n.02', 'synonyms': ['tank_(storage_vessel)', 'storage_tank'], 'id': 1059, 'def': 'a large (usually metallic) vessel for holding gases or liquids', 'name': 'tank_(storage_vessel)'}, {'frequency': 'f', 'synset': 'tank_top.n.01', 'synonyms': ['tank_top_(clothing)'], 'id': 1060, 'def': 'a tight-fitting sleeveless shirt with wide shoulder straps and low neck and no front opening', 'name': 'tank_top_(clothing)'}, {'frequency': 'f', 'synset': 'tape.n.01', 'synonyms': ['tape_(sticky_cloth_or_paper)'], 'id': 1061, 'def': 'a long thin piece of cloth or paper as used for binding or fastening', 'name': 'tape_(sticky_cloth_or_paper)'}, {'frequency': 'c', 'synset': 'tape.n.04', 'synonyms': ['tape_measure', 'measuring_tape'], 'id': 1062, 'def': 'measuring instrument consisting of a narrow strip (cloth or metal) marked in inches or centimeters and used for measuring lengths', 'name': 'tape_measure'}, {'frequency': 'c', 'synset': 'tapestry.n.02', 'synonyms': ['tapestry'], 'id': 1063, 'def': 'a heavy textile with a woven design; used for curtains and upholstery', 'name': 'tapestry'}, {'frequency': 'f', 'synset': 'tarpaulin.n.01', 'synonyms': ['tarp'], 'id': 1064, 'def': 'waterproofed canvas', 'name': 'tarp'}, {'frequency': 'c', 'synset': 'tartan.n.01', 'synonyms': ['tartan', 'plaid'], 'id': 1065, 'def': 'a cloth having a crisscross design', 'name': 'tartan'}, {'frequency': 'c', 'synset': 'tassel.n.01', 'synonyms': ['tassel'], 'id': 1066, 'def': 'adornment consisting of a bunch of cords fastened at one end', 'name': 'tassel'}, {'frequency': 'c', 'synset': 'tea_bag.n.01', 'synonyms': ['tea_bag'], 'id': 1067, 'def': 'a measured amount of tea in a bag for an individual serving of tea', 'name': 'tea_bag'}, {'frequency': 'c', 'synset': 'teacup.n.02', 'synonyms': ['teacup'], 'id': 1068, 'def': 'a cup from which tea is drunk', 'name': 'teacup'}, {'frequency': 'c', 'synset': 'teakettle.n.01', 'synonyms': ['teakettle'], 'id': 1069, 'def': 'kettle for boiling water to make tea', 'name': 'teakettle'}, {'frequency': 'f', 'synset': 'teapot.n.01', 'synonyms': ['teapot'], 'id': 1070, 'def': 'pot for brewing tea; usually has a spout and handle', 'name': 'teapot'}, {'frequency': 'f', 'synset': 'teddy.n.01', 'synonyms': ['teddy_bear'], 'id': 1071, 'def': "plaything consisting of a child's toy bear (usually plush and stuffed with soft materials)", 'name': 'teddy_bear'}, {'frequency': 'f', 'synset': 'telephone.n.01', 'synonyms': ['telephone', 'phone', 'telephone_set'], 'id': 1072, 'def': 'electronic device for communicating by voice over long distances (includes wired and wireless/cell phones)', 'name': 'telephone'}, {'frequency': 'c', 'synset': 'telephone_booth.n.01', 'synonyms': ['telephone_booth', 'phone_booth', 'call_box', 'telephone_box', 'telephone_kiosk'], 'id': 1073, 'def': 'booth for using a telephone', 'name': 'telephone_booth'}, {'frequency': 'f', 'synset': 'telephone_pole.n.01', 'synonyms': ['telephone_pole', 'telegraph_pole', 'telegraph_post'], 'id': 1074, 'def': 'tall pole supporting telephone wires', 'name': 'telephone_pole'}, {'frequency': 'r', 'synset': 'telephoto_lens.n.01', 'synonyms': ['telephoto_lens', 'zoom_lens'], 'id': 1075, 'def': 'a camera lens that magnifies the image', 'name': 'telephoto_lens'}, {'frequency': 'c', 'synset': 'television_camera.n.01', 'synonyms': ['television_camera', 'tv_camera'], 'id': 1076, 'def': 'television equipment for capturing and recording video', 'name': 'television_camera'}, {'frequency': 'f', 'synset': 'television_receiver.n.01', 'synonyms': ['television_set', 'tv', 'tv_set'], 'id': 1077, 'def': 'an electronic device that receives television signals and displays them on a screen', 'name': 'television_set'}, {'frequency': 'f', 'synset': 'tennis_ball.n.01', 'synonyms': ['tennis_ball'], 'id': 1078, 'def': 'ball about the size of a fist used in playing tennis', 'name': 'tennis_ball'}, {'frequency': 'f', 'synset': 'tennis_racket.n.01', 'synonyms': ['tennis_racket'], 'id': 1079, 'def': 'a racket used to play tennis', 'name': 'tennis_racket'}, {'frequency': 'r', 'synset': 'tequila.n.01', 'synonyms': ['tequila'], 'id': 1080, 'def': 'Mexican liquor made from fermented juices of an agave plant', 'name': 'tequila'}, {'frequency': 'c', 'synset': 'thermometer.n.01', 'synonyms': ['thermometer'], 'id': 1081, 'def': 'measuring instrument for measuring temperature', 'name': 'thermometer'}, {'frequency': 'c', 'synset': 'thermos.n.01', 'synonyms': ['thermos_bottle'], 'id': 1082, 'def': 'vacuum flask that preserves temperature of hot or cold drinks', 'name': 'thermos_bottle'}, {'frequency': 'f', 'synset': 'thermostat.n.01', 'synonyms': ['thermostat'], 'id': 1083, 'def': 'a regulator for automatically regulating temperature by starting or stopping the supply of heat', 'name': 'thermostat'}, {'frequency': 'r', 'synset': 'thimble.n.02', 'synonyms': ['thimble'], 'id': 1084, 'def': 'a small metal cap to protect the finger while sewing; can be used as a small container', 'name': 'thimble'}, {'frequency': 'c', 'synset': 'thread.n.01', 'synonyms': ['thread', 'yarn'], 'id': 1085, 'def': 'a fine cord of twisted fibers (of cotton or silk or wool or nylon etc.) used in sewing and weaving', 'name': 'thread'}, {'frequency': 'c', 'synset': 'thumbtack.n.01', 'synonyms': ['thumbtack', 'drawing_pin', 'pushpin'], 'id': 1086, 'def': 'a tack for attaching papers to a bulletin board or drawing board', 'name': 'thumbtack'}, {'frequency': 'c', 'synset': 'tiara.n.01', 'synonyms': ['tiara'], 'id': 1087, 'def': 'a jeweled headdress worn by women on formal occasions', 'name': 'tiara'}, {'frequency': 'c', 'synset': 'tiger.n.02', 'synonyms': ['tiger'], 'id': 1088, 'def': 'large feline of forests in most of Asia having a tawny coat with black stripes', 'name': 'tiger'}, {'frequency': 'c', 'synset': 'tights.n.01', 'synonyms': ['tights_(clothing)', 'leotards'], 'id': 1089, 'def': 'skintight knit hose covering the body from the waist to the feet worn by acrobats and dancers and as stockings by women and girls', 'name': 'tights_(clothing)'}, {'frequency': 'c', 'synset': 'timer.n.01', 'synonyms': ['timer', 'stopwatch'], 'id': 1090, 'def': 'a timepiece that measures a time interval and signals its end', 'name': 'timer'}, {'frequency': 'f', 'synset': 'tinfoil.n.01', 'synonyms': ['tinfoil'], 'id': 1091, 'def': 'foil made of tin or an alloy of tin and lead', 'name': 'tinfoil'}, {'frequency': 'c', 'synset': 'tinsel.n.01', 'synonyms': ['tinsel'], 'id': 1092, 'def': 'a showy decoration that is basically valueless', 'name': 'tinsel'}, {'frequency': 'f', 'synset': 'tissue.n.02', 'synonyms': ['tissue_paper'], 'id': 1093, 'def': 'a soft thin (usually translucent) paper', 'name': 'tissue_paper'}, {'frequency': 'c', 'synset': 'toast.n.01', 'synonyms': ['toast_(food)'], 'id': 1094, 'def': 'slice of bread that has been toasted', 'name': 'toast_(food)'}, {'frequency': 'f', 'synset': 'toaster.n.02', 'synonyms': ['toaster'], 'id': 1095, 'def': 'a kitchen appliance (usually electric) for toasting bread', 'name': 'toaster'}, {'frequency': 'f', 'synset': 'toaster_oven.n.01', 'synonyms': ['toaster_oven'], 'id': 1096, 'def': 'kitchen appliance consisting of a small electric oven for toasting or warming food', 'name': 'toaster_oven'}, {'frequency': 'f', 'synset': 'toilet.n.02', 'synonyms': ['toilet'], 'id': 1097, 'def': 'a plumbing fixture for defecation and urination', 'name': 'toilet'}, {'frequency': 'f', 'synset': 'toilet_tissue.n.01', 'synonyms': ['toilet_tissue', 'toilet_paper', 'bathroom_tissue'], 'id': 1098, 'def': 'a soft thin absorbent paper for use in toilets', 'name': 'toilet_tissue'}, {'frequency': 'f', 'synset': 'tomato.n.01', 'synonyms': ['tomato'], 'id': 1099, 'def': 'mildly acid red or yellow pulpy fruit eaten as a vegetable', 'name': 'tomato'}, {'frequency': 'f', 'synset': 'tongs.n.01', 'synonyms': ['tongs'], 'id': 1100, 'def': 'any of various devices for taking hold of objects; usually have two hinged legs with handles above and pointed hooks below', 'name': 'tongs'}, {'frequency': 'c', 'synset': 'toolbox.n.01', 'synonyms': ['toolbox'], 'id': 1101, 'def': 'a box or chest or cabinet for holding hand tools', 'name': 'toolbox'}, {'frequency': 'f', 'synset': 'toothbrush.n.01', 'synonyms': ['toothbrush'], 'id': 1102, 'def': 'small brush; has long handle; used to clean teeth', 'name': 'toothbrush'}, {'frequency': 'f', 'synset': 'toothpaste.n.01', 'synonyms': ['toothpaste'], 'id': 1103, 'def': 'a dentifrice in the form of a paste', 'name': 'toothpaste'}, {'frequency': 'f', 'synset': 'toothpick.n.01', 'synonyms': ['toothpick'], 'id': 1104, 'def': 'pick consisting of a small strip of wood or plastic; used to pick food from between the teeth', 'name': 'toothpick'}, {'frequency': 'f', 'synset': 'top.n.09', 'synonyms': ['cover'], 'id': 1105, 'def': 'covering for a hole (especially a hole in the top of a container)', 'name': 'cover'}, {'frequency': 'c', 'synset': 'tortilla.n.01', 'synonyms': ['tortilla'], 'id': 1106, 'def': 'thin unleavened pancake made from cornmeal or wheat flour', 'name': 'tortilla'}, {'frequency': 'c', 'synset': 'tow_truck.n.01', 'synonyms': ['tow_truck'], 'id': 1107, 'def': 'a truck equipped to hoist and pull wrecked cars (or to remove cars from no-parking zones)', 'name': 'tow_truck'}, {'frequency': 'f', 'synset': 'towel.n.01', 'synonyms': ['towel'], 'id': 1108, 'def': 'a rectangular piece of absorbent cloth (or paper) for drying or wiping', 'name': 'towel'}, {'frequency': 'f', 'synset': 'towel_rack.n.01', 'synonyms': ['towel_rack', 'towel_rail', 'towel_bar'], 'id': 1109, 'def': 'a rack consisting of one or more bars on which towels can be hung', 'name': 'towel_rack'}, {'frequency': 'f', 'synset': 'toy.n.03', 'synonyms': ['toy'], 'id': 1110, 'def': 'a device regarded as providing amusement', 'name': 'toy'}, {'frequency': 'c', 'synset': 'tractor.n.01', 'synonyms': ['tractor_(farm_equipment)'], 'id': 1111, 'def': 'a wheeled vehicle with large wheels; used in farming and other applications', 'name': 'tractor_(farm_equipment)'}, {'frequency': 'f', 'synset': 'traffic_light.n.01', 'synonyms': ['traffic_light'], 'id': 1112, 'def': 'a device to control vehicle traffic often consisting of three or more lights', 'name': 'traffic_light'}, {'frequency': 'c', 'synset': 'trail_bike.n.01', 'synonyms': ['dirt_bike'], 'id': 1113, 'def': 'a lightweight motorcycle equipped with rugged tires and suspension for off-road use', 'name': 'dirt_bike'}, {'frequency': 'f', 'synset': 'trailer_truck.n.01', 'synonyms': ['trailer_truck', 'tractor_trailer', 'trucking_rig', 'articulated_lorry', 'semi_truck'], 'id': 1114, 'def': 'a truck consisting of a tractor and trailer together', 'name': 'trailer_truck'}, {'frequency': 'f', 'synset': 'train.n.01', 'synonyms': ['train_(railroad_vehicle)', 'railroad_train'], 'id': 1115, 'def': 'public or private transport provided by a line of railway cars coupled together and drawn by a locomotive', 'name': 'train_(railroad_vehicle)'}, {'frequency': 'r', 'synset': 'trampoline.n.01', 'synonyms': ['trampoline'], 'id': 1116, 'def': 'gymnastic apparatus consisting of a strong canvas sheet attached with springs to a metal frame', 'name': 'trampoline'}, {'frequency': 'f', 'synset': 'tray.n.01', 'synonyms': ['tray'], 'id': 1117, 'def': 'an open receptacle for holding or displaying or serving articles or food', 'name': 'tray'}, {'frequency': 'r', 'synset': 'trench_coat.n.01', 'synonyms': ['trench_coat'], 'id': 1118, 'def': 'a military style raincoat; belted with deep pockets', 'name': 'trench_coat'}, {'frequency': 'r', 'synset': 'triangle.n.05', 'synonyms': ['triangle_(musical_instrument)'], 'id': 1119, 'def': 'a percussion instrument consisting of a metal bar bent in the shape of an open triangle', 'name': 'triangle_(musical_instrument)'}, {'frequency': 'c', 'synset': 'tricycle.n.01', 'synonyms': ['tricycle'], 'id': 1120, 'def': 'a vehicle with three wheels that is moved by foot pedals', 'name': 'tricycle'}, {'frequency': 'f', 'synset': 'tripod.n.01', 'synonyms': ['tripod'], 'id': 1121, 'def': 'a three-legged rack used for support', 'name': 'tripod'}, {'frequency': 'f', 'synset': 'trouser.n.01', 'synonyms': ['trousers', 'pants_(clothing)'], 'id': 1122, 'def': 'a garment extending from the waist to the knee or ankle, covering each leg separately', 'name': 'trousers'}, {'frequency': 'f', 'synset': 'truck.n.01', 'synonyms': ['truck'], 'id': 1123, 'def': 'an automotive vehicle suitable for hauling', 'name': 'truck'}, {'frequency': 'r', 'synset': 'truffle.n.03', 'synonyms': ['truffle_(chocolate)', 'chocolate_truffle'], 'id': 1124, 'def': 'creamy chocolate candy', 'name': 'truffle_(chocolate)'}, {'frequency': 'c', 'synset': 'trunk.n.02', 'synonyms': ['trunk'], 'id': 1125, 'def': 'luggage consisting of a large strong case used when traveling or for storage', 'name': 'trunk'}, {'frequency': 'r', 'synset': 'tub.n.02', 'synonyms': ['vat'], 'id': 1126, 'def': 'a large vessel for holding or storing liquids', 'name': 'vat'}, {'frequency': 'c', 'synset': 'turban.n.01', 'synonyms': ['turban'], 'id': 1127, 'def': 'a traditional headdress consisting of a long scarf wrapped around the head', 'name': 'turban'}, {'frequency': 'c', 'synset': 'turkey.n.04', 'synonyms': ['turkey_(food)'], 'id': 1128, 'def': 'flesh of large domesticated fowl usually roasted', 'name': 'turkey_(food)'}, {'frequency': 'r', 'synset': 'turnip.n.01', 'synonyms': ['turnip'], 'id': 1129, 'def': 'widely cultivated plant having a large fleshy edible white or yellow root', 'name': 'turnip'}, {'frequency': 'c', 'synset': 'turtle.n.02', 'synonyms': ['turtle'], 'id': 1130, 'def': 'any of various aquatic and land reptiles having a bony shell and flipper-like limbs for swimming', 'name': 'turtle'}, {'frequency': 'c', 'synset': 'turtleneck.n.01', 'synonyms': ['turtleneck_(clothing)', 'polo-neck'], 'id': 1131, 'def': 'a sweater or jersey with a high close-fitting collar', 'name': 'turtleneck_(clothing)'}, {'frequency': 'c', 'synset': 'typewriter.n.01', 'synonyms': ['typewriter'], 'id': 1132, 'def': 'hand-operated character printer for printing written messages one character at a time', 'name': 'typewriter'}, {'frequency': 'f', 'synset': 'umbrella.n.01', 'synonyms': ['umbrella'], 'id': 1133, 'def': 'a lightweight handheld collapsible canopy', 'name': 'umbrella'}, {'frequency': 'f', 'synset': 'underwear.n.01', 'synonyms': ['underwear', 'underclothes', 'underclothing', 'underpants'], 'id': 1134, 'def': 'undergarment worn next to the skin and under the outer garments', 'name': 'underwear'}, {'frequency': 'r', 'synset': 'unicycle.n.01', 'synonyms': ['unicycle'], 'id': 1135, 'def': 'a vehicle with a single wheel that is driven by pedals', 'name': 'unicycle'}, {'frequency': 'f', 'synset': 'urinal.n.01', 'synonyms': ['urinal'], 'id': 1136, 'def': 'a plumbing fixture (usually attached to the wall) used by men to urinate', 'name': 'urinal'}, {'frequency': 'c', 'synset': 'urn.n.01', 'synonyms': ['urn'], 'id': 1137, 'def': 'a large vase that usually has a pedestal or feet', 'name': 'urn'}, {'frequency': 'c', 'synset': 'vacuum.n.04', 'synonyms': ['vacuum_cleaner'], 'id': 1138, 'def': 'an electrical home appliance that cleans by suction', 'name': 'vacuum_cleaner'}, {'frequency': 'f', 'synset': 'vase.n.01', 'synonyms': ['vase'], 'id': 1139, 'def': 'an open jar of glass or porcelain used as an ornament or to hold flowers', 'name': 'vase'}, {'frequency': 'c', 'synset': 'vending_machine.n.01', 'synonyms': ['vending_machine'], 'id': 1140, 'def': 'a slot machine for selling goods', 'name': 'vending_machine'}, {'frequency': 'f', 'synset': 'vent.n.01', 'synonyms': ['vent', 'blowhole', 'air_vent'], 'id': 1141, 'def': 'a hole for the escape of gas or air', 'name': 'vent'}, {'frequency': 'f', 'synset': 'vest.n.01', 'synonyms': ['vest', 'waistcoat'], 'id': 1142, 'def': "a man's sleeveless garment worn underneath a coat", 'name': 'vest'}, {'frequency': 'c', 'synset': 'videotape.n.01', 'synonyms': ['videotape'], 'id': 1143, 'def': 'a video recording made on magnetic tape', 'name': 'videotape'}, {'frequency': 'r', 'synset': 'vinegar.n.01', 'synonyms': ['vinegar'], 'id': 1144, 'def': 'sour-tasting liquid produced usually by oxidation of the alcohol in wine or cider and used as a condiment or food preservative', 'name': 'vinegar'}, {'frequency': 'r', 'synset': 'violin.n.01', 'synonyms': ['violin', 'fiddle'], 'id': 1145, 'def': 'bowed stringed instrument that is the highest member of the violin family', 'name': 'violin'}, {'frequency': 'r', 'synset': 'vodka.n.01', 'synonyms': ['vodka'], 'id': 1146, 'def': 'unaged colorless liquor originating in Russia', 'name': 'vodka'}, {'frequency': 'c', 'synset': 'volleyball.n.02', 'synonyms': ['volleyball'], 'id': 1147, 'def': 'an inflated ball used in playing volleyball', 'name': 'volleyball'}, {'frequency': 'r', 'synset': 'vulture.n.01', 'synonyms': ['vulture'], 'id': 1148, 'def': 'any of various large birds of prey having naked heads and weak claws and feeding chiefly on carrion', 'name': 'vulture'}, {'frequency': 'c', 'synset': 'waffle.n.01', 'synonyms': ['waffle'], 'id': 1149, 'def': 'pancake batter baked in a waffle iron', 'name': 'waffle'}, {'frequency': 'r', 'synset': 'waffle_iron.n.01', 'synonyms': ['waffle_iron'], 'id': 1150, 'def': 'a kitchen appliance for baking waffles', 'name': 'waffle_iron'}, {'frequency': 'c', 'synset': 'wagon.n.01', 'synonyms': ['wagon'], 'id': 1151, 'def': 'any of various kinds of wheeled vehicles drawn by an animal or a tractor', 'name': 'wagon'}, {'frequency': 'c', 'synset': 'wagon_wheel.n.01', 'synonyms': ['wagon_wheel'], 'id': 1152, 'def': 'a wheel of a wagon', 'name': 'wagon_wheel'}, {'frequency': 'c', 'synset': 'walking_stick.n.01', 'synonyms': ['walking_stick'], 'id': 1153, 'def': 'a stick carried in the hand for support in walking', 'name': 'walking_stick'}, {'frequency': 'c', 'synset': 'wall_clock.n.01', 'synonyms': ['wall_clock'], 'id': 1154, 'def': 'a clock mounted on a wall', 'name': 'wall_clock'}, {'frequency': 'f', 'synset': 'wall_socket.n.01', 'synonyms': ['wall_socket', 'wall_plug', 'electric_outlet', 'electrical_outlet', 'outlet', 'electric_receptacle'], 'id': 1155, 'def': 'receptacle providing a place in a wiring system where current can be taken to run electrical devices', 'name': 'wall_socket'}, {'frequency': 'f', 'synset': 'wallet.n.01', 'synonyms': ['wallet', 'billfold'], 'id': 1156, 'def': 'a pocket-size case for holding papers and paper money', 'name': 'wallet'}, {'frequency': 'r', 'synset': 'walrus.n.01', 'synonyms': ['walrus'], 'id': 1157, 'def': 'either of two large northern marine mammals having ivory tusks and tough hide over thick blubber', 'name': 'walrus'}, {'frequency': 'r', 'synset': 'wardrobe.n.01', 'synonyms': ['wardrobe'], 'id': 1158, 'def': 'a tall piece of furniture that provides storage space for clothes; has a door and rails or hooks for hanging clothes', 'name': 'wardrobe'}, {'frequency': 'r', 'synset': 'washbasin.n.01', 'synonyms': ['washbasin', 'basin_(for_washing)', 'washbowl', 'washstand', 'handbasin'], 'id': 1159, 'def': 'a bathroom sink that is permanently installed and connected to a water supply and drainpipe; where you can wash your hands and face', 'name': 'washbasin'}, {'frequency': 'c', 'synset': 'washer.n.03', 'synonyms': ['automatic_washer', 'washing_machine'], 'id': 1160, 'def': 'a home appliance for washing clothes and linens automatically', 'name': 'automatic_washer'}, {'frequency': 'f', 'synset': 'watch.n.01', 'synonyms': ['watch', 'wristwatch'], 'id': 1161, 'def': 'a small, portable timepiece', 'name': 'watch'}, {'frequency': 'f', 'synset': 'water_bottle.n.01', 'synonyms': ['water_bottle'], 'id': 1162, 'def': 'a bottle for holding water', 'name': 'water_bottle'}, {'frequency': 'c', 'synset': 'water_cooler.n.01', 'synonyms': ['water_cooler'], 'id': 1163, 'def': 'a device for cooling and dispensing drinking water', 'name': 'water_cooler'}, {'frequency': 'c', 'synset': 'water_faucet.n.01', 'synonyms': ['water_faucet', 'water_tap', 'tap_(water_faucet)'], 'id': 1164, 'def': 'a faucet for drawing water from a pipe or cask', 'name': 'water_faucet'}, {'frequency': 'r', 'synset': 'water_heater.n.01', 'synonyms': ['water_heater', 'hot-water_heater'], 'id': 1165, 'def': 'a heater and storage tank to supply heated water', 'name': 'water_heater'}, {'frequency': 'c', 'synset': 'water_jug.n.01', 'synonyms': ['water_jug'], 'id': 1166, 'def': 'a jug that holds water', 'name': 'water_jug'}, {'frequency': 'r', 'synset': 'water_pistol.n.01', 'synonyms': ['water_gun', 'squirt_gun'], 'id': 1167, 'def': 'plaything consisting of a toy pistol that squirts water', 'name': 'water_gun'}, {'frequency': 'c', 'synset': 'water_scooter.n.01', 'synonyms': ['water_scooter', 'sea_scooter', 'jet_ski'], 'id': 1168, 'def': 'a motorboat resembling a motor scooter (NOT A SURFBOARD OR WATER SKI)', 'name': 'water_scooter'}, {'frequency': 'c', 'synset': 'water_ski.n.01', 'synonyms': ['water_ski'], 'id': 1169, 'def': 'broad ski for skimming over water towed by a speedboat (DO NOT MARK WATER)', 'name': 'water_ski'}, {'frequency': 'c', 'synset': 'water_tower.n.01', 'synonyms': ['water_tower'], 'id': 1170, 'def': 'a large reservoir for water', 'name': 'water_tower'}, {'frequency': 'c', 'synset': 'watering_can.n.01', 'synonyms': ['watering_can'], 'id': 1171, 'def': 'a container with a handle and a spout with a perforated nozzle; used to sprinkle water over plants', 'name': 'watering_can'}, {'frequency': 'f', 'synset': 'watermelon.n.02', 'synonyms': ['watermelon'], 'id': 1172, 'def': 'large oblong or roundish melon with a hard green rind and sweet watery red or occasionally yellowish pulp', 'name': 'watermelon'}, {'frequency': 'f', 'synset': 'weathervane.n.01', 'synonyms': ['weathervane', 'vane_(weathervane)', 'wind_vane'], 'id': 1173, 'def': 'mechanical device attached to an elevated structure; rotates freely to show the direction of the wind', 'name': 'weathervane'}, {'frequency': 'c', 'synset': 'webcam.n.01', 'synonyms': ['webcam'], 'id': 1174, 'def': 'a digital camera designed to take digital photographs and transmit them over the internet', 'name': 'webcam'}, {'frequency': 'c', 'synset': 'wedding_cake.n.01', 'synonyms': ['wedding_cake', 'bridecake'], 'id': 1175, 'def': 'a rich cake with two or more tiers and covered with frosting and decorations; served at a wedding reception', 'name': 'wedding_cake'}, {'frequency': 'c', 'synset': 'wedding_ring.n.01', 'synonyms': ['wedding_ring', 'wedding_band'], 'id': 1176, 'def': 'a ring given to the bride and/or groom at the wedding', 'name': 'wedding_ring'}, {'frequency': 'f', 'synset': 'wet_suit.n.01', 'synonyms': ['wet_suit'], 'id': 1177, 'def': 'a close-fitting garment made of a permeable material; worn in cold water to retain body heat', 'name': 'wet_suit'}, {'frequency': 'f', 'synset': 'wheel.n.01', 'synonyms': ['wheel'], 'id': 1178, 'def': 'a circular frame with spokes (or a solid disc) that can rotate on a shaft or axle', 'name': 'wheel'}, {'frequency': 'c', 'synset': 'wheelchair.n.01', 'synonyms': ['wheelchair'], 'id': 1179, 'def': 'a movable chair mounted on large wheels', 'name': 'wheelchair'}, {'frequency': 'c', 'synset': 'whipped_cream.n.01', 'synonyms': ['whipped_cream'], 'id': 1180, 'def': 'cream that has been beaten until light and fluffy', 'name': 'whipped_cream'}, {'frequency': 'c', 'synset': 'whistle.n.03', 'synonyms': ['whistle'], 'id': 1181, 'def': 'a small wind instrument that produces a whistling sound by blowing into it', 'name': 'whistle'}, {'frequency': 'c', 'synset': 'wig.n.01', 'synonyms': ['wig'], 'id': 1182, 'def': 'hairpiece covering the head and made of real or synthetic hair', 'name': 'wig'}, {'frequency': 'c', 'synset': 'wind_chime.n.01', 'synonyms': ['wind_chime'], 'id': 1183, 'def': 'a decorative arrangement of pieces of metal or glass or pottery that hang together loosely so the wind can cause them to tinkle', 'name': 'wind_chime'}, {'frequency': 'c', 'synset': 'windmill.n.01', 'synonyms': ['windmill'], 'id': 1184, 'def': 'A mill or turbine that is powered by wind', 'name': 'windmill'}, {'frequency': 'c', 'synset': 'window_box.n.01', 'synonyms': ['window_box_(for_plants)'], 'id': 1185, 'def': 'a container for growing plants on a windowsill', 'name': 'window_box_(for_plants)'}, {'frequency': 'f', 'synset': 'windshield_wiper.n.01', 'synonyms': ['windshield_wiper', 'windscreen_wiper', 'wiper_(for_windshield/screen)'], 'id': 1186, 'def': 'a mechanical device that cleans the windshield', 'name': 'windshield_wiper'}, {'frequency': 'c', 'synset': 'windsock.n.01', 'synonyms': ['windsock', 'air_sock', 'air-sleeve', 'wind_sleeve', 'wind_cone'], 'id': 1187, 'def': 'a truncated cloth cone mounted on a mast/pole; shows wind direction', 'name': 'windsock'}, {'frequency': 'f', 'synset': 'wine_bottle.n.01', 'synonyms': ['wine_bottle'], 'id': 1188, 'def': 'a bottle for holding wine', 'name': 'wine_bottle'}, {'frequency': 'c', 'synset': 'wine_bucket.n.01', 'synonyms': ['wine_bucket', 'wine_cooler'], 'id': 1189, 'def': 'a bucket of ice used to chill a bottle of wine', 'name': 'wine_bucket'}, {'frequency': 'f', 'synset': 'wineglass.n.01', 'synonyms': ['wineglass'], 'id': 1190, 'def': 'a glass that has a stem and in which wine is served', 'name': 'wineglass'}, {'frequency': 'f', 'synset': 'winker.n.02', 'synonyms': ['blinder_(for_horses)'], 'id': 1191, 'def': 'blinds that prevent a horse from seeing something on either side', 'name': 'blinder_(for_horses)'}, {'frequency': 'c', 'synset': 'wok.n.01', 'synonyms': ['wok'], 'id': 1192, 'def': 'pan with a convex bottom; used for frying in Chinese cooking', 'name': 'wok'}, {'frequency': 'r', 'synset': 'wolf.n.01', 'synonyms': ['wolf'], 'id': 1193, 'def': 'a wild carnivorous mammal of the dog family, living and hunting in packs', 'name': 'wolf'}, {'frequency': 'c', 'synset': 'wooden_spoon.n.02', 'synonyms': ['wooden_spoon'], 'id': 1194, 'def': 'a spoon made of wood', 'name': 'wooden_spoon'}, {'frequency': 'c', 'synset': 'wreath.n.01', 'synonyms': ['wreath'], 'id': 1195, 'def': 'an arrangement of flowers, leaves, or stems fastened in a ring', 'name': 'wreath'}, {'frequency': 'c', 'synset': 'wrench.n.03', 'synonyms': ['wrench', 'spanner'], 'id': 1196, 'def': 'a hand tool that is used to hold or twist a nut or bolt', 'name': 'wrench'}, {'frequency': 'f', 'synset': 'wristband.n.01', 'synonyms': ['wristband'], 'id': 1197, 'def': 'band consisting of a part of a sleeve that covers the wrist', 'name': 'wristband'}, {'frequency': 'f', 'synset': 'wristlet.n.01', 'synonyms': ['wristlet', 'wrist_band'], 'id': 1198, 'def': 'a band or bracelet worn around the wrist', 'name': 'wristlet'}, {'frequency': 'c', 'synset': 'yacht.n.01', 'synonyms': ['yacht'], 'id': 1199, 'def': 'an expensive vessel propelled by sail or power and used for cruising or racing', 'name': 'yacht'}, {'frequency': 'c', 'synset': 'yogurt.n.01', 'synonyms': ['yogurt', 'yoghurt', 'yoghourt'], 'id': 1200, 'def': 'a custard-like food made from curdled milk', 'name': 'yogurt'}, {'frequency': 'c', 'synset': 'yoke.n.07', 'synonyms': ['yoke_(animal_equipment)'], 'id': 1201, 'def': 'gear joining two animals at the neck; NOT egg yolk', 'name': 'yoke_(animal_equipment)'}, {'frequency': 'f', 'synset': 'zebra.n.01', 'synonyms': ['zebra'], 'id': 1202, 'def': 'any of several fleet black-and-white striped African equines', 'name': 'zebra'}, {'frequency': 'c', 'synset': 'zucchini.n.02', 'synonyms': ['zucchini', 'courgette'], 'id': 1203, 'def': 'small cucumber-shaped vegetable marrow; typically dark green', 'name': 'zucchini'}] # noqa
+# fmt: on
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/lvis_v1_category_image_count.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/lvis_v1_category_image_count.py
new file mode 100644
index 0000000000000000000000000000000000000000..31bf0cfcd5096ab87835db86a28671d474514c40
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/lvis_v1_category_image_count.py
@@ -0,0 +1,20 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+# Autogen with
+# with open("lvis_v1_train.json", "r") as f:
+# a = json.load(f)
+# c = a["categories"]
+# for x in c:
+# del x["name"]
+# del x["instance_count"]
+# del x["def"]
+# del x["synonyms"]
+# del x["frequency"]
+# del x["synset"]
+# LVIS_CATEGORY_IMAGE_COUNT = repr(c) + " # noqa"
+# with open("/tmp/lvis_category_image_count.py", "wt") as f:
+# f.write(f"LVIS_CATEGORY_IMAGE_COUNT = {LVIS_CATEGORY_IMAGE_COUNT}")
+# Then paste the contents of that file below
+
+# fmt: off
+LVIS_CATEGORY_IMAGE_COUNT = [{'id': 1, 'image_count': 64}, {'id': 2, 'image_count': 364}, {'id': 3, 'image_count': 1911}, {'id': 4, 'image_count': 149}, {'id': 5, 'image_count': 29}, {'id': 6, 'image_count': 26}, {'id': 7, 'image_count': 59}, {'id': 8, 'image_count': 22}, {'id': 9, 'image_count': 12}, {'id': 10, 'image_count': 28}, {'id': 11, 'image_count': 505}, {'id': 12, 'image_count': 1207}, {'id': 13, 'image_count': 4}, {'id': 14, 'image_count': 10}, {'id': 15, 'image_count': 500}, {'id': 16, 'image_count': 33}, {'id': 17, 'image_count': 3}, {'id': 18, 'image_count': 44}, {'id': 19, 'image_count': 561}, {'id': 20, 'image_count': 8}, {'id': 21, 'image_count': 9}, {'id': 22, 'image_count': 33}, {'id': 23, 'image_count': 1883}, {'id': 24, 'image_count': 98}, {'id': 25, 'image_count': 70}, {'id': 26, 'image_count': 46}, {'id': 27, 'image_count': 117}, {'id': 28, 'image_count': 41}, {'id': 29, 'image_count': 1395}, {'id': 30, 'image_count': 7}, {'id': 31, 'image_count': 1}, {'id': 32, 'image_count': 314}, {'id': 33, 'image_count': 31}, {'id': 34, 'image_count': 1905}, {'id': 35, 'image_count': 1859}, {'id': 36, 'image_count': 1623}, {'id': 37, 'image_count': 47}, {'id': 38, 'image_count': 3}, {'id': 39, 'image_count': 3}, {'id': 40, 'image_count': 1}, {'id': 41, 'image_count': 305}, {'id': 42, 'image_count': 6}, {'id': 43, 'image_count': 210}, {'id': 44, 'image_count': 36}, {'id': 45, 'image_count': 1787}, {'id': 46, 'image_count': 17}, {'id': 47, 'image_count': 51}, {'id': 48, 'image_count': 138}, {'id': 49, 'image_count': 3}, {'id': 50, 'image_count': 1470}, {'id': 51, 'image_count': 3}, {'id': 52, 'image_count': 2}, {'id': 53, 'image_count': 186}, {'id': 54, 'image_count': 76}, {'id': 55, 'image_count': 26}, {'id': 56, 'image_count': 303}, {'id': 57, 'image_count': 738}, {'id': 58, 'image_count': 1799}, {'id': 59, 'image_count': 1934}, {'id': 60, 'image_count': 1609}, {'id': 61, 'image_count': 1622}, {'id': 62, 'image_count': 41}, {'id': 63, 'image_count': 4}, {'id': 64, 'image_count': 11}, {'id': 65, 'image_count': 270}, {'id': 66, 'image_count': 349}, {'id': 67, 'image_count': 42}, {'id': 68, 'image_count': 823}, {'id': 69, 'image_count': 6}, {'id': 70, 'image_count': 48}, {'id': 71, 'image_count': 3}, {'id': 72, 'image_count': 42}, {'id': 73, 'image_count': 24}, {'id': 74, 'image_count': 16}, {'id': 75, 'image_count': 605}, {'id': 76, 'image_count': 646}, {'id': 77, 'image_count': 1765}, {'id': 78, 'image_count': 2}, {'id': 79, 'image_count': 125}, {'id': 80, 'image_count': 1420}, {'id': 81, 'image_count': 140}, {'id': 82, 'image_count': 4}, {'id': 83, 'image_count': 322}, {'id': 84, 'image_count': 60}, {'id': 85, 'image_count': 2}, {'id': 86, 'image_count': 231}, {'id': 87, 'image_count': 333}, {'id': 88, 'image_count': 1941}, {'id': 89, 'image_count': 367}, {'id': 90, 'image_count': 1922}, {'id': 91, 'image_count': 18}, {'id': 92, 'image_count': 81}, {'id': 93, 'image_count': 1}, {'id': 94, 'image_count': 1852}, {'id': 95, 'image_count': 430}, {'id': 96, 'image_count': 247}, {'id': 97, 'image_count': 94}, {'id': 98, 'image_count': 21}, {'id': 99, 'image_count': 1821}, {'id': 100, 'image_count': 16}, {'id': 101, 'image_count': 12}, {'id': 102, 'image_count': 25}, {'id': 103, 'image_count': 41}, {'id': 104, 'image_count': 244}, {'id': 105, 'image_count': 7}, {'id': 106, 'image_count': 1}, {'id': 107, 'image_count': 40}, {'id': 108, 'image_count': 40}, {'id': 109, 'image_count': 104}, {'id': 110, 'image_count': 1671}, {'id': 111, 'image_count': 49}, {'id': 112, 'image_count': 243}, {'id': 113, 'image_count': 2}, {'id': 114, 'image_count': 242}, {'id': 115, 'image_count': 271}, {'id': 116, 'image_count': 104}, {'id': 117, 'image_count': 8}, {'id': 118, 'image_count': 1758}, {'id': 119, 'image_count': 1}, {'id': 120, 'image_count': 48}, {'id': 121, 'image_count': 14}, {'id': 122, 'image_count': 40}, {'id': 123, 'image_count': 1}, {'id': 124, 'image_count': 37}, {'id': 125, 'image_count': 1510}, {'id': 126, 'image_count': 6}, {'id': 127, 'image_count': 1903}, {'id': 128, 'image_count': 70}, {'id': 129, 'image_count': 86}, {'id': 130, 'image_count': 7}, {'id': 131, 'image_count': 5}, {'id': 132, 'image_count': 1406}, {'id': 133, 'image_count': 1901}, {'id': 134, 'image_count': 15}, {'id': 135, 'image_count': 28}, {'id': 136, 'image_count': 6}, {'id': 137, 'image_count': 494}, {'id': 138, 'image_count': 234}, {'id': 139, 'image_count': 1922}, {'id': 140, 'image_count': 1}, {'id': 141, 'image_count': 35}, {'id': 142, 'image_count': 5}, {'id': 143, 'image_count': 1828}, {'id': 144, 'image_count': 8}, {'id': 145, 'image_count': 63}, {'id': 146, 'image_count': 1668}, {'id': 147, 'image_count': 4}, {'id': 148, 'image_count': 95}, {'id': 149, 'image_count': 17}, {'id': 150, 'image_count': 1567}, {'id': 151, 'image_count': 2}, {'id': 152, 'image_count': 103}, {'id': 153, 'image_count': 50}, {'id': 154, 'image_count': 1309}, {'id': 155, 'image_count': 6}, {'id': 156, 'image_count': 92}, {'id': 157, 'image_count': 19}, {'id': 158, 'image_count': 37}, {'id': 159, 'image_count': 4}, {'id': 160, 'image_count': 709}, {'id': 161, 'image_count': 9}, {'id': 162, 'image_count': 82}, {'id': 163, 'image_count': 15}, {'id': 164, 'image_count': 3}, {'id': 165, 'image_count': 61}, {'id': 166, 'image_count': 51}, {'id': 167, 'image_count': 5}, {'id': 168, 'image_count': 13}, {'id': 169, 'image_count': 642}, {'id': 170, 'image_count': 24}, {'id': 171, 'image_count': 255}, {'id': 172, 'image_count': 9}, {'id': 173, 'image_count': 1808}, {'id': 174, 'image_count': 31}, {'id': 175, 'image_count': 158}, {'id': 176, 'image_count': 80}, {'id': 177, 'image_count': 1884}, {'id': 178, 'image_count': 158}, {'id': 179, 'image_count': 2}, {'id': 180, 'image_count': 12}, {'id': 181, 'image_count': 1659}, {'id': 182, 'image_count': 7}, {'id': 183, 'image_count': 834}, {'id': 184, 'image_count': 57}, {'id': 185, 'image_count': 174}, {'id': 186, 'image_count': 95}, {'id': 187, 'image_count': 27}, {'id': 188, 'image_count': 22}, {'id': 189, 'image_count': 1391}, {'id': 190, 'image_count': 90}, {'id': 191, 'image_count': 40}, {'id': 192, 'image_count': 445}, {'id': 193, 'image_count': 21}, {'id': 194, 'image_count': 1132}, {'id': 195, 'image_count': 177}, {'id': 196, 'image_count': 4}, {'id': 197, 'image_count': 17}, {'id': 198, 'image_count': 84}, {'id': 199, 'image_count': 55}, {'id': 200, 'image_count': 30}, {'id': 201, 'image_count': 25}, {'id': 202, 'image_count': 2}, {'id': 203, 'image_count': 125}, {'id': 204, 'image_count': 1135}, {'id': 205, 'image_count': 19}, {'id': 206, 'image_count': 72}, {'id': 207, 'image_count': 1926}, {'id': 208, 'image_count': 159}, {'id': 209, 'image_count': 7}, {'id': 210, 'image_count': 1}, {'id': 211, 'image_count': 13}, {'id': 212, 'image_count': 35}, {'id': 213, 'image_count': 18}, {'id': 214, 'image_count': 8}, {'id': 215, 'image_count': 6}, {'id': 216, 'image_count': 35}, {'id': 217, 'image_count': 1222}, {'id': 218, 'image_count': 103}, {'id': 219, 'image_count': 28}, {'id': 220, 'image_count': 63}, {'id': 221, 'image_count': 28}, {'id': 222, 'image_count': 5}, {'id': 223, 'image_count': 7}, {'id': 224, 'image_count': 14}, {'id': 225, 'image_count': 1918}, {'id': 226, 'image_count': 133}, {'id': 227, 'image_count': 16}, {'id': 228, 'image_count': 27}, {'id': 229, 'image_count': 110}, {'id': 230, 'image_count': 1895}, {'id': 231, 'image_count': 4}, {'id': 232, 'image_count': 1927}, {'id': 233, 'image_count': 8}, {'id': 234, 'image_count': 1}, {'id': 235, 'image_count': 263}, {'id': 236, 'image_count': 10}, {'id': 237, 'image_count': 2}, {'id': 238, 'image_count': 3}, {'id': 239, 'image_count': 87}, {'id': 240, 'image_count': 9}, {'id': 241, 'image_count': 71}, {'id': 242, 'image_count': 13}, {'id': 243, 'image_count': 18}, {'id': 244, 'image_count': 2}, {'id': 245, 'image_count': 5}, {'id': 246, 'image_count': 45}, {'id': 247, 'image_count': 1}, {'id': 248, 'image_count': 23}, {'id': 249, 'image_count': 32}, {'id': 250, 'image_count': 4}, {'id': 251, 'image_count': 1}, {'id': 252, 'image_count': 858}, {'id': 253, 'image_count': 661}, {'id': 254, 'image_count': 168}, {'id': 255, 'image_count': 210}, {'id': 256, 'image_count': 65}, {'id': 257, 'image_count': 4}, {'id': 258, 'image_count': 2}, {'id': 259, 'image_count': 159}, {'id': 260, 'image_count': 31}, {'id': 261, 'image_count': 811}, {'id': 262, 'image_count': 1}, {'id': 263, 'image_count': 42}, {'id': 264, 'image_count': 27}, {'id': 265, 'image_count': 2}, {'id': 266, 'image_count': 5}, {'id': 267, 'image_count': 95}, {'id': 268, 'image_count': 32}, {'id': 269, 'image_count': 1}, {'id': 270, 'image_count': 1}, {'id': 271, 'image_count': 1844}, {'id': 272, 'image_count': 897}, {'id': 273, 'image_count': 31}, {'id': 274, 'image_count': 23}, {'id': 275, 'image_count': 1}, {'id': 276, 'image_count': 202}, {'id': 277, 'image_count': 746}, {'id': 278, 'image_count': 44}, {'id': 279, 'image_count': 14}, {'id': 280, 'image_count': 26}, {'id': 281, 'image_count': 1}, {'id': 282, 'image_count': 2}, {'id': 283, 'image_count': 25}, {'id': 284, 'image_count': 238}, {'id': 285, 'image_count': 592}, {'id': 286, 'image_count': 26}, {'id': 287, 'image_count': 5}, {'id': 288, 'image_count': 42}, {'id': 289, 'image_count': 13}, {'id': 290, 'image_count': 46}, {'id': 291, 'image_count': 1}, {'id': 292, 'image_count': 8}, {'id': 293, 'image_count': 34}, {'id': 294, 'image_count': 5}, {'id': 295, 'image_count': 1}, {'id': 296, 'image_count': 1871}, {'id': 297, 'image_count': 717}, {'id': 298, 'image_count': 1010}, {'id': 299, 'image_count': 679}, {'id': 300, 'image_count': 3}, {'id': 301, 'image_count': 4}, {'id': 302, 'image_count': 1}, {'id': 303, 'image_count': 166}, {'id': 304, 'image_count': 2}, {'id': 305, 'image_count': 266}, {'id': 306, 'image_count': 101}, {'id': 307, 'image_count': 6}, {'id': 308, 'image_count': 14}, {'id': 309, 'image_count': 133}, {'id': 310, 'image_count': 2}, {'id': 311, 'image_count': 38}, {'id': 312, 'image_count': 95}, {'id': 313, 'image_count': 1}, {'id': 314, 'image_count': 12}, {'id': 315, 'image_count': 49}, {'id': 316, 'image_count': 5}, {'id': 317, 'image_count': 5}, {'id': 318, 'image_count': 16}, {'id': 319, 'image_count': 216}, {'id': 320, 'image_count': 12}, {'id': 321, 'image_count': 1}, {'id': 322, 'image_count': 54}, {'id': 323, 'image_count': 5}, {'id': 324, 'image_count': 245}, {'id': 325, 'image_count': 12}, {'id': 326, 'image_count': 7}, {'id': 327, 'image_count': 35}, {'id': 328, 'image_count': 36}, {'id': 329, 'image_count': 32}, {'id': 330, 'image_count': 1027}, {'id': 331, 'image_count': 10}, {'id': 332, 'image_count': 12}, {'id': 333, 'image_count': 1}, {'id': 334, 'image_count': 67}, {'id': 335, 'image_count': 71}, {'id': 336, 'image_count': 30}, {'id': 337, 'image_count': 48}, {'id': 338, 'image_count': 249}, {'id': 339, 'image_count': 13}, {'id': 340, 'image_count': 29}, {'id': 341, 'image_count': 14}, {'id': 342, 'image_count': 236}, {'id': 343, 'image_count': 15}, {'id': 344, 'image_count': 1521}, {'id': 345, 'image_count': 25}, {'id': 346, 'image_count': 249}, {'id': 347, 'image_count': 139}, {'id': 348, 'image_count': 2}, {'id': 349, 'image_count': 2}, {'id': 350, 'image_count': 1890}, {'id': 351, 'image_count': 1240}, {'id': 352, 'image_count': 1}, {'id': 353, 'image_count': 9}, {'id': 354, 'image_count': 1}, {'id': 355, 'image_count': 3}, {'id': 356, 'image_count': 11}, {'id': 357, 'image_count': 4}, {'id': 358, 'image_count': 236}, {'id': 359, 'image_count': 44}, {'id': 360, 'image_count': 19}, {'id': 361, 'image_count': 1100}, {'id': 362, 'image_count': 7}, {'id': 363, 'image_count': 69}, {'id': 364, 'image_count': 2}, {'id': 365, 'image_count': 8}, {'id': 366, 'image_count': 5}, {'id': 367, 'image_count': 227}, {'id': 368, 'image_count': 6}, {'id': 369, 'image_count': 106}, {'id': 370, 'image_count': 81}, {'id': 371, 'image_count': 17}, {'id': 372, 'image_count': 134}, {'id': 373, 'image_count': 312}, {'id': 374, 'image_count': 8}, {'id': 375, 'image_count': 271}, {'id': 376, 'image_count': 2}, {'id': 377, 'image_count': 103}, {'id': 378, 'image_count': 1938}, {'id': 379, 'image_count': 574}, {'id': 380, 'image_count': 120}, {'id': 381, 'image_count': 2}, {'id': 382, 'image_count': 2}, {'id': 383, 'image_count': 13}, {'id': 384, 'image_count': 29}, {'id': 385, 'image_count': 1710}, {'id': 386, 'image_count': 66}, {'id': 387, 'image_count': 1008}, {'id': 388, 'image_count': 1}, {'id': 389, 'image_count': 3}, {'id': 390, 'image_count': 1942}, {'id': 391, 'image_count': 19}, {'id': 392, 'image_count': 1488}, {'id': 393, 'image_count': 46}, {'id': 394, 'image_count': 106}, {'id': 395, 'image_count': 115}, {'id': 396, 'image_count': 19}, {'id': 397, 'image_count': 2}, {'id': 398, 'image_count': 1}, {'id': 399, 'image_count': 28}, {'id': 400, 'image_count': 9}, {'id': 401, 'image_count': 192}, {'id': 402, 'image_count': 12}, {'id': 403, 'image_count': 21}, {'id': 404, 'image_count': 247}, {'id': 405, 'image_count': 6}, {'id': 406, 'image_count': 64}, {'id': 407, 'image_count': 7}, {'id': 408, 'image_count': 40}, {'id': 409, 'image_count': 542}, {'id': 410, 'image_count': 2}, {'id': 411, 'image_count': 1898}, {'id': 412, 'image_count': 36}, {'id': 413, 'image_count': 4}, {'id': 414, 'image_count': 1}, {'id': 415, 'image_count': 191}, {'id': 416, 'image_count': 6}, {'id': 417, 'image_count': 41}, {'id': 418, 'image_count': 39}, {'id': 419, 'image_count': 46}, {'id': 420, 'image_count': 1}, {'id': 421, 'image_count': 1451}, {'id': 422, 'image_count': 1878}, {'id': 423, 'image_count': 11}, {'id': 424, 'image_count': 82}, {'id': 425, 'image_count': 18}, {'id': 426, 'image_count': 1}, {'id': 427, 'image_count': 7}, {'id': 428, 'image_count': 3}, {'id': 429, 'image_count': 575}, {'id': 430, 'image_count': 1907}, {'id': 431, 'image_count': 8}, {'id': 432, 'image_count': 4}, {'id': 433, 'image_count': 32}, {'id': 434, 'image_count': 11}, {'id': 435, 'image_count': 4}, {'id': 436, 'image_count': 54}, {'id': 437, 'image_count': 202}, {'id': 438, 'image_count': 32}, {'id': 439, 'image_count': 3}, {'id': 440, 'image_count': 130}, {'id': 441, 'image_count': 119}, {'id': 442, 'image_count': 141}, {'id': 443, 'image_count': 29}, {'id': 444, 'image_count': 525}, {'id': 445, 'image_count': 1323}, {'id': 446, 'image_count': 2}, {'id': 447, 'image_count': 113}, {'id': 448, 'image_count': 16}, {'id': 449, 'image_count': 7}, {'id': 450, 'image_count': 35}, {'id': 451, 'image_count': 1908}, {'id': 452, 'image_count': 353}, {'id': 453, 'image_count': 18}, {'id': 454, 'image_count': 14}, {'id': 455, 'image_count': 77}, {'id': 456, 'image_count': 8}, {'id': 457, 'image_count': 37}, {'id': 458, 'image_count': 1}, {'id': 459, 'image_count': 346}, {'id': 460, 'image_count': 19}, {'id': 461, 'image_count': 1779}, {'id': 462, 'image_count': 23}, {'id': 463, 'image_count': 25}, {'id': 464, 'image_count': 67}, {'id': 465, 'image_count': 19}, {'id': 466, 'image_count': 28}, {'id': 467, 'image_count': 4}, {'id': 468, 'image_count': 27}, {'id': 469, 'image_count': 1861}, {'id': 470, 'image_count': 11}, {'id': 471, 'image_count': 13}, {'id': 472, 'image_count': 13}, {'id': 473, 'image_count': 32}, {'id': 474, 'image_count': 1767}, {'id': 475, 'image_count': 42}, {'id': 476, 'image_count': 17}, {'id': 477, 'image_count': 128}, {'id': 478, 'image_count': 1}, {'id': 479, 'image_count': 9}, {'id': 480, 'image_count': 10}, {'id': 481, 'image_count': 4}, {'id': 482, 'image_count': 9}, {'id': 483, 'image_count': 18}, {'id': 484, 'image_count': 41}, {'id': 485, 'image_count': 28}, {'id': 486, 'image_count': 3}, {'id': 487, 'image_count': 65}, {'id': 488, 'image_count': 9}, {'id': 489, 'image_count': 23}, {'id': 490, 'image_count': 24}, {'id': 491, 'image_count': 1}, {'id': 492, 'image_count': 2}, {'id': 493, 'image_count': 59}, {'id': 494, 'image_count': 48}, {'id': 495, 'image_count': 17}, {'id': 496, 'image_count': 1877}, {'id': 497, 'image_count': 18}, {'id': 498, 'image_count': 1920}, {'id': 499, 'image_count': 50}, {'id': 500, 'image_count': 1890}, {'id': 501, 'image_count': 99}, {'id': 502, 'image_count': 1530}, {'id': 503, 'image_count': 3}, {'id': 504, 'image_count': 11}, {'id': 505, 'image_count': 19}, {'id': 506, 'image_count': 3}, {'id': 507, 'image_count': 63}, {'id': 508, 'image_count': 5}, {'id': 509, 'image_count': 6}, {'id': 510, 'image_count': 233}, {'id': 511, 'image_count': 54}, {'id': 512, 'image_count': 36}, {'id': 513, 'image_count': 10}, {'id': 514, 'image_count': 124}, {'id': 515, 'image_count': 101}, {'id': 516, 'image_count': 3}, {'id': 517, 'image_count': 363}, {'id': 518, 'image_count': 3}, {'id': 519, 'image_count': 30}, {'id': 520, 'image_count': 18}, {'id': 521, 'image_count': 199}, {'id': 522, 'image_count': 97}, {'id': 523, 'image_count': 32}, {'id': 524, 'image_count': 121}, {'id': 525, 'image_count': 16}, {'id': 526, 'image_count': 12}, {'id': 527, 'image_count': 2}, {'id': 528, 'image_count': 214}, {'id': 529, 'image_count': 48}, {'id': 530, 'image_count': 26}, {'id': 531, 'image_count': 13}, {'id': 532, 'image_count': 4}, {'id': 533, 'image_count': 11}, {'id': 534, 'image_count': 123}, {'id': 535, 'image_count': 7}, {'id': 536, 'image_count': 200}, {'id': 537, 'image_count': 91}, {'id': 538, 'image_count': 9}, {'id': 539, 'image_count': 72}, {'id': 540, 'image_count': 1886}, {'id': 541, 'image_count': 4}, {'id': 542, 'image_count': 1}, {'id': 543, 'image_count': 1}, {'id': 544, 'image_count': 1932}, {'id': 545, 'image_count': 4}, {'id': 546, 'image_count': 56}, {'id': 547, 'image_count': 854}, {'id': 548, 'image_count': 755}, {'id': 549, 'image_count': 1843}, {'id': 550, 'image_count': 96}, {'id': 551, 'image_count': 7}, {'id': 552, 'image_count': 74}, {'id': 553, 'image_count': 66}, {'id': 554, 'image_count': 57}, {'id': 555, 'image_count': 44}, {'id': 556, 'image_count': 1905}, {'id': 557, 'image_count': 4}, {'id': 558, 'image_count': 90}, {'id': 559, 'image_count': 1635}, {'id': 560, 'image_count': 8}, {'id': 561, 'image_count': 5}, {'id': 562, 'image_count': 50}, {'id': 563, 'image_count': 545}, {'id': 564, 'image_count': 20}, {'id': 565, 'image_count': 193}, {'id': 566, 'image_count': 285}, {'id': 567, 'image_count': 3}, {'id': 568, 'image_count': 1}, {'id': 569, 'image_count': 1904}, {'id': 570, 'image_count': 294}, {'id': 571, 'image_count': 3}, {'id': 572, 'image_count': 5}, {'id': 573, 'image_count': 24}, {'id': 574, 'image_count': 2}, {'id': 575, 'image_count': 2}, {'id': 576, 'image_count': 16}, {'id': 577, 'image_count': 8}, {'id': 578, 'image_count': 154}, {'id': 579, 'image_count': 66}, {'id': 580, 'image_count': 1}, {'id': 581, 'image_count': 24}, {'id': 582, 'image_count': 1}, {'id': 583, 'image_count': 4}, {'id': 584, 'image_count': 75}, {'id': 585, 'image_count': 6}, {'id': 586, 'image_count': 126}, {'id': 587, 'image_count': 24}, {'id': 588, 'image_count': 22}, {'id': 589, 'image_count': 1872}, {'id': 590, 'image_count': 16}, {'id': 591, 'image_count': 423}, {'id': 592, 'image_count': 1927}, {'id': 593, 'image_count': 38}, {'id': 594, 'image_count': 3}, {'id': 595, 'image_count': 1945}, {'id': 596, 'image_count': 35}, {'id': 597, 'image_count': 1}, {'id': 598, 'image_count': 13}, {'id': 599, 'image_count': 9}, {'id': 600, 'image_count': 14}, {'id': 601, 'image_count': 37}, {'id': 602, 'image_count': 3}, {'id': 603, 'image_count': 4}, {'id': 604, 'image_count': 100}, {'id': 605, 'image_count': 195}, {'id': 606, 'image_count': 1}, {'id': 607, 'image_count': 12}, {'id': 608, 'image_count': 24}, {'id': 609, 'image_count': 489}, {'id': 610, 'image_count': 10}, {'id': 611, 'image_count': 1689}, {'id': 612, 'image_count': 42}, {'id': 613, 'image_count': 81}, {'id': 614, 'image_count': 894}, {'id': 615, 'image_count': 1868}, {'id': 616, 'image_count': 7}, {'id': 617, 'image_count': 1567}, {'id': 618, 'image_count': 10}, {'id': 619, 'image_count': 8}, {'id': 620, 'image_count': 7}, {'id': 621, 'image_count': 629}, {'id': 622, 'image_count': 89}, {'id': 623, 'image_count': 15}, {'id': 624, 'image_count': 134}, {'id': 625, 'image_count': 4}, {'id': 626, 'image_count': 1802}, {'id': 627, 'image_count': 595}, {'id': 628, 'image_count': 1210}, {'id': 629, 'image_count': 48}, {'id': 630, 'image_count': 418}, {'id': 631, 'image_count': 1846}, {'id': 632, 'image_count': 5}, {'id': 633, 'image_count': 221}, {'id': 634, 'image_count': 10}, {'id': 635, 'image_count': 7}, {'id': 636, 'image_count': 76}, {'id': 637, 'image_count': 22}, {'id': 638, 'image_count': 10}, {'id': 639, 'image_count': 341}, {'id': 640, 'image_count': 1}, {'id': 641, 'image_count': 705}, {'id': 642, 'image_count': 1900}, {'id': 643, 'image_count': 188}, {'id': 644, 'image_count': 227}, {'id': 645, 'image_count': 861}, {'id': 646, 'image_count': 6}, {'id': 647, 'image_count': 115}, {'id': 648, 'image_count': 5}, {'id': 649, 'image_count': 43}, {'id': 650, 'image_count': 14}, {'id': 651, 'image_count': 6}, {'id': 652, 'image_count': 15}, {'id': 653, 'image_count': 1167}, {'id': 654, 'image_count': 15}, {'id': 655, 'image_count': 994}, {'id': 656, 'image_count': 28}, {'id': 657, 'image_count': 2}, {'id': 658, 'image_count': 338}, {'id': 659, 'image_count': 334}, {'id': 660, 'image_count': 15}, {'id': 661, 'image_count': 102}, {'id': 662, 'image_count': 1}, {'id': 663, 'image_count': 8}, {'id': 664, 'image_count': 1}, {'id': 665, 'image_count': 1}, {'id': 666, 'image_count': 28}, {'id': 667, 'image_count': 91}, {'id': 668, 'image_count': 260}, {'id': 669, 'image_count': 131}, {'id': 670, 'image_count': 128}, {'id': 671, 'image_count': 3}, {'id': 672, 'image_count': 10}, {'id': 673, 'image_count': 39}, {'id': 674, 'image_count': 2}, {'id': 675, 'image_count': 925}, {'id': 676, 'image_count': 354}, {'id': 677, 'image_count': 31}, {'id': 678, 'image_count': 10}, {'id': 679, 'image_count': 215}, {'id': 680, 'image_count': 71}, {'id': 681, 'image_count': 43}, {'id': 682, 'image_count': 28}, {'id': 683, 'image_count': 34}, {'id': 684, 'image_count': 16}, {'id': 685, 'image_count': 273}, {'id': 686, 'image_count': 2}, {'id': 687, 'image_count': 999}, {'id': 688, 'image_count': 4}, {'id': 689, 'image_count': 107}, {'id': 690, 'image_count': 2}, {'id': 691, 'image_count': 1}, {'id': 692, 'image_count': 454}, {'id': 693, 'image_count': 9}, {'id': 694, 'image_count': 1901}, {'id': 695, 'image_count': 61}, {'id': 696, 'image_count': 91}, {'id': 697, 'image_count': 46}, {'id': 698, 'image_count': 1402}, {'id': 699, 'image_count': 74}, {'id': 700, 'image_count': 421}, {'id': 701, 'image_count': 226}, {'id': 702, 'image_count': 10}, {'id': 703, 'image_count': 1720}, {'id': 704, 'image_count': 261}, {'id': 705, 'image_count': 1337}, {'id': 706, 'image_count': 293}, {'id': 707, 'image_count': 62}, {'id': 708, 'image_count': 814}, {'id': 709, 'image_count': 407}, {'id': 710, 'image_count': 6}, {'id': 711, 'image_count': 16}, {'id': 712, 'image_count': 7}, {'id': 713, 'image_count': 1791}, {'id': 714, 'image_count': 2}, {'id': 715, 'image_count': 1915}, {'id': 716, 'image_count': 1940}, {'id': 717, 'image_count': 13}, {'id': 718, 'image_count': 16}, {'id': 719, 'image_count': 448}, {'id': 720, 'image_count': 12}, {'id': 721, 'image_count': 18}, {'id': 722, 'image_count': 4}, {'id': 723, 'image_count': 71}, {'id': 724, 'image_count': 189}, {'id': 725, 'image_count': 74}, {'id': 726, 'image_count': 103}, {'id': 727, 'image_count': 3}, {'id': 728, 'image_count': 110}, {'id': 729, 'image_count': 5}, {'id': 730, 'image_count': 9}, {'id': 731, 'image_count': 15}, {'id': 732, 'image_count': 25}, {'id': 733, 'image_count': 7}, {'id': 734, 'image_count': 647}, {'id': 735, 'image_count': 824}, {'id': 736, 'image_count': 100}, {'id': 737, 'image_count': 47}, {'id': 738, 'image_count': 121}, {'id': 739, 'image_count': 731}, {'id': 740, 'image_count': 73}, {'id': 741, 'image_count': 49}, {'id': 742, 'image_count': 23}, {'id': 743, 'image_count': 4}, {'id': 744, 'image_count': 62}, {'id': 745, 'image_count': 118}, {'id': 746, 'image_count': 99}, {'id': 747, 'image_count': 40}, {'id': 748, 'image_count': 1036}, {'id': 749, 'image_count': 105}, {'id': 750, 'image_count': 21}, {'id': 751, 'image_count': 229}, {'id': 752, 'image_count': 7}, {'id': 753, 'image_count': 72}, {'id': 754, 'image_count': 9}, {'id': 755, 'image_count': 10}, {'id': 756, 'image_count': 328}, {'id': 757, 'image_count': 468}, {'id': 758, 'image_count': 1}, {'id': 759, 'image_count': 2}, {'id': 760, 'image_count': 24}, {'id': 761, 'image_count': 11}, {'id': 762, 'image_count': 72}, {'id': 763, 'image_count': 17}, {'id': 764, 'image_count': 10}, {'id': 765, 'image_count': 17}, {'id': 766, 'image_count': 489}, {'id': 767, 'image_count': 47}, {'id': 768, 'image_count': 93}, {'id': 769, 'image_count': 1}, {'id': 770, 'image_count': 12}, {'id': 771, 'image_count': 228}, {'id': 772, 'image_count': 5}, {'id': 773, 'image_count': 76}, {'id': 774, 'image_count': 71}, {'id': 775, 'image_count': 30}, {'id': 776, 'image_count': 109}, {'id': 777, 'image_count': 14}, {'id': 778, 'image_count': 1}, {'id': 779, 'image_count': 8}, {'id': 780, 'image_count': 26}, {'id': 781, 'image_count': 339}, {'id': 782, 'image_count': 153}, {'id': 783, 'image_count': 2}, {'id': 784, 'image_count': 3}, {'id': 785, 'image_count': 8}, {'id': 786, 'image_count': 47}, {'id': 787, 'image_count': 8}, {'id': 788, 'image_count': 6}, {'id': 789, 'image_count': 116}, {'id': 790, 'image_count': 69}, {'id': 791, 'image_count': 13}, {'id': 792, 'image_count': 6}, {'id': 793, 'image_count': 1928}, {'id': 794, 'image_count': 79}, {'id': 795, 'image_count': 14}, {'id': 796, 'image_count': 7}, {'id': 797, 'image_count': 20}, {'id': 798, 'image_count': 114}, {'id': 799, 'image_count': 221}, {'id': 800, 'image_count': 502}, {'id': 801, 'image_count': 62}, {'id': 802, 'image_count': 87}, {'id': 803, 'image_count': 4}, {'id': 804, 'image_count': 1912}, {'id': 805, 'image_count': 7}, {'id': 806, 'image_count': 186}, {'id': 807, 'image_count': 18}, {'id': 808, 'image_count': 4}, {'id': 809, 'image_count': 3}, {'id': 810, 'image_count': 7}, {'id': 811, 'image_count': 1413}, {'id': 812, 'image_count': 7}, {'id': 813, 'image_count': 12}, {'id': 814, 'image_count': 248}, {'id': 815, 'image_count': 4}, {'id': 816, 'image_count': 1881}, {'id': 817, 'image_count': 529}, {'id': 818, 'image_count': 1932}, {'id': 819, 'image_count': 50}, {'id': 820, 'image_count': 3}, {'id': 821, 'image_count': 28}, {'id': 822, 'image_count': 10}, {'id': 823, 'image_count': 5}, {'id': 824, 'image_count': 5}, {'id': 825, 'image_count': 18}, {'id': 826, 'image_count': 14}, {'id': 827, 'image_count': 1890}, {'id': 828, 'image_count': 660}, {'id': 829, 'image_count': 8}, {'id': 830, 'image_count': 25}, {'id': 831, 'image_count': 10}, {'id': 832, 'image_count': 218}, {'id': 833, 'image_count': 36}, {'id': 834, 'image_count': 16}, {'id': 835, 'image_count': 808}, {'id': 836, 'image_count': 479}, {'id': 837, 'image_count': 1404}, {'id': 838, 'image_count': 307}, {'id': 839, 'image_count': 57}, {'id': 840, 'image_count': 28}, {'id': 841, 'image_count': 80}, {'id': 842, 'image_count': 11}, {'id': 843, 'image_count': 92}, {'id': 844, 'image_count': 20}, {'id': 845, 'image_count': 194}, {'id': 846, 'image_count': 23}, {'id': 847, 'image_count': 52}, {'id': 848, 'image_count': 673}, {'id': 849, 'image_count': 2}, {'id': 850, 'image_count': 2}, {'id': 851, 'image_count': 1}, {'id': 852, 'image_count': 2}, {'id': 853, 'image_count': 8}, {'id': 854, 'image_count': 80}, {'id': 855, 'image_count': 3}, {'id': 856, 'image_count': 3}, {'id': 857, 'image_count': 15}, {'id': 858, 'image_count': 2}, {'id': 859, 'image_count': 10}, {'id': 860, 'image_count': 386}, {'id': 861, 'image_count': 65}, {'id': 862, 'image_count': 3}, {'id': 863, 'image_count': 35}, {'id': 864, 'image_count': 5}, {'id': 865, 'image_count': 180}, {'id': 866, 'image_count': 99}, {'id': 867, 'image_count': 49}, {'id': 868, 'image_count': 28}, {'id': 869, 'image_count': 1}, {'id': 870, 'image_count': 52}, {'id': 871, 'image_count': 36}, {'id': 872, 'image_count': 70}, {'id': 873, 'image_count': 6}, {'id': 874, 'image_count': 29}, {'id': 875, 'image_count': 24}, {'id': 876, 'image_count': 1115}, {'id': 877, 'image_count': 61}, {'id': 878, 'image_count': 18}, {'id': 879, 'image_count': 18}, {'id': 880, 'image_count': 665}, {'id': 881, 'image_count': 1096}, {'id': 882, 'image_count': 29}, {'id': 883, 'image_count': 8}, {'id': 884, 'image_count': 14}, {'id': 885, 'image_count': 1622}, {'id': 886, 'image_count': 2}, {'id': 887, 'image_count': 3}, {'id': 888, 'image_count': 32}, {'id': 889, 'image_count': 55}, {'id': 890, 'image_count': 1}, {'id': 891, 'image_count': 10}, {'id': 892, 'image_count': 10}, {'id': 893, 'image_count': 47}, {'id': 894, 'image_count': 3}, {'id': 895, 'image_count': 29}, {'id': 896, 'image_count': 342}, {'id': 897, 'image_count': 25}, {'id': 898, 'image_count': 1469}, {'id': 899, 'image_count': 521}, {'id': 900, 'image_count': 347}, {'id': 901, 'image_count': 35}, {'id': 902, 'image_count': 7}, {'id': 903, 'image_count': 207}, {'id': 904, 'image_count': 108}, {'id': 905, 'image_count': 2}, {'id': 906, 'image_count': 34}, {'id': 907, 'image_count': 12}, {'id': 908, 'image_count': 10}, {'id': 909, 'image_count': 13}, {'id': 910, 'image_count': 361}, {'id': 911, 'image_count': 1023}, {'id': 912, 'image_count': 782}, {'id': 913, 'image_count': 2}, {'id': 914, 'image_count': 5}, {'id': 915, 'image_count': 247}, {'id': 916, 'image_count': 221}, {'id': 917, 'image_count': 4}, {'id': 918, 'image_count': 8}, {'id': 919, 'image_count': 158}, {'id': 920, 'image_count': 3}, {'id': 921, 'image_count': 752}, {'id': 922, 'image_count': 64}, {'id': 923, 'image_count': 707}, {'id': 924, 'image_count': 143}, {'id': 925, 'image_count': 1}, {'id': 926, 'image_count': 49}, {'id': 927, 'image_count': 126}, {'id': 928, 'image_count': 76}, {'id': 929, 'image_count': 11}, {'id': 930, 'image_count': 11}, {'id': 931, 'image_count': 4}, {'id': 932, 'image_count': 39}, {'id': 933, 'image_count': 11}, {'id': 934, 'image_count': 13}, {'id': 935, 'image_count': 91}, {'id': 936, 'image_count': 14}, {'id': 937, 'image_count': 5}, {'id': 938, 'image_count': 3}, {'id': 939, 'image_count': 10}, {'id': 940, 'image_count': 18}, {'id': 941, 'image_count': 9}, {'id': 942, 'image_count': 6}, {'id': 943, 'image_count': 951}, {'id': 944, 'image_count': 2}, {'id': 945, 'image_count': 1}, {'id': 946, 'image_count': 19}, {'id': 947, 'image_count': 1942}, {'id': 948, 'image_count': 1916}, {'id': 949, 'image_count': 139}, {'id': 950, 'image_count': 43}, {'id': 951, 'image_count': 1969}, {'id': 952, 'image_count': 5}, {'id': 953, 'image_count': 134}, {'id': 954, 'image_count': 74}, {'id': 955, 'image_count': 381}, {'id': 956, 'image_count': 1}, {'id': 957, 'image_count': 381}, {'id': 958, 'image_count': 6}, {'id': 959, 'image_count': 1826}, {'id': 960, 'image_count': 28}, {'id': 961, 'image_count': 1635}, {'id': 962, 'image_count': 1967}, {'id': 963, 'image_count': 16}, {'id': 964, 'image_count': 1926}, {'id': 965, 'image_count': 1789}, {'id': 966, 'image_count': 401}, {'id': 967, 'image_count': 1968}, {'id': 968, 'image_count': 1167}, {'id': 969, 'image_count': 1}, {'id': 970, 'image_count': 56}, {'id': 971, 'image_count': 17}, {'id': 972, 'image_count': 1}, {'id': 973, 'image_count': 58}, {'id': 974, 'image_count': 9}, {'id': 975, 'image_count': 8}, {'id': 976, 'image_count': 1124}, {'id': 977, 'image_count': 31}, {'id': 978, 'image_count': 16}, {'id': 979, 'image_count': 491}, {'id': 980, 'image_count': 432}, {'id': 981, 'image_count': 1945}, {'id': 982, 'image_count': 1899}, {'id': 983, 'image_count': 5}, {'id': 984, 'image_count': 28}, {'id': 985, 'image_count': 7}, {'id': 986, 'image_count': 146}, {'id': 987, 'image_count': 1}, {'id': 988, 'image_count': 25}, {'id': 989, 'image_count': 22}, {'id': 990, 'image_count': 1}, {'id': 991, 'image_count': 10}, {'id': 992, 'image_count': 9}, {'id': 993, 'image_count': 308}, {'id': 994, 'image_count': 4}, {'id': 995, 'image_count': 1969}, {'id': 996, 'image_count': 45}, {'id': 997, 'image_count': 12}, {'id': 998, 'image_count': 1}, {'id': 999, 'image_count': 85}, {'id': 1000, 'image_count': 1127}, {'id': 1001, 'image_count': 11}, {'id': 1002, 'image_count': 60}, {'id': 1003, 'image_count': 1}, {'id': 1004, 'image_count': 16}, {'id': 1005, 'image_count': 1}, {'id': 1006, 'image_count': 65}, {'id': 1007, 'image_count': 13}, {'id': 1008, 'image_count': 655}, {'id': 1009, 'image_count': 51}, {'id': 1010, 'image_count': 1}, {'id': 1011, 'image_count': 673}, {'id': 1012, 'image_count': 5}, {'id': 1013, 'image_count': 36}, {'id': 1014, 'image_count': 54}, {'id': 1015, 'image_count': 5}, {'id': 1016, 'image_count': 8}, {'id': 1017, 'image_count': 305}, {'id': 1018, 'image_count': 297}, {'id': 1019, 'image_count': 1053}, {'id': 1020, 'image_count': 223}, {'id': 1021, 'image_count': 1037}, {'id': 1022, 'image_count': 63}, {'id': 1023, 'image_count': 1881}, {'id': 1024, 'image_count': 507}, {'id': 1025, 'image_count': 333}, {'id': 1026, 'image_count': 1911}, {'id': 1027, 'image_count': 1765}, {'id': 1028, 'image_count': 1}, {'id': 1029, 'image_count': 5}, {'id': 1030, 'image_count': 1}, {'id': 1031, 'image_count': 9}, {'id': 1032, 'image_count': 2}, {'id': 1033, 'image_count': 151}, {'id': 1034, 'image_count': 82}, {'id': 1035, 'image_count': 1931}, {'id': 1036, 'image_count': 41}, {'id': 1037, 'image_count': 1895}, {'id': 1038, 'image_count': 24}, {'id': 1039, 'image_count': 22}, {'id': 1040, 'image_count': 35}, {'id': 1041, 'image_count': 69}, {'id': 1042, 'image_count': 962}, {'id': 1043, 'image_count': 588}, {'id': 1044, 'image_count': 21}, {'id': 1045, 'image_count': 825}, {'id': 1046, 'image_count': 52}, {'id': 1047, 'image_count': 5}, {'id': 1048, 'image_count': 5}, {'id': 1049, 'image_count': 5}, {'id': 1050, 'image_count': 1860}, {'id': 1051, 'image_count': 56}, {'id': 1052, 'image_count': 1582}, {'id': 1053, 'image_count': 7}, {'id': 1054, 'image_count': 2}, {'id': 1055, 'image_count': 1562}, {'id': 1056, 'image_count': 1885}, {'id': 1057, 'image_count': 1}, {'id': 1058, 'image_count': 5}, {'id': 1059, 'image_count': 137}, {'id': 1060, 'image_count': 1094}, {'id': 1061, 'image_count': 134}, {'id': 1062, 'image_count': 29}, {'id': 1063, 'image_count': 22}, {'id': 1064, 'image_count': 522}, {'id': 1065, 'image_count': 50}, {'id': 1066, 'image_count': 68}, {'id': 1067, 'image_count': 16}, {'id': 1068, 'image_count': 40}, {'id': 1069, 'image_count': 35}, {'id': 1070, 'image_count': 135}, {'id': 1071, 'image_count': 1413}, {'id': 1072, 'image_count': 772}, {'id': 1073, 'image_count': 50}, {'id': 1074, 'image_count': 1015}, {'id': 1075, 'image_count': 1}, {'id': 1076, 'image_count': 65}, {'id': 1077, 'image_count': 1900}, {'id': 1078, 'image_count': 1302}, {'id': 1079, 'image_count': 1977}, {'id': 1080, 'image_count': 2}, {'id': 1081, 'image_count': 29}, {'id': 1082, 'image_count': 36}, {'id': 1083, 'image_count': 138}, {'id': 1084, 'image_count': 4}, {'id': 1085, 'image_count': 67}, {'id': 1086, 'image_count': 26}, {'id': 1087, 'image_count': 25}, {'id': 1088, 'image_count': 33}, {'id': 1089, 'image_count': 37}, {'id': 1090, 'image_count': 50}, {'id': 1091, 'image_count': 270}, {'id': 1092, 'image_count': 12}, {'id': 1093, 'image_count': 316}, {'id': 1094, 'image_count': 41}, {'id': 1095, 'image_count': 224}, {'id': 1096, 'image_count': 105}, {'id': 1097, 'image_count': 1925}, {'id': 1098, 'image_count': 1021}, {'id': 1099, 'image_count': 1213}, {'id': 1100, 'image_count': 172}, {'id': 1101, 'image_count': 28}, {'id': 1102, 'image_count': 745}, {'id': 1103, 'image_count': 187}, {'id': 1104, 'image_count': 147}, {'id': 1105, 'image_count': 136}, {'id': 1106, 'image_count': 34}, {'id': 1107, 'image_count': 41}, {'id': 1108, 'image_count': 636}, {'id': 1109, 'image_count': 570}, {'id': 1110, 'image_count': 1149}, {'id': 1111, 'image_count': 61}, {'id': 1112, 'image_count': 1890}, {'id': 1113, 'image_count': 18}, {'id': 1114, 'image_count': 143}, {'id': 1115, 'image_count': 1517}, {'id': 1116, 'image_count': 7}, {'id': 1117, 'image_count': 943}, {'id': 1118, 'image_count': 6}, {'id': 1119, 'image_count': 1}, {'id': 1120, 'image_count': 11}, {'id': 1121, 'image_count': 101}, {'id': 1122, 'image_count': 1909}, {'id': 1123, 'image_count': 800}, {'id': 1124, 'image_count': 1}, {'id': 1125, 'image_count': 44}, {'id': 1126, 'image_count': 3}, {'id': 1127, 'image_count': 44}, {'id': 1128, 'image_count': 31}, {'id': 1129, 'image_count': 7}, {'id': 1130, 'image_count': 20}, {'id': 1131, 'image_count': 11}, {'id': 1132, 'image_count': 13}, {'id': 1133, 'image_count': 1924}, {'id': 1134, 'image_count': 113}, {'id': 1135, 'image_count': 2}, {'id': 1136, 'image_count': 139}, {'id': 1137, 'image_count': 12}, {'id': 1138, 'image_count': 37}, {'id': 1139, 'image_count': 1866}, {'id': 1140, 'image_count': 47}, {'id': 1141, 'image_count': 1468}, {'id': 1142, 'image_count': 729}, {'id': 1143, 'image_count': 24}, {'id': 1144, 'image_count': 1}, {'id': 1145, 'image_count': 10}, {'id': 1146, 'image_count': 3}, {'id': 1147, 'image_count': 14}, {'id': 1148, 'image_count': 4}, {'id': 1149, 'image_count': 29}, {'id': 1150, 'image_count': 4}, {'id': 1151, 'image_count': 70}, {'id': 1152, 'image_count': 46}, {'id': 1153, 'image_count': 14}, {'id': 1154, 'image_count': 48}, {'id': 1155, 'image_count': 1855}, {'id': 1156, 'image_count': 113}, {'id': 1157, 'image_count': 1}, {'id': 1158, 'image_count': 1}, {'id': 1159, 'image_count': 10}, {'id': 1160, 'image_count': 54}, {'id': 1161, 'image_count': 1923}, {'id': 1162, 'image_count': 630}, {'id': 1163, 'image_count': 31}, {'id': 1164, 'image_count': 69}, {'id': 1165, 'image_count': 7}, {'id': 1166, 'image_count': 11}, {'id': 1167, 'image_count': 1}, {'id': 1168, 'image_count': 30}, {'id': 1169, 'image_count': 50}, {'id': 1170, 'image_count': 45}, {'id': 1171, 'image_count': 28}, {'id': 1172, 'image_count': 114}, {'id': 1173, 'image_count': 193}, {'id': 1174, 'image_count': 21}, {'id': 1175, 'image_count': 91}, {'id': 1176, 'image_count': 31}, {'id': 1177, 'image_count': 1469}, {'id': 1178, 'image_count': 1924}, {'id': 1179, 'image_count': 87}, {'id': 1180, 'image_count': 77}, {'id': 1181, 'image_count': 11}, {'id': 1182, 'image_count': 47}, {'id': 1183, 'image_count': 21}, {'id': 1184, 'image_count': 47}, {'id': 1185, 'image_count': 70}, {'id': 1186, 'image_count': 1838}, {'id': 1187, 'image_count': 19}, {'id': 1188, 'image_count': 531}, {'id': 1189, 'image_count': 11}, {'id': 1190, 'image_count': 941}, {'id': 1191, 'image_count': 113}, {'id': 1192, 'image_count': 26}, {'id': 1193, 'image_count': 5}, {'id': 1194, 'image_count': 56}, {'id': 1195, 'image_count': 73}, {'id': 1196, 'image_count': 32}, {'id': 1197, 'image_count': 128}, {'id': 1198, 'image_count': 623}, {'id': 1199, 'image_count': 12}, {'id': 1200, 'image_count': 52}, {'id': 1201, 'image_count': 11}, {'id': 1202, 'image_count': 1674}, {'id': 1203, 'image_count': 81}] # noqa
+# fmt: on
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/pascal_voc.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/pascal_voc.py
new file mode 100644
index 0000000000000000000000000000000000000000..919cc4920394d3cb87ad5232adcbedc250e4db26
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/pascal_voc.py
@@ -0,0 +1,82 @@
+# -*- coding: utf-8 -*-
+# Copyright (c) Facebook, Inc. and its affiliates.
+
+import numpy as np
+import os
+import xml.etree.ElementTree as ET
+from typing import List, Tuple, Union
+
+from annotator.oneformer.detectron2.data import DatasetCatalog, MetadataCatalog
+from annotator.oneformer.detectron2.structures import BoxMode
+from annotator.oneformer.detectron2.utils.file_io import PathManager
+
+__all__ = ["load_voc_instances", "register_pascal_voc"]
+
+
+# fmt: off
+CLASS_NAMES = (
+ "aeroplane", "bicycle", "bird", "boat", "bottle", "bus", "car", "cat",
+ "chair", "cow", "diningtable", "dog", "horse", "motorbike", "person",
+ "pottedplant", "sheep", "sofa", "train", "tvmonitor"
+)
+# fmt: on
+
+
+def load_voc_instances(dirname: str, split: str, class_names: Union[List[str], Tuple[str, ...]]):
+ """
+ Load Pascal VOC detection annotations to Detectron2 format.
+
+ Args:
+ dirname: Contain "Annotations", "ImageSets", "JPEGImages"
+ split (str): one of "train", "test", "val", "trainval"
+ class_names: list or tuple of class names
+ """
+ with PathManager.open(os.path.join(dirname, "ImageSets", "Main", split + ".txt")) as f:
+ fileids = np.loadtxt(f, dtype=np.str)
+
+ # Needs to read many small annotation files. Makes sense at local
+ annotation_dirname = PathManager.get_local_path(os.path.join(dirname, "Annotations/"))
+ dicts = []
+ for fileid in fileids:
+ anno_file = os.path.join(annotation_dirname, fileid + ".xml")
+ jpeg_file = os.path.join(dirname, "JPEGImages", fileid + ".jpg")
+
+ with PathManager.open(anno_file) as f:
+ tree = ET.parse(f)
+
+ r = {
+ "file_name": jpeg_file,
+ "image_id": fileid,
+ "height": int(tree.findall("./size/height")[0].text),
+ "width": int(tree.findall("./size/width")[0].text),
+ }
+ instances = []
+
+ for obj in tree.findall("object"):
+ cls = obj.find("name").text
+ # We include "difficult" samples in training.
+ # Based on limited experiments, they don't hurt accuracy.
+ # difficult = int(obj.find("difficult").text)
+ # if difficult == 1:
+ # continue
+ bbox = obj.find("bndbox")
+ bbox = [float(bbox.find(x).text) for x in ["xmin", "ymin", "xmax", "ymax"]]
+ # Original annotations are integers in the range [1, W or H]
+ # Assuming they mean 1-based pixel indices (inclusive),
+ # a box with annotation (xmin=1, xmax=W) covers the whole image.
+ # In coordinate space this is represented by (xmin=0, xmax=W)
+ bbox[0] -= 1.0
+ bbox[1] -= 1.0
+ instances.append(
+ {"category_id": class_names.index(cls), "bbox": bbox, "bbox_mode": BoxMode.XYXY_ABS}
+ )
+ r["annotations"] = instances
+ dicts.append(r)
+ return dicts
+
+
+def register_pascal_voc(name, dirname, split, year, class_names=CLASS_NAMES):
+ DatasetCatalog.register(name, lambda: load_voc_instances(dirname, split, class_names))
+ MetadataCatalog.get(name).set(
+ thing_classes=list(class_names), dirname=dirname, year=year, split=split
+ )
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/register_coco.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/register_coco.py
new file mode 100644
index 0000000000000000000000000000000000000000..e564438d5bf016bcdbb65b4bbdc215d79f579f8a
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/datasets/register_coco.py
@@ -0,0 +1,3 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+from .coco import register_coco_instances # noqa
+from .coco_panoptic import register_coco_panoptic_separated # noqa
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/detection_utils.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/detection_utils.py
new file mode 100644
index 0000000000000000000000000000000000000000..99ce45f52bab8ff87dba3e9e008947eef2f7c33e
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/detection_utils.py
@@ -0,0 +1,659 @@
+# -*- coding: utf-8 -*-
+# Copyright (c) Facebook, Inc. and its affiliates.
+
+"""
+Common data processing utilities that are used in a
+typical object detection data pipeline.
+"""
+import logging
+import numpy as np
+from typing import List, Union
+import pycocotools.mask as mask_util
+import torch
+from PIL import Image
+
+from annotator.oneformer.detectron2.structures import (
+ BitMasks,
+ Boxes,
+ BoxMode,
+ Instances,
+ Keypoints,
+ PolygonMasks,
+ RotatedBoxes,
+ polygons_to_bitmask,
+)
+from annotator.oneformer.detectron2.utils.file_io import PathManager
+
+from . import transforms as T
+from .catalog import MetadataCatalog
+
+__all__ = [
+ "SizeMismatchError",
+ "convert_image_to_rgb",
+ "check_image_size",
+ "transform_proposals",
+ "transform_instance_annotations",
+ "annotations_to_instances",
+ "annotations_to_instances_rotated",
+ "build_augmentation",
+ "build_transform_gen",
+ "create_keypoint_hflip_indices",
+ "filter_empty_instances",
+ "read_image",
+]
+
+
+class SizeMismatchError(ValueError):
+ """
+ When loaded image has difference width/height compared with annotation.
+ """
+
+
+# https://en.wikipedia.org/wiki/YUV#SDTV_with_BT.601
+_M_RGB2YUV = [[0.299, 0.587, 0.114], [-0.14713, -0.28886, 0.436], [0.615, -0.51499, -0.10001]]
+_M_YUV2RGB = [[1.0, 0.0, 1.13983], [1.0, -0.39465, -0.58060], [1.0, 2.03211, 0.0]]
+
+# https://www.exiv2.org/tags.html
+_EXIF_ORIENT = 274 # exif 'Orientation' tag
+
+
+def convert_PIL_to_numpy(image, format):
+ """
+ Convert PIL image to numpy array of target format.
+
+ Args:
+ image (PIL.Image): a PIL image
+ format (str): the format of output image
+
+ Returns:
+ (np.ndarray): also see `read_image`
+ """
+ if format is not None:
+ # PIL only supports RGB, so convert to RGB and flip channels over below
+ conversion_format = format
+ if format in ["BGR", "YUV-BT.601"]:
+ conversion_format = "RGB"
+ image = image.convert(conversion_format)
+ image = np.asarray(image)
+ # PIL squeezes out the channel dimension for "L", so make it HWC
+ if format == "L":
+ image = np.expand_dims(image, -1)
+
+ # handle formats not supported by PIL
+ elif format == "BGR":
+ # flip channels if needed
+ image = image[:, :, ::-1]
+ elif format == "YUV-BT.601":
+ image = image / 255.0
+ image = np.dot(image, np.array(_M_RGB2YUV).T)
+
+ return image
+
+
+def convert_image_to_rgb(image, format):
+ """
+ Convert an image from given format to RGB.
+
+ Args:
+ image (np.ndarray or Tensor): an HWC image
+ format (str): the format of input image, also see `read_image`
+
+ Returns:
+ (np.ndarray): (H,W,3) RGB image in 0-255 range, can be either float or uint8
+ """
+ if isinstance(image, torch.Tensor):
+ image = image.cpu().numpy()
+ if format == "BGR":
+ image = image[:, :, [2, 1, 0]]
+ elif format == "YUV-BT.601":
+ image = np.dot(image, np.array(_M_YUV2RGB).T)
+ image = image * 255.0
+ else:
+ if format == "L":
+ image = image[:, :, 0]
+ image = image.astype(np.uint8)
+ image = np.asarray(Image.fromarray(image, mode=format).convert("RGB"))
+ return image
+
+
+def _apply_exif_orientation(image):
+ """
+ Applies the exif orientation correctly.
+
+ This code exists per the bug:
+ https://github.com/python-pillow/Pillow/issues/3973
+ with the function `ImageOps.exif_transpose`. The Pillow source raises errors with
+ various methods, especially `tobytes`
+
+ Function based on:
+ https://github.com/wkentaro/labelme/blob/v4.5.4/labelme/utils/image.py#L59
+ https://github.com/python-pillow/Pillow/blob/7.1.2/src/PIL/ImageOps.py#L527
+
+ Args:
+ image (PIL.Image): a PIL image
+
+ Returns:
+ (PIL.Image): the PIL image with exif orientation applied, if applicable
+ """
+ if not hasattr(image, "getexif"):
+ return image
+
+ try:
+ exif = image.getexif()
+ except Exception: # https://github.com/facebookresearch/detectron2/issues/1885
+ exif = None
+
+ if exif is None:
+ return image
+
+ orientation = exif.get(_EXIF_ORIENT)
+
+ method = {
+ 2: Image.FLIP_LEFT_RIGHT,
+ 3: Image.ROTATE_180,
+ 4: Image.FLIP_TOP_BOTTOM,
+ 5: Image.TRANSPOSE,
+ 6: Image.ROTATE_270,
+ 7: Image.TRANSVERSE,
+ 8: Image.ROTATE_90,
+ }.get(orientation)
+
+ if method is not None:
+ return image.transpose(method)
+ return image
+
+
+def read_image(file_name, format=None):
+ """
+ Read an image into the given format.
+ Will apply rotation and flipping if the image has such exif information.
+
+ Args:
+ file_name (str): image file path
+ format (str): one of the supported image modes in PIL, or "BGR" or "YUV-BT.601".
+
+ Returns:
+ image (np.ndarray):
+ an HWC image in the given format, which is 0-255, uint8 for
+ supported image modes in PIL or "BGR"; float (0-1 for Y) for YUV-BT.601.
+ """
+ with PathManager.open(file_name, "rb") as f:
+ image = Image.open(f)
+
+ # work around this bug: https://github.com/python-pillow/Pillow/issues/3973
+ image = _apply_exif_orientation(image)
+ return convert_PIL_to_numpy(image, format)
+
+
+def check_image_size(dataset_dict, image):
+ """
+ Raise an error if the image does not match the size specified in the dict.
+ """
+ if "width" in dataset_dict or "height" in dataset_dict:
+ image_wh = (image.shape[1], image.shape[0])
+ expected_wh = (dataset_dict["width"], dataset_dict["height"])
+ if not image_wh == expected_wh:
+ raise SizeMismatchError(
+ "Mismatched image shape{}, got {}, expect {}.".format(
+ " for image " + dataset_dict["file_name"]
+ if "file_name" in dataset_dict
+ else "",
+ image_wh,
+ expected_wh,
+ )
+ + " Please check the width/height in your annotation."
+ )
+
+ # To ensure bbox always remap to original image size
+ if "width" not in dataset_dict:
+ dataset_dict["width"] = image.shape[1]
+ if "height" not in dataset_dict:
+ dataset_dict["height"] = image.shape[0]
+
+
+def transform_proposals(dataset_dict, image_shape, transforms, *, proposal_topk, min_box_size=0):
+ """
+ Apply transformations to the proposals in dataset_dict, if any.
+
+ Args:
+ dataset_dict (dict): a dict read from the dataset, possibly
+ contains fields "proposal_boxes", "proposal_objectness_logits", "proposal_bbox_mode"
+ image_shape (tuple): height, width
+ transforms (TransformList):
+ proposal_topk (int): only keep top-K scoring proposals
+ min_box_size (int): proposals with either side smaller than this
+ threshold are removed
+
+ The input dict is modified in-place, with abovementioned keys removed. A new
+ key "proposals" will be added. Its value is an `Instances`
+ object which contains the transformed proposals in its field
+ "proposal_boxes" and "objectness_logits".
+ """
+ if "proposal_boxes" in dataset_dict:
+ # Transform proposal boxes
+ boxes = transforms.apply_box(
+ BoxMode.convert(
+ dataset_dict.pop("proposal_boxes"),
+ dataset_dict.pop("proposal_bbox_mode"),
+ BoxMode.XYXY_ABS,
+ )
+ )
+ boxes = Boxes(boxes)
+ objectness_logits = torch.as_tensor(
+ dataset_dict.pop("proposal_objectness_logits").astype("float32")
+ )
+
+ boxes.clip(image_shape)
+ keep = boxes.nonempty(threshold=min_box_size)
+ boxes = boxes[keep]
+ objectness_logits = objectness_logits[keep]
+
+ proposals = Instances(image_shape)
+ proposals.proposal_boxes = boxes[:proposal_topk]
+ proposals.objectness_logits = objectness_logits[:proposal_topk]
+ dataset_dict["proposals"] = proposals
+
+
+def get_bbox(annotation):
+ """
+ Get bbox from data
+ Args:
+ annotation (dict): dict of instance annotations for a single instance.
+ Returns:
+ bbox (ndarray): x1, y1, x2, y2 coordinates
+ """
+ # bbox is 1d (per-instance bounding box)
+ bbox = BoxMode.convert(annotation["bbox"], annotation["bbox_mode"], BoxMode.XYXY_ABS)
+ return bbox
+
+
+def transform_instance_annotations(
+ annotation, transforms, image_size, *, keypoint_hflip_indices=None
+):
+ """
+ Apply transforms to box, segmentation and keypoints annotations of a single instance.
+
+ It will use `transforms.apply_box` for the box, and
+ `transforms.apply_coords` for segmentation polygons & keypoints.
+ If you need anything more specially designed for each data structure,
+ you'll need to implement your own version of this function or the transforms.
+
+ Args:
+ annotation (dict): dict of instance annotations for a single instance.
+ It will be modified in-place.
+ transforms (TransformList or list[Transform]):
+ image_size (tuple): the height, width of the transformed image
+ keypoint_hflip_indices (ndarray[int]): see `create_keypoint_hflip_indices`.
+
+ Returns:
+ dict:
+ the same input dict with fields "bbox", "segmentation", "keypoints"
+ transformed according to `transforms`.
+ The "bbox_mode" field will be set to XYXY_ABS.
+ """
+ if isinstance(transforms, (tuple, list)):
+ transforms = T.TransformList(transforms)
+ # bbox is 1d (per-instance bounding box)
+ bbox = BoxMode.convert(annotation["bbox"], annotation["bbox_mode"], BoxMode.XYXY_ABS)
+ # clip transformed bbox to image size
+ bbox = transforms.apply_box(np.array([bbox]))[0].clip(min=0)
+ annotation["bbox"] = np.minimum(bbox, list(image_size + image_size)[::-1])
+ annotation["bbox_mode"] = BoxMode.XYXY_ABS
+
+ if "segmentation" in annotation:
+ # each instance contains 1 or more polygons
+ segm = annotation["segmentation"]
+ if isinstance(segm, list):
+ # polygons
+ polygons = [np.asarray(p).reshape(-1, 2) for p in segm]
+ annotation["segmentation"] = [
+ p.reshape(-1) for p in transforms.apply_polygons(polygons)
+ ]
+ elif isinstance(segm, dict):
+ # RLE
+ mask = mask_util.decode(segm)
+ mask = transforms.apply_segmentation(mask)
+ assert tuple(mask.shape[:2]) == image_size
+ annotation["segmentation"] = mask
+ else:
+ raise ValueError(
+ "Cannot transform segmentation of type '{}'!"
+ "Supported types are: polygons as list[list[float] or ndarray],"
+ " COCO-style RLE as a dict.".format(type(segm))
+ )
+
+ if "keypoints" in annotation:
+ keypoints = transform_keypoint_annotations(
+ annotation["keypoints"], transforms, image_size, keypoint_hflip_indices
+ )
+ annotation["keypoints"] = keypoints
+
+ return annotation
+
+
+def transform_keypoint_annotations(keypoints, transforms, image_size, keypoint_hflip_indices=None):
+ """
+ Transform keypoint annotations of an image.
+ If a keypoint is transformed out of image boundary, it will be marked "unlabeled" (visibility=0)
+
+ Args:
+ keypoints (list[float]): Nx3 float in Detectron2's Dataset format.
+ Each point is represented by (x, y, visibility).
+ transforms (TransformList):
+ image_size (tuple): the height, width of the transformed image
+ keypoint_hflip_indices (ndarray[int]): see `create_keypoint_hflip_indices`.
+ When `transforms` includes horizontal flip, will use the index
+ mapping to flip keypoints.
+ """
+ # (N*3,) -> (N, 3)
+ keypoints = np.asarray(keypoints, dtype="float64").reshape(-1, 3)
+ keypoints_xy = transforms.apply_coords(keypoints[:, :2])
+
+ # Set all out-of-boundary points to "unlabeled"
+ inside = (keypoints_xy >= np.array([0, 0])) & (keypoints_xy <= np.array(image_size[::-1]))
+ inside = inside.all(axis=1)
+ keypoints[:, :2] = keypoints_xy
+ keypoints[:, 2][~inside] = 0
+
+ # This assumes that HorizFlipTransform is the only one that does flip
+ do_hflip = sum(isinstance(t, T.HFlipTransform) for t in transforms.transforms) % 2 == 1
+
+ # Alternative way: check if probe points was horizontally flipped.
+ # probe = np.asarray([[0.0, 0.0], [image_width, 0.0]])
+ # probe_aug = transforms.apply_coords(probe.copy())
+ # do_hflip = np.sign(probe[1][0] - probe[0][0]) != np.sign(probe_aug[1][0] - probe_aug[0][0]) # noqa
+
+ # If flipped, swap each keypoint with its opposite-handed equivalent
+ if do_hflip:
+ if keypoint_hflip_indices is None:
+ raise ValueError("Cannot flip keypoints without providing flip indices!")
+ if len(keypoints) != len(keypoint_hflip_indices):
+ raise ValueError(
+ "Keypoint data has {} points, but metadata "
+ "contains {} points!".format(len(keypoints), len(keypoint_hflip_indices))
+ )
+ keypoints = keypoints[np.asarray(keypoint_hflip_indices, dtype=np.int32), :]
+
+ # Maintain COCO convention that if visibility == 0 (unlabeled), then x, y = 0
+ keypoints[keypoints[:, 2] == 0] = 0
+ return keypoints
+
+
+def annotations_to_instances(annos, image_size, mask_format="polygon"):
+ """
+ Create an :class:`Instances` object used by the models,
+ from instance annotations in the dataset dict.
+
+ Args:
+ annos (list[dict]): a list of instance annotations in one image, each
+ element for one instance.
+ image_size (tuple): height, width
+
+ Returns:
+ Instances:
+ It will contain fields "gt_boxes", "gt_classes",
+ "gt_masks", "gt_keypoints", if they can be obtained from `annos`.
+ This is the format that builtin models expect.
+ """
+ boxes = (
+ np.stack(
+ [BoxMode.convert(obj["bbox"], obj["bbox_mode"], BoxMode.XYXY_ABS) for obj in annos]
+ )
+ if len(annos)
+ else np.zeros((0, 4))
+ )
+ target = Instances(image_size)
+ target.gt_boxes = Boxes(boxes)
+
+ classes = [int(obj["category_id"]) for obj in annos]
+ classes = torch.tensor(classes, dtype=torch.int64)
+ target.gt_classes = classes
+
+ if len(annos) and "segmentation" in annos[0]:
+ segms = [obj["segmentation"] for obj in annos]
+ if mask_format == "polygon":
+ try:
+ masks = PolygonMasks(segms)
+ except ValueError as e:
+ raise ValueError(
+ "Failed to use mask_format=='polygon' from the given annotations!"
+ ) from e
+ else:
+ assert mask_format == "bitmask", mask_format
+ masks = []
+ for segm in segms:
+ if isinstance(segm, list):
+ # polygon
+ masks.append(polygons_to_bitmask(segm, *image_size))
+ elif isinstance(segm, dict):
+ # COCO RLE
+ masks.append(mask_util.decode(segm))
+ elif isinstance(segm, np.ndarray):
+ assert segm.ndim == 2, "Expect segmentation of 2 dimensions, got {}.".format(
+ segm.ndim
+ )
+ # mask array
+ masks.append(segm)
+ else:
+ raise ValueError(
+ "Cannot convert segmentation of type '{}' to BitMasks!"
+ "Supported types are: polygons as list[list[float] or ndarray],"
+ " COCO-style RLE as a dict, or a binary segmentation mask "
+ " in a 2D numpy array of shape HxW.".format(type(segm))
+ )
+ # torch.from_numpy does not support array with negative stride.
+ masks = BitMasks(
+ torch.stack([torch.from_numpy(np.ascontiguousarray(x)) for x in masks])
+ )
+ target.gt_masks = masks
+
+ if len(annos) and "keypoints" in annos[0]:
+ kpts = [obj.get("keypoints", []) for obj in annos]
+ target.gt_keypoints = Keypoints(kpts)
+
+ return target
+
+
+def annotations_to_instances_rotated(annos, image_size):
+ """
+ Create an :class:`Instances` object used by the models,
+ from instance annotations in the dataset dict.
+ Compared to `annotations_to_instances`, this function is for rotated boxes only
+
+ Args:
+ annos (list[dict]): a list of instance annotations in one image, each
+ element for one instance.
+ image_size (tuple): height, width
+
+ Returns:
+ Instances:
+ Containing fields "gt_boxes", "gt_classes",
+ if they can be obtained from `annos`.
+ This is the format that builtin models expect.
+ """
+ boxes = [obj["bbox"] for obj in annos]
+ target = Instances(image_size)
+ boxes = target.gt_boxes = RotatedBoxes(boxes)
+ boxes.clip(image_size)
+
+ classes = [obj["category_id"] for obj in annos]
+ classes = torch.tensor(classes, dtype=torch.int64)
+ target.gt_classes = classes
+
+ return target
+
+
+def filter_empty_instances(
+ instances, by_box=True, by_mask=True, box_threshold=1e-5, return_mask=False
+):
+ """
+ Filter out empty instances in an `Instances` object.
+
+ Args:
+ instances (Instances):
+ by_box (bool): whether to filter out instances with empty boxes
+ by_mask (bool): whether to filter out instances with empty masks
+ box_threshold (float): minimum width and height to be considered non-empty
+ return_mask (bool): whether to return boolean mask of filtered instances
+
+ Returns:
+ Instances: the filtered instances.
+ tensor[bool], optional: boolean mask of filtered instances
+ """
+ assert by_box or by_mask
+ r = []
+ if by_box:
+ r.append(instances.gt_boxes.nonempty(threshold=box_threshold))
+ if instances.has("gt_masks") and by_mask:
+ r.append(instances.gt_masks.nonempty())
+
+ # TODO: can also filter visible keypoints
+
+ if not r:
+ return instances
+ m = r[0]
+ for x in r[1:]:
+ m = m & x
+ if return_mask:
+ return instances[m], m
+ return instances[m]
+
+
+def create_keypoint_hflip_indices(dataset_names: Union[str, List[str]]) -> List[int]:
+ """
+ Args:
+ dataset_names: list of dataset names
+
+ Returns:
+ list[int]: a list of size=#keypoints, storing the
+ horizontally-flipped keypoint indices.
+ """
+ if isinstance(dataset_names, str):
+ dataset_names = [dataset_names]
+
+ check_metadata_consistency("keypoint_names", dataset_names)
+ check_metadata_consistency("keypoint_flip_map", dataset_names)
+
+ meta = MetadataCatalog.get(dataset_names[0])
+ names = meta.keypoint_names
+ # TODO flip -> hflip
+ flip_map = dict(meta.keypoint_flip_map)
+ flip_map.update({v: k for k, v in flip_map.items()})
+ flipped_names = [i if i not in flip_map else flip_map[i] for i in names]
+ flip_indices = [names.index(i) for i in flipped_names]
+ return flip_indices
+
+
+def get_fed_loss_cls_weights(dataset_names: Union[str, List[str]], freq_weight_power=1.0):
+ """
+ Get frequency weight for each class sorted by class id.
+ We now calcualte freqency weight using image_count to the power freq_weight_power.
+
+ Args:
+ dataset_names: list of dataset names
+ freq_weight_power: power value
+ """
+ if isinstance(dataset_names, str):
+ dataset_names = [dataset_names]
+
+ check_metadata_consistency("class_image_count", dataset_names)
+
+ meta = MetadataCatalog.get(dataset_names[0])
+ class_freq_meta = meta.class_image_count
+ class_freq = torch.tensor(
+ [c["image_count"] for c in sorted(class_freq_meta, key=lambda x: x["id"])]
+ )
+ class_freq_weight = class_freq.float() ** freq_weight_power
+ return class_freq_weight
+
+
+def gen_crop_transform_with_instance(crop_size, image_size, instance):
+ """
+ Generate a CropTransform so that the cropping region contains
+ the center of the given instance.
+
+ Args:
+ crop_size (tuple): h, w in pixels
+ image_size (tuple): h, w
+ instance (dict): an annotation dict of one instance, in Detectron2's
+ dataset format.
+ """
+ crop_size = np.asarray(crop_size, dtype=np.int32)
+ bbox = BoxMode.convert(instance["bbox"], instance["bbox_mode"], BoxMode.XYXY_ABS)
+ center_yx = (bbox[1] + bbox[3]) * 0.5, (bbox[0] + bbox[2]) * 0.5
+ assert (
+ image_size[0] >= center_yx[0] and image_size[1] >= center_yx[1]
+ ), "The annotation bounding box is outside of the image!"
+ assert (
+ image_size[0] >= crop_size[0] and image_size[1] >= crop_size[1]
+ ), "Crop size is larger than image size!"
+
+ min_yx = np.maximum(np.floor(center_yx).astype(np.int32) - crop_size, 0)
+ max_yx = np.maximum(np.asarray(image_size, dtype=np.int32) - crop_size, 0)
+ max_yx = np.minimum(max_yx, np.ceil(center_yx).astype(np.int32))
+
+ y0 = np.random.randint(min_yx[0], max_yx[0] + 1)
+ x0 = np.random.randint(min_yx[1], max_yx[1] + 1)
+ return T.CropTransform(x0, y0, crop_size[1], crop_size[0])
+
+
+def check_metadata_consistency(key, dataset_names):
+ """
+ Check that the datasets have consistent metadata.
+
+ Args:
+ key (str): a metadata key
+ dataset_names (list[str]): a list of dataset names
+
+ Raises:
+ AttributeError: if the key does not exist in the metadata
+ ValueError: if the given datasets do not have the same metadata values defined by key
+ """
+ if len(dataset_names) == 0:
+ return
+ logger = logging.getLogger(__name__)
+ entries_per_dataset = [getattr(MetadataCatalog.get(d), key) for d in dataset_names]
+ for idx, entry in enumerate(entries_per_dataset):
+ if entry != entries_per_dataset[0]:
+ logger.error(
+ "Metadata '{}' for dataset '{}' is '{}'".format(key, dataset_names[idx], str(entry))
+ )
+ logger.error(
+ "Metadata '{}' for dataset '{}' is '{}'".format(
+ key, dataset_names[0], str(entries_per_dataset[0])
+ )
+ )
+ raise ValueError("Datasets have different metadata '{}'!".format(key))
+
+
+def build_augmentation(cfg, is_train):
+ """
+ Create a list of default :class:`Augmentation` from config.
+ Now it includes resizing and flipping.
+
+ Returns:
+ list[Augmentation]
+ """
+ if is_train:
+ min_size = cfg.INPUT.MIN_SIZE_TRAIN
+ max_size = cfg.INPUT.MAX_SIZE_TRAIN
+ sample_style = cfg.INPUT.MIN_SIZE_TRAIN_SAMPLING
+ else:
+ min_size = cfg.INPUT.MIN_SIZE_TEST
+ max_size = cfg.INPUT.MAX_SIZE_TEST
+ sample_style = "choice"
+ augmentation = [T.ResizeShortestEdge(min_size, max_size, sample_style)]
+ if is_train and cfg.INPUT.RANDOM_FLIP != "none":
+ augmentation.append(
+ T.RandomFlip(
+ horizontal=cfg.INPUT.RANDOM_FLIP == "horizontal",
+ vertical=cfg.INPUT.RANDOM_FLIP == "vertical",
+ )
+ )
+ return augmentation
+
+
+build_transform_gen = build_augmentation
+"""
+Alias for backward-compatibility.
+"""
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/samplers/__init__.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/samplers/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..85c9f1a9df8a4038fbd4246239b699402e382309
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/samplers/__init__.py
@@ -0,0 +1,17 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+from .distributed_sampler import (
+ InferenceSampler,
+ RandomSubsetTrainingSampler,
+ RepeatFactorTrainingSampler,
+ TrainingSampler,
+)
+
+from .grouped_batch_sampler import GroupedBatchSampler
+
+__all__ = [
+ "GroupedBatchSampler",
+ "TrainingSampler",
+ "RandomSubsetTrainingSampler",
+ "InferenceSampler",
+ "RepeatFactorTrainingSampler",
+]
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/samplers/distributed_sampler.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/samplers/distributed_sampler.py
new file mode 100644
index 0000000000000000000000000000000000000000..cd4724eac8fbff2456bd26f95e6fea5e914b73e2
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/samplers/distributed_sampler.py
@@ -0,0 +1,278 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import itertools
+import logging
+import math
+from collections import defaultdict
+from typing import Optional
+import torch
+from torch.utils.data.sampler import Sampler
+
+from annotator.oneformer.detectron2.utils import comm
+
+logger = logging.getLogger(__name__)
+
+
+class TrainingSampler(Sampler):
+ """
+ In training, we only care about the "infinite stream" of training data.
+ So this sampler produces an infinite stream of indices and
+ all workers cooperate to correctly shuffle the indices and sample different indices.
+
+ The samplers in each worker effectively produces `indices[worker_id::num_workers]`
+ where `indices` is an infinite stream of indices consisting of
+ `shuffle(range(size)) + shuffle(range(size)) + ...` (if shuffle is True)
+ or `range(size) + range(size) + ...` (if shuffle is False)
+
+ Note that this sampler does not shard based on pytorch DataLoader worker id.
+ A sampler passed to pytorch DataLoader is used only with map-style dataset
+ and will not be executed inside workers.
+ But if this sampler is used in a way that it gets execute inside a dataloader
+ worker, then extra work needs to be done to shard its outputs based on worker id.
+ This is required so that workers don't produce identical data.
+ :class:`ToIterableDataset` implements this logic.
+ This note is true for all samplers in detectron2.
+ """
+
+ def __init__(self, size: int, shuffle: bool = True, seed: Optional[int] = None):
+ """
+ Args:
+ size (int): the total number of data of the underlying dataset to sample from
+ shuffle (bool): whether to shuffle the indices or not
+ seed (int): the initial seed of the shuffle. Must be the same
+ across all workers. If None, will use a random seed shared
+ among workers (require synchronization among all workers).
+ """
+ if not isinstance(size, int):
+ raise TypeError(f"TrainingSampler(size=) expects an int. Got type {type(size)}.")
+ if size <= 0:
+ raise ValueError(f"TrainingSampler(size=) expects a positive int. Got {size}.")
+ self._size = size
+ self._shuffle = shuffle
+ if seed is None:
+ seed = comm.shared_random_seed()
+ self._seed = int(seed)
+
+ self._rank = comm.get_rank()
+ self._world_size = comm.get_world_size()
+
+ def __iter__(self):
+ start = self._rank
+ yield from itertools.islice(self._infinite_indices(), start, None, self._world_size)
+
+ def _infinite_indices(self):
+ g = torch.Generator()
+ g.manual_seed(self._seed)
+ while True:
+ if self._shuffle:
+ yield from torch.randperm(self._size, generator=g).tolist()
+ else:
+ yield from torch.arange(self._size).tolist()
+
+
+class RandomSubsetTrainingSampler(TrainingSampler):
+ """
+ Similar to TrainingSampler, but only sample a random subset of indices.
+ This is useful when you want to estimate the accuracy vs data-number curves by
+ training the model with different subset_ratio.
+ """
+
+ def __init__(
+ self,
+ size: int,
+ subset_ratio: float,
+ shuffle: bool = True,
+ seed_shuffle: Optional[int] = None,
+ seed_subset: Optional[int] = None,
+ ):
+ """
+ Args:
+ size (int): the total number of data of the underlying dataset to sample from
+ subset_ratio (float): the ratio of subset data to sample from the underlying dataset
+ shuffle (bool): whether to shuffle the indices or not
+ seed_shuffle (int): the initial seed of the shuffle. Must be the same
+ across all workers. If None, will use a random seed shared
+ among workers (require synchronization among all workers).
+ seed_subset (int): the seed to randomize the subset to be sampled.
+ Must be the same across all workers. If None, will use a random seed shared
+ among workers (require synchronization among all workers).
+ """
+ super().__init__(size=size, shuffle=shuffle, seed=seed_shuffle)
+
+ assert 0.0 < subset_ratio <= 1.0
+ self._size_subset = int(size * subset_ratio)
+ assert self._size_subset > 0
+ if seed_subset is None:
+ seed_subset = comm.shared_random_seed()
+ self._seed_subset = int(seed_subset)
+
+ # randomly generate the subset indexes to be sampled from
+ g = torch.Generator()
+ g.manual_seed(self._seed_subset)
+ indexes_randperm = torch.randperm(self._size, generator=g)
+ self._indexes_subset = indexes_randperm[: self._size_subset]
+
+ logger.info("Using RandomSubsetTrainingSampler......")
+ logger.info(f"Randomly sample {self._size_subset} data from the original {self._size} data")
+
+ def _infinite_indices(self):
+ g = torch.Generator()
+ g.manual_seed(self._seed) # self._seed equals seed_shuffle from __init__()
+ while True:
+ if self._shuffle:
+ # generate a random permutation to shuffle self._indexes_subset
+ randperm = torch.randperm(self._size_subset, generator=g)
+ yield from self._indexes_subset[randperm].tolist()
+ else:
+ yield from self._indexes_subset.tolist()
+
+
+class RepeatFactorTrainingSampler(Sampler):
+ """
+ Similar to TrainingSampler, but a sample may appear more times than others based
+ on its "repeat factor". This is suitable for training on class imbalanced datasets like LVIS.
+ """
+
+ def __init__(self, repeat_factors, *, shuffle=True, seed=None):
+ """
+ Args:
+ repeat_factors (Tensor): a float vector, the repeat factor for each indice. When it's
+ full of ones, it is equivalent to ``TrainingSampler(len(repeat_factors), ...)``.
+ shuffle (bool): whether to shuffle the indices or not
+ seed (int): the initial seed of the shuffle. Must be the same
+ across all workers. If None, will use a random seed shared
+ among workers (require synchronization among all workers).
+ """
+ self._shuffle = shuffle
+ if seed is None:
+ seed = comm.shared_random_seed()
+ self._seed = int(seed)
+
+ self._rank = comm.get_rank()
+ self._world_size = comm.get_world_size()
+
+ # Split into whole number (_int_part) and fractional (_frac_part) parts.
+ self._int_part = torch.trunc(repeat_factors)
+ self._frac_part = repeat_factors - self._int_part
+
+ @staticmethod
+ def repeat_factors_from_category_frequency(dataset_dicts, repeat_thresh):
+ """
+ Compute (fractional) per-image repeat factors based on category frequency.
+ The repeat factor for an image is a function of the frequency of the rarest
+ category labeled in that image. The "frequency of category c" in [0, 1] is defined
+ as the fraction of images in the training set (without repeats) in which category c
+ appears.
+ See :paper:`lvis` (>= v2) Appendix B.2.
+
+ Args:
+ dataset_dicts (list[dict]): annotations in Detectron2 dataset format.
+ repeat_thresh (float): frequency threshold below which data is repeated.
+ If the frequency is half of `repeat_thresh`, the image will be
+ repeated twice.
+
+ Returns:
+ torch.Tensor:
+ the i-th element is the repeat factor for the dataset image at index i.
+ """
+ # 1. For each category c, compute the fraction of images that contain it: f(c)
+ category_freq = defaultdict(int)
+ for dataset_dict in dataset_dicts: # For each image (without repeats)
+ cat_ids = {ann["category_id"] for ann in dataset_dict["annotations"]}
+ for cat_id in cat_ids:
+ category_freq[cat_id] += 1
+ num_images = len(dataset_dicts)
+ for k, v in category_freq.items():
+ category_freq[k] = v / num_images
+
+ # 2. For each category c, compute the category-level repeat factor:
+ # r(c) = max(1, sqrt(t / f(c)))
+ category_rep = {
+ cat_id: max(1.0, math.sqrt(repeat_thresh / cat_freq))
+ for cat_id, cat_freq in category_freq.items()
+ }
+
+ # 3. For each image I, compute the image-level repeat factor:
+ # r(I) = max_{c in I} r(c)
+ rep_factors = []
+ for dataset_dict in dataset_dicts:
+ cat_ids = {ann["category_id"] for ann in dataset_dict["annotations"]}
+ rep_factor = max({category_rep[cat_id] for cat_id in cat_ids}, default=1.0)
+ rep_factors.append(rep_factor)
+
+ return torch.tensor(rep_factors, dtype=torch.float32)
+
+ def _get_epoch_indices(self, generator):
+ """
+ Create a list of dataset indices (with repeats) to use for one epoch.
+
+ Args:
+ generator (torch.Generator): pseudo random number generator used for
+ stochastic rounding.
+
+ Returns:
+ torch.Tensor: list of dataset indices to use in one epoch. Each index
+ is repeated based on its calculated repeat factor.
+ """
+ # Since repeat factors are fractional, we use stochastic rounding so
+ # that the target repeat factor is achieved in expectation over the
+ # course of training
+ rands = torch.rand(len(self._frac_part), generator=generator)
+ rep_factors = self._int_part + (rands < self._frac_part).float()
+ # Construct a list of indices in which we repeat images as specified
+ indices = []
+ for dataset_index, rep_factor in enumerate(rep_factors):
+ indices.extend([dataset_index] * int(rep_factor.item()))
+ return torch.tensor(indices, dtype=torch.int64)
+
+ def __iter__(self):
+ start = self._rank
+ yield from itertools.islice(self._infinite_indices(), start, None, self._world_size)
+
+ def _infinite_indices(self):
+ g = torch.Generator()
+ g.manual_seed(self._seed)
+ while True:
+ # Sample indices with repeats determined by stochastic rounding; each
+ # "epoch" may have a slightly different size due to the rounding.
+ indices = self._get_epoch_indices(g)
+ if self._shuffle:
+ randperm = torch.randperm(len(indices), generator=g)
+ yield from indices[randperm].tolist()
+ else:
+ yield from indices.tolist()
+
+
+class InferenceSampler(Sampler):
+ """
+ Produce indices for inference across all workers.
+ Inference needs to run on the __exact__ set of samples,
+ therefore when the total number of samples is not divisible by the number of workers,
+ this sampler produces different number of samples on different workers.
+ """
+
+ def __init__(self, size: int):
+ """
+ Args:
+ size (int): the total number of data of the underlying dataset to sample from
+ """
+ self._size = size
+ assert size > 0
+ self._rank = comm.get_rank()
+ self._world_size = comm.get_world_size()
+ self._local_indices = self._get_local_indices(size, self._world_size, self._rank)
+
+ @staticmethod
+ def _get_local_indices(total_size, world_size, rank):
+ shard_size = total_size // world_size
+ left = total_size % world_size
+ shard_sizes = [shard_size + int(r < left) for r in range(world_size)]
+
+ begin = sum(shard_sizes[:rank])
+ end = min(sum(shard_sizes[: rank + 1]), total_size)
+ return range(begin, end)
+
+ def __iter__(self):
+ yield from self._local_indices
+
+ def __len__(self):
+ return len(self._local_indices)
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/samplers/grouped_batch_sampler.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/samplers/grouped_batch_sampler.py
new file mode 100644
index 0000000000000000000000000000000000000000..5b247730aacd04dd0c752664acde3257c4eddd71
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/samplers/grouped_batch_sampler.py
@@ -0,0 +1,47 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import numpy as np
+from torch.utils.data.sampler import BatchSampler, Sampler
+
+
+class GroupedBatchSampler(BatchSampler):
+ """
+ Wraps another sampler to yield a mini-batch of indices.
+ It enforces that the batch only contain elements from the same group.
+ It also tries to provide mini-batches which follows an ordering which is
+ as close as possible to the ordering from the original sampler.
+ """
+
+ def __init__(self, sampler, group_ids, batch_size):
+ """
+ Args:
+ sampler (Sampler): Base sampler.
+ group_ids (list[int]): If the sampler produces indices in range [0, N),
+ `group_ids` must be a list of `N` ints which contains the group id of each sample.
+ The group ids must be a set of integers in the range [0, num_groups).
+ batch_size (int): Size of mini-batch.
+ """
+ if not isinstance(sampler, Sampler):
+ raise ValueError(
+ "sampler should be an instance of "
+ "torch.utils.data.Sampler, but got sampler={}".format(sampler)
+ )
+ self.sampler = sampler
+ self.group_ids = np.asarray(group_ids)
+ assert self.group_ids.ndim == 1
+ self.batch_size = batch_size
+ groups = np.unique(self.group_ids).tolist()
+
+ # buffer the indices of each group until batch size is reached
+ self.buffer_per_group = {k: [] for k in groups}
+
+ def __iter__(self):
+ for idx in self.sampler:
+ group_id = self.group_ids[idx]
+ group_buffer = self.buffer_per_group[group_id]
+ group_buffer.append(idx)
+ if len(group_buffer) == self.batch_size:
+ yield group_buffer[:] # yield a copy of the list
+ del group_buffer[:]
+
+ def __len__(self):
+ raise NotImplementedError("len() of GroupedBatchSampler is not well-defined.")
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/transforms/__init__.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/transforms/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..e91c6cdfacd6992a7a1e80c7d2e4b38b2cf7dcde
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/transforms/__init__.py
@@ -0,0 +1,14 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+from fvcore.transforms.transform import Transform, TransformList # order them first
+from fvcore.transforms.transform import *
+from .transform import *
+from .augmentation import *
+from .augmentation_impl import *
+
+__all__ = [k for k in globals().keys() if not k.startswith("_")]
+
+
+from annotator.oneformer.detectron2.utils.env import fixup_module_metadata
+
+fixup_module_metadata(__name__, globals(), __all__)
+del fixup_module_metadata
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/transforms/augmentation.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/transforms/augmentation.py
new file mode 100644
index 0000000000000000000000000000000000000000..63dd41aef658c9b51c7246880399405a029c5580
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/transforms/augmentation.py
@@ -0,0 +1,380 @@
+# -*- coding: utf-8 -*-
+# Copyright (c) Facebook, Inc. and its affiliates.
+
+import inspect
+import numpy as np
+import pprint
+from typing import Any, List, Optional, Tuple, Union
+from fvcore.transforms.transform import Transform, TransformList
+
+"""
+See "Data Augmentation" tutorial for an overview of the system:
+https://detectron2.readthedocs.io/tutorials/augmentation.html
+"""
+
+
+__all__ = [
+ "Augmentation",
+ "AugmentationList",
+ "AugInput",
+ "TransformGen",
+ "apply_transform_gens",
+ "StandardAugInput",
+ "apply_augmentations",
+]
+
+
+def _check_img_dtype(img):
+ assert isinstance(img, np.ndarray), "[Augmentation] Needs an numpy array, but got a {}!".format(
+ type(img)
+ )
+ assert not isinstance(img.dtype, np.integer) or (
+ img.dtype == np.uint8
+ ), "[Augmentation] Got image of type {}, use uint8 or floating points instead!".format(
+ img.dtype
+ )
+ assert img.ndim in [2, 3], img.ndim
+
+
+def _get_aug_input_args(aug, aug_input) -> List[Any]:
+ """
+ Get the arguments to be passed to ``aug.get_transform`` from the input ``aug_input``.
+ """
+ if aug.input_args is None:
+ # Decide what attributes are needed automatically
+ prms = list(inspect.signature(aug.get_transform).parameters.items())
+ # The default behavior is: if there is one parameter, then its "image"
+ # (work automatically for majority of use cases, and also avoid BC breaking),
+ # Otherwise, use the argument names.
+ if len(prms) == 1:
+ names = ("image",)
+ else:
+ names = []
+ for name, prm in prms:
+ if prm.kind in (
+ inspect.Parameter.VAR_POSITIONAL,
+ inspect.Parameter.VAR_KEYWORD,
+ ):
+ raise TypeError(
+ f""" \
+The default implementation of `{type(aug)}.__call__` does not allow \
+`{type(aug)}.get_transform` to use variable-length arguments (*args, **kwargs)! \
+If arguments are unknown, reimplement `__call__` instead. \
+"""
+ )
+ names.append(name)
+ aug.input_args = tuple(names)
+
+ args = []
+ for f in aug.input_args:
+ try:
+ args.append(getattr(aug_input, f))
+ except AttributeError as e:
+ raise AttributeError(
+ f"{type(aug)}.get_transform needs input attribute '{f}', "
+ f"but it is not an attribute of {type(aug_input)}!"
+ ) from e
+ return args
+
+
+class Augmentation:
+ """
+ Augmentation defines (often random) policies/strategies to generate :class:`Transform`
+ from data. It is often used for pre-processing of input data.
+
+ A "policy" that generates a :class:`Transform` may, in the most general case,
+ need arbitrary information from input data in order to determine what transforms
+ to apply. Therefore, each :class:`Augmentation` instance defines the arguments
+ needed by its :meth:`get_transform` method. When called with the positional arguments,
+ the :meth:`get_transform` method executes the policy.
+
+ Note that :class:`Augmentation` defines the policies to create a :class:`Transform`,
+ but not how to execute the actual transform operations to those data.
+ Its :meth:`__call__` method will use :meth:`AugInput.transform` to execute the transform.
+
+ The returned `Transform` object is meant to describe deterministic transformation, which means
+ it can be re-applied on associated data, e.g. the geometry of an image and its segmentation
+ masks need to be transformed together.
+ (If such re-application is not needed, then determinism is not a crucial requirement.)
+ """
+
+ input_args: Optional[Tuple[str]] = None
+ """
+ Stores the attribute names needed by :meth:`get_transform`, e.g. ``("image", "sem_seg")``.
+ By default, it is just a tuple of argument names in :meth:`self.get_transform`, which often only
+ contain "image". As long as the argument name convention is followed, there is no need for
+ users to touch this attribute.
+ """
+
+ def _init(self, params=None):
+ if params:
+ for k, v in params.items():
+ if k != "self" and not k.startswith("_"):
+ setattr(self, k, v)
+
+ def get_transform(self, *args) -> Transform:
+ """
+ Execute the policy based on input data, and decide what transform to apply to inputs.
+
+ Args:
+ args: Any fixed-length positional arguments. By default, the name of the arguments
+ should exist in the :class:`AugInput` to be used.
+
+ Returns:
+ Transform: Returns the deterministic transform to apply to the input.
+
+ Examples:
+ ::
+ class MyAug:
+ # if a policy needs to know both image and semantic segmentation
+ def get_transform(image, sem_seg) -> T.Transform:
+ pass
+ tfm: Transform = MyAug().get_transform(image, sem_seg)
+ new_image = tfm.apply_image(image)
+
+ Notes:
+ Users can freely use arbitrary new argument names in custom
+ :meth:`get_transform` method, as long as they are available in the
+ input data. In detectron2 we use the following convention:
+
+ * image: (H,W) or (H,W,C) ndarray of type uint8 in range [0, 255], or
+ floating point in range [0, 1] or [0, 255].
+ * boxes: (N,4) ndarray of float32. It represents the instance bounding boxes
+ of N instances. Each is in XYXY format in unit of absolute coordinates.
+ * sem_seg: (H,W) ndarray of type uint8. Each element is an integer label of pixel.
+
+ We do not specify convention for other types and do not include builtin
+ :class:`Augmentation` that uses other types in detectron2.
+ """
+ raise NotImplementedError
+
+ def __call__(self, aug_input) -> Transform:
+ """
+ Augment the given `aug_input` **in-place**, and return the transform that's used.
+
+ This method will be called to apply the augmentation. In most augmentation, it
+ is enough to use the default implementation, which calls :meth:`get_transform`
+ using the inputs. But a subclass can overwrite it to have more complicated logic.
+
+ Args:
+ aug_input (AugInput): an object that has attributes needed by this augmentation
+ (defined by ``self.get_transform``). Its ``transform`` method will be called
+ to in-place transform it.
+
+ Returns:
+ Transform: the transform that is applied on the input.
+ """
+ args = _get_aug_input_args(self, aug_input)
+ tfm = self.get_transform(*args)
+ assert isinstance(tfm, (Transform, TransformList)), (
+ f"{type(self)}.get_transform must return an instance of Transform! "
+ f"Got {type(tfm)} instead."
+ )
+ aug_input.transform(tfm)
+ return tfm
+
+ def _rand_range(self, low=1.0, high=None, size=None):
+ """
+ Uniform float random number between low and high.
+ """
+ if high is None:
+ low, high = 0, low
+ if size is None:
+ size = []
+ return np.random.uniform(low, high, size)
+
+ def __repr__(self):
+ """
+ Produce something like:
+ "MyAugmentation(field1={self.field1}, field2={self.field2})"
+ """
+ try:
+ sig = inspect.signature(self.__init__)
+ classname = type(self).__name__
+ argstr = []
+ for name, param in sig.parameters.items():
+ assert (
+ param.kind != param.VAR_POSITIONAL and param.kind != param.VAR_KEYWORD
+ ), "The default __repr__ doesn't support *args or **kwargs"
+ assert hasattr(self, name), (
+ "Attribute {} not found! "
+ "Default __repr__ only works if attributes match the constructor.".format(name)
+ )
+ attr = getattr(self, name)
+ default = param.default
+ if default is attr:
+ continue
+ attr_str = pprint.pformat(attr)
+ if "\n" in attr_str:
+ # don't show it if pformat decides to use >1 lines
+ attr_str = "..."
+ argstr.append("{}={}".format(name, attr_str))
+ return "{}({})".format(classname, ", ".join(argstr))
+ except AssertionError:
+ return super().__repr__()
+
+ __str__ = __repr__
+
+
+class _TransformToAug(Augmentation):
+ def __init__(self, tfm: Transform):
+ self.tfm = tfm
+
+ def get_transform(self, *args):
+ return self.tfm
+
+ def __repr__(self):
+ return repr(self.tfm)
+
+ __str__ = __repr__
+
+
+def _transform_to_aug(tfm_or_aug):
+ """
+ Wrap Transform into Augmentation.
+ Private, used internally to implement augmentations.
+ """
+ assert isinstance(tfm_or_aug, (Transform, Augmentation)), tfm_or_aug
+ if isinstance(tfm_or_aug, Augmentation):
+ return tfm_or_aug
+ else:
+ return _TransformToAug(tfm_or_aug)
+
+
+class AugmentationList(Augmentation):
+ """
+ Apply a sequence of augmentations.
+
+ It has ``__call__`` method to apply the augmentations.
+
+ Note that :meth:`get_transform` method is impossible (will throw error if called)
+ for :class:`AugmentationList`, because in order to apply a sequence of augmentations,
+ the kth augmentation must be applied first, to provide inputs needed by the (k+1)th
+ augmentation.
+ """
+
+ def __init__(self, augs):
+ """
+ Args:
+ augs (list[Augmentation or Transform]):
+ """
+ super().__init__()
+ self.augs = [_transform_to_aug(x) for x in augs]
+
+ def __call__(self, aug_input) -> TransformList:
+ tfms = []
+ for x in self.augs:
+ tfm = x(aug_input)
+ tfms.append(tfm)
+ return TransformList(tfms)
+
+ def __repr__(self):
+ msgs = [str(x) for x in self.augs]
+ return "AugmentationList[{}]".format(", ".join(msgs))
+
+ __str__ = __repr__
+
+
+class AugInput:
+ """
+ Input that can be used with :meth:`Augmentation.__call__`.
+ This is a standard implementation for the majority of use cases.
+ This class provides the standard attributes **"image", "boxes", "sem_seg"**
+ defined in :meth:`__init__` and they may be needed by different augmentations.
+ Most augmentation policies do not need attributes beyond these three.
+
+ After applying augmentations to these attributes (using :meth:`AugInput.transform`),
+ the returned transforms can then be used to transform other data structures that users have.
+
+ Examples:
+ ::
+ input = AugInput(image, boxes=boxes)
+ tfms = augmentation(input)
+ transformed_image = input.image
+ transformed_boxes = input.boxes
+ transformed_other_data = tfms.apply_other(other_data)
+
+ An extended project that works with new data types may implement augmentation policies
+ that need other inputs. An algorithm may need to transform inputs in a way different
+ from the standard approach defined in this class. In those rare situations, users can
+ implement a class similar to this class, that satify the following condition:
+
+ * The input must provide access to these data in the form of attribute access
+ (``getattr``). For example, if an :class:`Augmentation` to be applied needs "image"
+ and "sem_seg" arguments, its input must have the attribute "image" and "sem_seg".
+ * The input must have a ``transform(tfm: Transform) -> None`` method which
+ in-place transforms all its attributes.
+ """
+
+ # TODO maybe should support more builtin data types here
+ def __init__(
+ self,
+ image: np.ndarray,
+ *,
+ boxes: Optional[np.ndarray] = None,
+ sem_seg: Optional[np.ndarray] = None,
+ ):
+ """
+ Args:
+ image (ndarray): (H,W) or (H,W,C) ndarray of type uint8 in range [0, 255], or
+ floating point in range [0, 1] or [0, 255]. The meaning of C is up
+ to users.
+ boxes (ndarray or None): Nx4 float32 boxes in XYXY_ABS mode
+ sem_seg (ndarray or None): HxW uint8 semantic segmentation mask. Each element
+ is an integer label of pixel.
+ """
+ _check_img_dtype(image)
+ self.image = image
+ self.boxes = boxes
+ self.sem_seg = sem_seg
+
+ def transform(self, tfm: Transform) -> None:
+ """
+ In-place transform all attributes of this class.
+
+ By "in-place", it means after calling this method, accessing an attribute such
+ as ``self.image`` will return transformed data.
+ """
+ self.image = tfm.apply_image(self.image)
+ if self.boxes is not None:
+ self.boxes = tfm.apply_box(self.boxes)
+ if self.sem_seg is not None:
+ self.sem_seg = tfm.apply_segmentation(self.sem_seg)
+
+ def apply_augmentations(
+ self, augmentations: List[Union[Augmentation, Transform]]
+ ) -> TransformList:
+ """
+ Equivalent of ``AugmentationList(augmentations)(self)``
+ """
+ return AugmentationList(augmentations)(self)
+
+
+def apply_augmentations(augmentations: List[Union[Transform, Augmentation]], inputs):
+ """
+ Use ``T.AugmentationList(augmentations)(inputs)`` instead.
+ """
+ if isinstance(inputs, np.ndarray):
+ # handle the common case of image-only Augmentation, also for backward compatibility
+ image_only = True
+ inputs = AugInput(inputs)
+ else:
+ image_only = False
+ tfms = inputs.apply_augmentations(augmentations)
+ return inputs.image if image_only else inputs, tfms
+
+
+apply_transform_gens = apply_augmentations
+"""
+Alias for backward-compatibility.
+"""
+
+TransformGen = Augmentation
+"""
+Alias for Augmentation, since it is something that generates :class:`Transform`s
+"""
+
+StandardAugInput = AugInput
+"""
+Alias for compatibility. It's not worth the complexity to have two classes.
+"""
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/transforms/augmentation_impl.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/transforms/augmentation_impl.py
new file mode 100644
index 0000000000000000000000000000000000000000..965f0a947d7c3ff03b0990f1a645703d470227de
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/transforms/augmentation_impl.py
@@ -0,0 +1,736 @@
+# -*- coding: utf-8 -*-
+# Copyright (c) Facebook, Inc. and its affiliates.
+"""
+Implement many useful :class:`Augmentation`.
+"""
+import numpy as np
+import sys
+from numpy import random
+from typing import Tuple
+import torch
+from fvcore.transforms.transform import (
+ BlendTransform,
+ CropTransform,
+ HFlipTransform,
+ NoOpTransform,
+ PadTransform,
+ Transform,
+ TransformList,
+ VFlipTransform,
+)
+from PIL import Image
+
+from annotator.oneformer.detectron2.structures import Boxes, pairwise_iou
+
+from .augmentation import Augmentation, _transform_to_aug
+from .transform import ExtentTransform, ResizeTransform, RotationTransform
+
+__all__ = [
+ "FixedSizeCrop",
+ "RandomApply",
+ "RandomBrightness",
+ "RandomContrast",
+ "RandomCrop",
+ "RandomExtent",
+ "RandomFlip",
+ "RandomSaturation",
+ "RandomLighting",
+ "RandomRotation",
+ "Resize",
+ "ResizeScale",
+ "ResizeShortestEdge",
+ "RandomCrop_CategoryAreaConstraint",
+ "RandomResize",
+ "MinIoURandomCrop",
+]
+
+
+class RandomApply(Augmentation):
+ """
+ Randomly apply an augmentation with a given probability.
+ """
+
+ def __init__(self, tfm_or_aug, prob=0.5):
+ """
+ Args:
+ tfm_or_aug (Transform, Augmentation): the transform or augmentation
+ to be applied. It can either be a `Transform` or `Augmentation`
+ instance.
+ prob (float): probability between 0.0 and 1.0 that
+ the wrapper transformation is applied
+ """
+ super().__init__()
+ self.aug = _transform_to_aug(tfm_or_aug)
+ assert 0.0 <= prob <= 1.0, f"Probablity must be between 0.0 and 1.0 (given: {prob})"
+ self.prob = prob
+
+ def get_transform(self, *args):
+ do = self._rand_range() < self.prob
+ if do:
+ return self.aug.get_transform(*args)
+ else:
+ return NoOpTransform()
+
+ def __call__(self, aug_input):
+ do = self._rand_range() < self.prob
+ if do:
+ return self.aug(aug_input)
+ else:
+ return NoOpTransform()
+
+
+class RandomFlip(Augmentation):
+ """
+ Flip the image horizontally or vertically with the given probability.
+ """
+
+ def __init__(self, prob=0.5, *, horizontal=True, vertical=False):
+ """
+ Args:
+ prob (float): probability of flip.
+ horizontal (boolean): whether to apply horizontal flipping
+ vertical (boolean): whether to apply vertical flipping
+ """
+ super().__init__()
+
+ if horizontal and vertical:
+ raise ValueError("Cannot do both horiz and vert. Please use two Flip instead.")
+ if not horizontal and not vertical:
+ raise ValueError("At least one of horiz or vert has to be True!")
+ self._init(locals())
+
+ def get_transform(self, image):
+ h, w = image.shape[:2]
+ do = self._rand_range() < self.prob
+ if do:
+ if self.horizontal:
+ return HFlipTransform(w)
+ elif self.vertical:
+ return VFlipTransform(h)
+ else:
+ return NoOpTransform()
+
+
+class Resize(Augmentation):
+ """Resize image to a fixed target size"""
+
+ def __init__(self, shape, interp=Image.BILINEAR):
+ """
+ Args:
+ shape: (h, w) tuple or a int
+ interp: PIL interpolation method
+ """
+ if isinstance(shape, int):
+ shape = (shape, shape)
+ shape = tuple(shape)
+ self._init(locals())
+
+ def get_transform(self, image):
+ return ResizeTransform(
+ image.shape[0], image.shape[1], self.shape[0], self.shape[1], self.interp
+ )
+
+
+class ResizeShortestEdge(Augmentation):
+ """
+ Resize the image while keeping the aspect ratio unchanged.
+ It attempts to scale the shorter edge to the given `short_edge_length`,
+ as long as the longer edge does not exceed `max_size`.
+ If `max_size` is reached, then downscale so that the longer edge does not exceed max_size.
+ """
+
+ @torch.jit.unused
+ def __init__(
+ self, short_edge_length, max_size=sys.maxsize, sample_style="range", interp=Image.BILINEAR
+ ):
+ """
+ Args:
+ short_edge_length (list[int]): If ``sample_style=="range"``,
+ a [min, max] interval from which to sample the shortest edge length.
+ If ``sample_style=="choice"``, a list of shortest edge lengths to sample from.
+ max_size (int): maximum allowed longest edge length.
+ sample_style (str): either "range" or "choice".
+ """
+ super().__init__()
+ assert sample_style in ["range", "choice"], sample_style
+
+ self.is_range = sample_style == "range"
+ if isinstance(short_edge_length, int):
+ short_edge_length = (short_edge_length, short_edge_length)
+ if self.is_range:
+ assert len(short_edge_length) == 2, (
+ "short_edge_length must be two values using 'range' sample style."
+ f" Got {short_edge_length}!"
+ )
+ self._init(locals())
+
+ @torch.jit.unused
+ def get_transform(self, image):
+ h, w = image.shape[:2]
+ if self.is_range:
+ size = np.random.randint(self.short_edge_length[0], self.short_edge_length[1] + 1)
+ else:
+ size = np.random.choice(self.short_edge_length)
+ if size == 0:
+ return NoOpTransform()
+
+ newh, neww = ResizeShortestEdge.get_output_shape(h, w, size, self.max_size)
+ return ResizeTransform(h, w, newh, neww, self.interp)
+
+ @staticmethod
+ def get_output_shape(
+ oldh: int, oldw: int, short_edge_length: int, max_size: int
+ ) -> Tuple[int, int]:
+ """
+ Compute the output size given input size and target short edge length.
+ """
+ h, w = oldh, oldw
+ size = short_edge_length * 1.0
+ scale = size / min(h, w)
+ if h < w:
+ newh, neww = size, scale * w
+ else:
+ newh, neww = scale * h, size
+ if max(newh, neww) > max_size:
+ scale = max_size * 1.0 / max(newh, neww)
+ newh = newh * scale
+ neww = neww * scale
+ neww = int(neww + 0.5)
+ newh = int(newh + 0.5)
+ return (newh, neww)
+
+
+class ResizeScale(Augmentation):
+ """
+ Takes target size as input and randomly scales the given target size between `min_scale`
+ and `max_scale`. It then scales the input image such that it fits inside the scaled target
+ box, keeping the aspect ratio constant.
+ This implements the resize part of the Google's 'resize_and_crop' data augmentation:
+ https://github.com/tensorflow/tpu/blob/master/models/official/detection/utils/input_utils.py#L127
+ """
+
+ def __init__(
+ self,
+ min_scale: float,
+ max_scale: float,
+ target_height: int,
+ target_width: int,
+ interp: int = Image.BILINEAR,
+ ):
+ """
+ Args:
+ min_scale: minimum image scale range.
+ max_scale: maximum image scale range.
+ target_height: target image height.
+ target_width: target image width.
+ interp: image interpolation method.
+ """
+ super().__init__()
+ self._init(locals())
+
+ def _get_resize(self, image: np.ndarray, scale: float) -> Transform:
+ input_size = image.shape[:2]
+
+ # Compute new target size given a scale.
+ target_size = (self.target_height, self.target_width)
+ target_scale_size = np.multiply(target_size, scale)
+
+ # Compute actual rescaling applied to input image and output size.
+ output_scale = np.minimum(
+ target_scale_size[0] / input_size[0], target_scale_size[1] / input_size[1]
+ )
+ output_size = np.round(np.multiply(input_size, output_scale)).astype(int)
+
+ return ResizeTransform(
+ input_size[0], input_size[1], output_size[0], output_size[1], self.interp
+ )
+
+ def get_transform(self, image: np.ndarray) -> Transform:
+ random_scale = np.random.uniform(self.min_scale, self.max_scale)
+ return self._get_resize(image, random_scale)
+
+
+class RandomRotation(Augmentation):
+ """
+ This method returns a copy of this image, rotated the given
+ number of degrees counter clockwise around the given center.
+ """
+
+ def __init__(self, angle, expand=True, center=None, sample_style="range", interp=None):
+ """
+ Args:
+ angle (list[float]): If ``sample_style=="range"``,
+ a [min, max] interval from which to sample the angle (in degrees).
+ If ``sample_style=="choice"``, a list of angles to sample from
+ expand (bool): choose if the image should be resized to fit the whole
+ rotated image (default), or simply cropped
+ center (list[[float, float]]): If ``sample_style=="range"``,
+ a [[minx, miny], [maxx, maxy]] relative interval from which to sample the center,
+ [0, 0] being the top left of the image and [1, 1] the bottom right.
+ If ``sample_style=="choice"``, a list of centers to sample from
+ Default: None, which means that the center of rotation is the center of the image
+ center has no effect if expand=True because it only affects shifting
+ """
+ super().__init__()
+ assert sample_style in ["range", "choice"], sample_style
+ self.is_range = sample_style == "range"
+ if isinstance(angle, (float, int)):
+ angle = (angle, angle)
+ if center is not None and isinstance(center[0], (float, int)):
+ center = (center, center)
+ self._init(locals())
+
+ def get_transform(self, image):
+ h, w = image.shape[:2]
+ center = None
+ if self.is_range:
+ angle = np.random.uniform(self.angle[0], self.angle[1])
+ if self.center is not None:
+ center = (
+ np.random.uniform(self.center[0][0], self.center[1][0]),
+ np.random.uniform(self.center[0][1], self.center[1][1]),
+ )
+ else:
+ angle = np.random.choice(self.angle)
+ if self.center is not None:
+ center = np.random.choice(self.center)
+
+ if center is not None:
+ center = (w * center[0], h * center[1]) # Convert to absolute coordinates
+
+ if angle % 360 == 0:
+ return NoOpTransform()
+
+ return RotationTransform(h, w, angle, expand=self.expand, center=center, interp=self.interp)
+
+
+class FixedSizeCrop(Augmentation):
+ """
+ If `crop_size` is smaller than the input image size, then it uses a random crop of
+ the crop size. If `crop_size` is larger than the input image size, then it pads
+ the right and the bottom of the image to the crop size if `pad` is True, otherwise
+ it returns the smaller image.
+ """
+
+ def __init__(
+ self,
+ crop_size: Tuple[int],
+ pad: bool = True,
+ pad_value: float = 128.0,
+ seg_pad_value: int = 255,
+ ):
+ """
+ Args:
+ crop_size: target image (height, width).
+ pad: if True, will pad images smaller than `crop_size` up to `crop_size`
+ pad_value: the padding value to the image.
+ seg_pad_value: the padding value to the segmentation mask.
+ """
+ super().__init__()
+ self._init(locals())
+
+ def _get_crop(self, image: np.ndarray) -> Transform:
+ # Compute the image scale and scaled size.
+ input_size = image.shape[:2]
+ output_size = self.crop_size
+
+ # Add random crop if the image is scaled up.
+ max_offset = np.subtract(input_size, output_size)
+ max_offset = np.maximum(max_offset, 0)
+ offset = np.multiply(max_offset, np.random.uniform(0.0, 1.0))
+ offset = np.round(offset).astype(int)
+ return CropTransform(
+ offset[1], offset[0], output_size[1], output_size[0], input_size[1], input_size[0]
+ )
+
+ def _get_pad(self, image: np.ndarray) -> Transform:
+ # Compute the image scale and scaled size.
+ input_size = image.shape[:2]
+ output_size = self.crop_size
+
+ # Add padding if the image is scaled down.
+ pad_size = np.subtract(output_size, input_size)
+ pad_size = np.maximum(pad_size, 0)
+ original_size = np.minimum(input_size, output_size)
+ return PadTransform(
+ 0,
+ 0,
+ pad_size[1],
+ pad_size[0],
+ original_size[1],
+ original_size[0],
+ self.pad_value,
+ self.seg_pad_value,
+ )
+
+ def get_transform(self, image: np.ndarray) -> TransformList:
+ transforms = [self._get_crop(image)]
+ if self.pad:
+ transforms.append(self._get_pad(image))
+ return TransformList(transforms)
+
+
+class RandomCrop(Augmentation):
+ """
+ Randomly crop a rectangle region out of an image.
+ """
+
+ def __init__(self, crop_type: str, crop_size):
+ """
+ Args:
+ crop_type (str): one of "relative_range", "relative", "absolute", "absolute_range".
+ crop_size (tuple[float, float]): two floats, explained below.
+
+ - "relative": crop a (H * crop_size[0], W * crop_size[1]) region from an input image of
+ size (H, W). crop size should be in (0, 1]
+ - "relative_range": uniformly sample two values from [crop_size[0], 1]
+ and [crop_size[1]], 1], and use them as in "relative" crop type.
+ - "absolute" crop a (crop_size[0], crop_size[1]) region from input image.
+ crop_size must be smaller than the input image size.
+ - "absolute_range", for an input of size (H, W), uniformly sample H_crop in
+ [crop_size[0], min(H, crop_size[1])] and W_crop in [crop_size[0], min(W, crop_size[1])].
+ Then crop a region (H_crop, W_crop).
+ """
+ # TODO style of relative_range and absolute_range are not consistent:
+ # one takes (h, w) but another takes (min, max)
+ super().__init__()
+ assert crop_type in ["relative_range", "relative", "absolute", "absolute_range"]
+ self._init(locals())
+
+ def get_transform(self, image):
+ h, w = image.shape[:2]
+ croph, cropw = self.get_crop_size((h, w))
+ assert h >= croph and w >= cropw, "Shape computation in {} has bugs.".format(self)
+ h0 = np.random.randint(h - croph + 1)
+ w0 = np.random.randint(w - cropw + 1)
+ return CropTransform(w0, h0, cropw, croph)
+
+ def get_crop_size(self, image_size):
+ """
+ Args:
+ image_size (tuple): height, width
+
+ Returns:
+ crop_size (tuple): height, width in absolute pixels
+ """
+ h, w = image_size
+ if self.crop_type == "relative":
+ ch, cw = self.crop_size
+ return int(h * ch + 0.5), int(w * cw + 0.5)
+ elif self.crop_type == "relative_range":
+ crop_size = np.asarray(self.crop_size, dtype=np.float32)
+ ch, cw = crop_size + np.random.rand(2) * (1 - crop_size)
+ return int(h * ch + 0.5), int(w * cw + 0.5)
+ elif self.crop_type == "absolute":
+ return (min(self.crop_size[0], h), min(self.crop_size[1], w))
+ elif self.crop_type == "absolute_range":
+ assert self.crop_size[0] <= self.crop_size[1]
+ ch = np.random.randint(min(h, self.crop_size[0]), min(h, self.crop_size[1]) + 1)
+ cw = np.random.randint(min(w, self.crop_size[0]), min(w, self.crop_size[1]) + 1)
+ return ch, cw
+ else:
+ raise NotImplementedError("Unknown crop type {}".format(self.crop_type))
+
+
+class RandomCrop_CategoryAreaConstraint(Augmentation):
+ """
+ Similar to :class:`RandomCrop`, but find a cropping window such that no single category
+ occupies a ratio of more than `single_category_max_area` in semantic segmentation ground
+ truth, which can cause unstability in training. The function attempts to find such a valid
+ cropping window for at most 10 times.
+ """
+
+ def __init__(
+ self,
+ crop_type: str,
+ crop_size,
+ single_category_max_area: float = 1.0,
+ ignored_category: int = None,
+ ):
+ """
+ Args:
+ crop_type, crop_size: same as in :class:`RandomCrop`
+ single_category_max_area: the maximum allowed area ratio of a
+ category. Set to 1.0 to disable
+ ignored_category: allow this category in the semantic segmentation
+ ground truth to exceed the area ratio. Usually set to the category
+ that's ignored in training.
+ """
+ self.crop_aug = RandomCrop(crop_type, crop_size)
+ self._init(locals())
+
+ def get_transform(self, image, sem_seg):
+ if self.single_category_max_area >= 1.0:
+ return self.crop_aug.get_transform(image)
+ else:
+ h, w = sem_seg.shape
+ for _ in range(10):
+ crop_size = self.crop_aug.get_crop_size((h, w))
+ y0 = np.random.randint(h - crop_size[0] + 1)
+ x0 = np.random.randint(w - crop_size[1] + 1)
+ sem_seg_temp = sem_seg[y0 : y0 + crop_size[0], x0 : x0 + crop_size[1]]
+ labels, cnt = np.unique(sem_seg_temp, return_counts=True)
+ if self.ignored_category is not None:
+ cnt = cnt[labels != self.ignored_category]
+ if len(cnt) > 1 and np.max(cnt) < np.sum(cnt) * self.single_category_max_area:
+ break
+ crop_tfm = CropTransform(x0, y0, crop_size[1], crop_size[0])
+ return crop_tfm
+
+
+class RandomExtent(Augmentation):
+ """
+ Outputs an image by cropping a random "subrect" of the source image.
+
+ The subrect can be parameterized to include pixels outside the source image,
+ in which case they will be set to zeros (i.e. black). The size of the output
+ image will vary with the size of the random subrect.
+ """
+
+ def __init__(self, scale_range, shift_range):
+ """
+ Args:
+ output_size (h, w): Dimensions of output image
+ scale_range (l, h): Range of input-to-output size scaling factor
+ shift_range (x, y): Range of shifts of the cropped subrect. The rect
+ is shifted by [w / 2 * Uniform(-x, x), h / 2 * Uniform(-y, y)],
+ where (w, h) is the (width, height) of the input image. Set each
+ component to zero to crop at the image's center.
+ """
+ super().__init__()
+ self._init(locals())
+
+ def get_transform(self, image):
+ img_h, img_w = image.shape[:2]
+
+ # Initialize src_rect to fit the input image.
+ src_rect = np.array([-0.5 * img_w, -0.5 * img_h, 0.5 * img_w, 0.5 * img_h])
+
+ # Apply a random scaling to the src_rect.
+ src_rect *= np.random.uniform(self.scale_range[0], self.scale_range[1])
+
+ # Apply a random shift to the coordinates origin.
+ src_rect[0::2] += self.shift_range[0] * img_w * (np.random.rand() - 0.5)
+ src_rect[1::2] += self.shift_range[1] * img_h * (np.random.rand() - 0.5)
+
+ # Map src_rect coordinates into image coordinates (center at corner).
+ src_rect[0::2] += 0.5 * img_w
+ src_rect[1::2] += 0.5 * img_h
+
+ return ExtentTransform(
+ src_rect=(src_rect[0], src_rect[1], src_rect[2], src_rect[3]),
+ output_size=(int(src_rect[3] - src_rect[1]), int(src_rect[2] - src_rect[0])),
+ )
+
+
+class RandomContrast(Augmentation):
+ """
+ Randomly transforms image contrast.
+
+ Contrast intensity is uniformly sampled in (intensity_min, intensity_max).
+ - intensity < 1 will reduce contrast
+ - intensity = 1 will preserve the input image
+ - intensity > 1 will increase contrast
+
+ See: https://pillow.readthedocs.io/en/3.0.x/reference/ImageEnhance.html
+ """
+
+ def __init__(self, intensity_min, intensity_max):
+ """
+ Args:
+ intensity_min (float): Minimum augmentation
+ intensity_max (float): Maximum augmentation
+ """
+ super().__init__()
+ self._init(locals())
+
+ def get_transform(self, image):
+ w = np.random.uniform(self.intensity_min, self.intensity_max)
+ return BlendTransform(src_image=image.mean(), src_weight=1 - w, dst_weight=w)
+
+
+class RandomBrightness(Augmentation):
+ """
+ Randomly transforms image brightness.
+
+ Brightness intensity is uniformly sampled in (intensity_min, intensity_max).
+ - intensity < 1 will reduce brightness
+ - intensity = 1 will preserve the input image
+ - intensity > 1 will increase brightness
+
+ See: https://pillow.readthedocs.io/en/3.0.x/reference/ImageEnhance.html
+ """
+
+ def __init__(self, intensity_min, intensity_max):
+ """
+ Args:
+ intensity_min (float): Minimum augmentation
+ intensity_max (float): Maximum augmentation
+ """
+ super().__init__()
+ self._init(locals())
+
+ def get_transform(self, image):
+ w = np.random.uniform(self.intensity_min, self.intensity_max)
+ return BlendTransform(src_image=0, src_weight=1 - w, dst_weight=w)
+
+
+class RandomSaturation(Augmentation):
+ """
+ Randomly transforms saturation of an RGB image.
+ Input images are assumed to have 'RGB' channel order.
+
+ Saturation intensity is uniformly sampled in (intensity_min, intensity_max).
+ - intensity < 1 will reduce saturation (make the image more grayscale)
+ - intensity = 1 will preserve the input image
+ - intensity > 1 will increase saturation
+
+ See: https://pillow.readthedocs.io/en/3.0.x/reference/ImageEnhance.html
+ """
+
+ def __init__(self, intensity_min, intensity_max):
+ """
+ Args:
+ intensity_min (float): Minimum augmentation (1 preserves input).
+ intensity_max (float): Maximum augmentation (1 preserves input).
+ """
+ super().__init__()
+ self._init(locals())
+
+ def get_transform(self, image):
+ assert image.shape[-1] == 3, "RandomSaturation only works on RGB images"
+ w = np.random.uniform(self.intensity_min, self.intensity_max)
+ grayscale = image.dot([0.299, 0.587, 0.114])[:, :, np.newaxis]
+ return BlendTransform(src_image=grayscale, src_weight=1 - w, dst_weight=w)
+
+
+class RandomLighting(Augmentation):
+ """
+ The "lighting" augmentation described in AlexNet, using fixed PCA over ImageNet.
+ Input images are assumed to have 'RGB' channel order.
+
+ The degree of color jittering is randomly sampled via a normal distribution,
+ with standard deviation given by the scale parameter.
+ """
+
+ def __init__(self, scale):
+ """
+ Args:
+ scale (float): Standard deviation of principal component weighting.
+ """
+ super().__init__()
+ self._init(locals())
+ self.eigen_vecs = np.array(
+ [[-0.5675, 0.7192, 0.4009], [-0.5808, -0.0045, -0.8140], [-0.5836, -0.6948, 0.4203]]
+ )
+ self.eigen_vals = np.array([0.2175, 0.0188, 0.0045])
+
+ def get_transform(self, image):
+ assert image.shape[-1] == 3, "RandomLighting only works on RGB images"
+ weights = np.random.normal(scale=self.scale, size=3)
+ return BlendTransform(
+ src_image=self.eigen_vecs.dot(weights * self.eigen_vals), src_weight=1.0, dst_weight=1.0
+ )
+
+
+class RandomResize(Augmentation):
+ """Randomly resize image to a target size in shape_list"""
+
+ def __init__(self, shape_list, interp=Image.BILINEAR):
+ """
+ Args:
+ shape_list: a list of shapes in (h, w)
+ interp: PIL interpolation method
+ """
+ self.shape_list = shape_list
+ self._init(locals())
+
+ def get_transform(self, image):
+ shape_idx = np.random.randint(low=0, high=len(self.shape_list))
+ h, w = self.shape_list[shape_idx]
+ return ResizeTransform(image.shape[0], image.shape[1], h, w, self.interp)
+
+
+class MinIoURandomCrop(Augmentation):
+ """Random crop the image & bboxes, the cropped patches have minimum IoU
+ requirement with original image & bboxes, the IoU threshold is randomly
+ selected from min_ious.
+
+ Args:
+ min_ious (tuple): minimum IoU threshold for all intersections with
+ bounding boxes
+ min_crop_size (float): minimum crop's size (i.e. h,w := a*h, a*w,
+ where a >= min_crop_size)
+ mode_trials: number of trials for sampling min_ious threshold
+ crop_trials: number of trials for sampling crop_size after cropping
+ """
+
+ def __init__(
+ self,
+ min_ious=(0.1, 0.3, 0.5, 0.7, 0.9),
+ min_crop_size=0.3,
+ mode_trials=1000,
+ crop_trials=50,
+ ):
+ self.min_ious = min_ious
+ self.sample_mode = (1, *min_ious, 0)
+ self.min_crop_size = min_crop_size
+ self.mode_trials = mode_trials
+ self.crop_trials = crop_trials
+
+ def get_transform(self, image, boxes):
+ """Call function to crop images and bounding boxes with minimum IoU
+ constraint.
+
+ Args:
+ boxes: ground truth boxes in (x1, y1, x2, y2) format
+ """
+ if boxes is None:
+ return NoOpTransform()
+ h, w, c = image.shape
+ for _ in range(self.mode_trials):
+ mode = random.choice(self.sample_mode)
+ self.mode = mode
+ if mode == 1:
+ return NoOpTransform()
+
+ min_iou = mode
+ for _ in range(self.crop_trials):
+ new_w = random.uniform(self.min_crop_size * w, w)
+ new_h = random.uniform(self.min_crop_size * h, h)
+
+ # h / w in [0.5, 2]
+ if new_h / new_w < 0.5 or new_h / new_w > 2:
+ continue
+
+ left = random.uniform(w - new_w)
+ top = random.uniform(h - new_h)
+
+ patch = np.array((int(left), int(top), int(left + new_w), int(top + new_h)))
+ # Line or point crop is not allowed
+ if patch[2] == patch[0] or patch[3] == patch[1]:
+ continue
+ overlaps = pairwise_iou(
+ Boxes(patch.reshape(-1, 4)), Boxes(boxes.reshape(-1, 4))
+ ).reshape(-1)
+ if len(overlaps) > 0 and overlaps.min() < min_iou:
+ continue
+
+ # center of boxes should inside the crop img
+ # only adjust boxes and instance masks when the gt is not empty
+ if len(overlaps) > 0:
+ # adjust boxes
+ def is_center_of_bboxes_in_patch(boxes, patch):
+ center = (boxes[:, :2] + boxes[:, 2:]) / 2
+ mask = (
+ (center[:, 0] > patch[0])
+ * (center[:, 1] > patch[1])
+ * (center[:, 0] < patch[2])
+ * (center[:, 1] < patch[3])
+ )
+ return mask
+
+ mask = is_center_of_bboxes_in_patch(boxes, patch)
+ if not mask.any():
+ continue
+ return CropTransform(int(left), int(top), int(new_w), int(new_h))
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/transforms/transform.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/transforms/transform.py
new file mode 100644
index 0000000000000000000000000000000000000000..de44b991d7ab0d920ffb769e1402f08e358d37f7
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/data/transforms/transform.py
@@ -0,0 +1,351 @@
+# -*- coding: utf-8 -*-
+# Copyright (c) Facebook, Inc. and its affiliates.
+
+"""
+See "Data Augmentation" tutorial for an overview of the system:
+https://detectron2.readthedocs.io/tutorials/augmentation.html
+"""
+
+import numpy as np
+import torch
+import torch.nn.functional as F
+from fvcore.transforms.transform import (
+ CropTransform,
+ HFlipTransform,
+ NoOpTransform,
+ Transform,
+ TransformList,
+)
+from PIL import Image
+
+try:
+ import cv2 # noqa
+except ImportError:
+ # OpenCV is an optional dependency at the moment
+ pass
+
+__all__ = [
+ "ExtentTransform",
+ "ResizeTransform",
+ "RotationTransform",
+ "ColorTransform",
+ "PILColorTransform",
+]
+
+
+class ExtentTransform(Transform):
+ """
+ Extracts a subregion from the source image and scales it to the output size.
+
+ The fill color is used to map pixels from the source rect that fall outside
+ the source image.
+
+ See: https://pillow.readthedocs.io/en/latest/PIL.html#PIL.ImageTransform.ExtentTransform
+ """
+
+ def __init__(self, src_rect, output_size, interp=Image.LINEAR, fill=0):
+ """
+ Args:
+ src_rect (x0, y0, x1, y1): src coordinates
+ output_size (h, w): dst image size
+ interp: PIL interpolation methods
+ fill: Fill color used when src_rect extends outside image
+ """
+ super().__init__()
+ self._set_attributes(locals())
+
+ def apply_image(self, img, interp=None):
+ h, w = self.output_size
+ if len(img.shape) > 2 and img.shape[2] == 1:
+ pil_image = Image.fromarray(img[:, :, 0], mode="L")
+ else:
+ pil_image = Image.fromarray(img)
+ pil_image = pil_image.transform(
+ size=(w, h),
+ method=Image.EXTENT,
+ data=self.src_rect,
+ resample=interp if interp else self.interp,
+ fill=self.fill,
+ )
+ ret = np.asarray(pil_image)
+ if len(img.shape) > 2 and img.shape[2] == 1:
+ ret = np.expand_dims(ret, -1)
+ return ret
+
+ def apply_coords(self, coords):
+ # Transform image center from source coordinates into output coordinates
+ # and then map the new origin to the corner of the output image.
+ h, w = self.output_size
+ x0, y0, x1, y1 = self.src_rect
+ new_coords = coords.astype(np.float32)
+ new_coords[:, 0] -= 0.5 * (x0 + x1)
+ new_coords[:, 1] -= 0.5 * (y0 + y1)
+ new_coords[:, 0] *= w / (x1 - x0)
+ new_coords[:, 1] *= h / (y1 - y0)
+ new_coords[:, 0] += 0.5 * w
+ new_coords[:, 1] += 0.5 * h
+ return new_coords
+
+ def apply_segmentation(self, segmentation):
+ segmentation = self.apply_image(segmentation, interp=Image.NEAREST)
+ return segmentation
+
+
+class ResizeTransform(Transform):
+ """
+ Resize the image to a target size.
+ """
+
+ def __init__(self, h, w, new_h, new_w, interp=None):
+ """
+ Args:
+ h, w (int): original image size
+ new_h, new_w (int): new image size
+ interp: PIL interpolation methods, defaults to bilinear.
+ """
+ # TODO decide on PIL vs opencv
+ super().__init__()
+ if interp is None:
+ interp = Image.BILINEAR
+ self._set_attributes(locals())
+
+ def apply_image(self, img, interp=None):
+ assert img.shape[:2] == (self.h, self.w)
+ assert len(img.shape) <= 4
+ interp_method = interp if interp is not None else self.interp
+
+ if img.dtype == np.uint8:
+ if len(img.shape) > 2 and img.shape[2] == 1:
+ pil_image = Image.fromarray(img[:, :, 0], mode="L")
+ else:
+ pil_image = Image.fromarray(img)
+ pil_image = pil_image.resize((self.new_w, self.new_h), interp_method)
+ ret = np.asarray(pil_image)
+ if len(img.shape) > 2 and img.shape[2] == 1:
+ ret = np.expand_dims(ret, -1)
+ else:
+ # PIL only supports uint8
+ if any(x < 0 for x in img.strides):
+ img = np.ascontiguousarray(img)
+ img = torch.from_numpy(img)
+ shape = list(img.shape)
+ shape_4d = shape[:2] + [1] * (4 - len(shape)) + shape[2:]
+ img = img.view(shape_4d).permute(2, 3, 0, 1) # hw(c) -> nchw
+ _PIL_RESIZE_TO_INTERPOLATE_MODE = {
+ Image.NEAREST: "nearest",
+ Image.BILINEAR: "bilinear",
+ Image.BICUBIC: "bicubic",
+ }
+ mode = _PIL_RESIZE_TO_INTERPOLATE_MODE[interp_method]
+ align_corners = None if mode == "nearest" else False
+ img = F.interpolate(
+ img, (self.new_h, self.new_w), mode=mode, align_corners=align_corners
+ )
+ shape[:2] = (self.new_h, self.new_w)
+ ret = img.permute(2, 3, 0, 1).view(shape).numpy() # nchw -> hw(c)
+
+ return ret
+
+ def apply_coords(self, coords):
+ coords[:, 0] = coords[:, 0] * (self.new_w * 1.0 / self.w)
+ coords[:, 1] = coords[:, 1] * (self.new_h * 1.0 / self.h)
+ return coords
+
+ def apply_segmentation(self, segmentation):
+ segmentation = self.apply_image(segmentation, interp=Image.NEAREST)
+ return segmentation
+
+ def inverse(self):
+ return ResizeTransform(self.new_h, self.new_w, self.h, self.w, self.interp)
+
+
+class RotationTransform(Transform):
+ """
+ This method returns a copy of this image, rotated the given
+ number of degrees counter clockwise around its center.
+ """
+
+ def __init__(self, h, w, angle, expand=True, center=None, interp=None):
+ """
+ Args:
+ h, w (int): original image size
+ angle (float): degrees for rotation
+ expand (bool): choose if the image should be resized to fit the whole
+ rotated image (default), or simply cropped
+ center (tuple (width, height)): coordinates of the rotation center
+ if left to None, the center will be fit to the center of each image
+ center has no effect if expand=True because it only affects shifting
+ interp: cv2 interpolation method, default cv2.INTER_LINEAR
+ """
+ super().__init__()
+ image_center = np.array((w / 2, h / 2))
+ if center is None:
+ center = image_center
+ if interp is None:
+ interp = cv2.INTER_LINEAR
+ abs_cos, abs_sin = (abs(np.cos(np.deg2rad(angle))), abs(np.sin(np.deg2rad(angle))))
+ if expand:
+ # find the new width and height bounds
+ bound_w, bound_h = np.rint(
+ [h * abs_sin + w * abs_cos, h * abs_cos + w * abs_sin]
+ ).astype(int)
+ else:
+ bound_w, bound_h = w, h
+
+ self._set_attributes(locals())
+ self.rm_coords = self.create_rotation_matrix()
+ # Needed because of this problem https://github.com/opencv/opencv/issues/11784
+ self.rm_image = self.create_rotation_matrix(offset=-0.5)
+
+ def apply_image(self, img, interp=None):
+ """
+ img should be a numpy array, formatted as Height * Width * Nchannels
+ """
+ if len(img) == 0 or self.angle % 360 == 0:
+ return img
+ assert img.shape[:2] == (self.h, self.w)
+ interp = interp if interp is not None else self.interp
+ return cv2.warpAffine(img, self.rm_image, (self.bound_w, self.bound_h), flags=interp)
+
+ def apply_coords(self, coords):
+ """
+ coords should be a N * 2 array-like, containing N couples of (x, y) points
+ """
+ coords = np.asarray(coords, dtype=float)
+ if len(coords) == 0 or self.angle % 360 == 0:
+ return coords
+ return cv2.transform(coords[:, np.newaxis, :], self.rm_coords)[:, 0, :]
+
+ def apply_segmentation(self, segmentation):
+ segmentation = self.apply_image(segmentation, interp=cv2.INTER_NEAREST)
+ return segmentation
+
+ def create_rotation_matrix(self, offset=0):
+ center = (self.center[0] + offset, self.center[1] + offset)
+ rm = cv2.getRotationMatrix2D(tuple(center), self.angle, 1)
+ if self.expand:
+ # Find the coordinates of the center of rotation in the new image
+ # The only point for which we know the future coordinates is the center of the image
+ rot_im_center = cv2.transform(self.image_center[None, None, :] + offset, rm)[0, 0, :]
+ new_center = np.array([self.bound_w / 2, self.bound_h / 2]) + offset - rot_im_center
+ # shift the rotation center to the new coordinates
+ rm[:, 2] += new_center
+ return rm
+
+ def inverse(self):
+ """
+ The inverse is to rotate it back with expand, and crop to get the original shape.
+ """
+ if not self.expand: # Not possible to inverse if a part of the image is lost
+ raise NotImplementedError()
+ rotation = RotationTransform(
+ self.bound_h, self.bound_w, -self.angle, True, None, self.interp
+ )
+ crop = CropTransform(
+ (rotation.bound_w - self.w) // 2, (rotation.bound_h - self.h) // 2, self.w, self.h
+ )
+ return TransformList([rotation, crop])
+
+
+class ColorTransform(Transform):
+ """
+ Generic wrapper for any photometric transforms.
+ These transformations should only affect the color space and
+ not the coordinate space of the image (e.g. annotation
+ coordinates such as bounding boxes should not be changed)
+ """
+
+ def __init__(self, op):
+ """
+ Args:
+ op (Callable): operation to be applied to the image,
+ which takes in an ndarray and returns an ndarray.
+ """
+ if not callable(op):
+ raise ValueError("op parameter should be callable")
+ super().__init__()
+ self._set_attributes(locals())
+
+ def apply_image(self, img):
+ return self.op(img)
+
+ def apply_coords(self, coords):
+ return coords
+
+ def inverse(self):
+ return NoOpTransform()
+
+ def apply_segmentation(self, segmentation):
+ return segmentation
+
+
+class PILColorTransform(ColorTransform):
+ """
+ Generic wrapper for PIL Photometric image transforms,
+ which affect the color space and not the coordinate
+ space of the image
+ """
+
+ def __init__(self, op):
+ """
+ Args:
+ op (Callable): operation to be applied to the image,
+ which takes in a PIL Image and returns a transformed
+ PIL Image.
+ For reference on possible operations see:
+ - https://pillow.readthedocs.io/en/stable/
+ """
+ if not callable(op):
+ raise ValueError("op parameter should be callable")
+ super().__init__(op)
+
+ def apply_image(self, img):
+ img = Image.fromarray(img)
+ return np.asarray(super().apply_image(img))
+
+
+def HFlip_rotated_box(transform, rotated_boxes):
+ """
+ Apply the horizontal flip transform on rotated boxes.
+
+ Args:
+ rotated_boxes (ndarray): Nx5 floating point array of
+ (x_center, y_center, width, height, angle_degrees) format
+ in absolute coordinates.
+ """
+ # Transform x_center
+ rotated_boxes[:, 0] = transform.width - rotated_boxes[:, 0]
+ # Transform angle
+ rotated_boxes[:, 4] = -rotated_boxes[:, 4]
+ return rotated_boxes
+
+
+def Resize_rotated_box(transform, rotated_boxes):
+ """
+ Apply the resizing transform on rotated boxes. For details of how these (approximation)
+ formulas are derived, please refer to :meth:`RotatedBoxes.scale`.
+
+ Args:
+ rotated_boxes (ndarray): Nx5 floating point array of
+ (x_center, y_center, width, height, angle_degrees) format
+ in absolute coordinates.
+ """
+ scale_factor_x = transform.new_w * 1.0 / transform.w
+ scale_factor_y = transform.new_h * 1.0 / transform.h
+ rotated_boxes[:, 0] *= scale_factor_x
+ rotated_boxes[:, 1] *= scale_factor_y
+ theta = rotated_boxes[:, 4] * np.pi / 180.0
+ c = np.cos(theta)
+ s = np.sin(theta)
+ rotated_boxes[:, 2] *= np.sqrt(np.square(scale_factor_x * c) + np.square(scale_factor_y * s))
+ rotated_boxes[:, 3] *= np.sqrt(np.square(scale_factor_x * s) + np.square(scale_factor_y * c))
+ rotated_boxes[:, 4] = np.arctan2(scale_factor_x * s, scale_factor_y * c) * 180 / np.pi
+
+ return rotated_boxes
+
+
+HFlipTransform.register_type("rotated_box", HFlip_rotated_box)
+ResizeTransform.register_type("rotated_box", Resize_rotated_box)
+
+# not necessary any more with latest fvcore
+NoOpTransform.register_type("rotated_box", lambda t, x: x)
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/engine/__init__.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/engine/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..08a61572b4c7d09c8d400e903a96cbf5b2cc4763
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/engine/__init__.py
@@ -0,0 +1,12 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+
+from .launch import *
+from .train_loop import *
+
+__all__ = [k for k in globals().keys() if not k.startswith("_")]
+
+
+# prefer to let hooks and defaults live in separate namespaces (therefore not in __all__)
+# but still make them available here
+from .hooks import *
+from .defaults import *
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/engine/defaults.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/engine/defaults.py
new file mode 100644
index 0000000000000000000000000000000000000000..51d49148ca7b048402a63490bf7df83a43c65d9f
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/engine/defaults.py
@@ -0,0 +1,715 @@
+# -*- coding: utf-8 -*-
+# Copyright (c) Facebook, Inc. and its affiliates.
+
+"""
+This file contains components with some default boilerplate logic user may need
+in training / testing. They will not work for everyone, but many users may find them useful.
+
+The behavior of functions/classes in this file is subject to change,
+since they are meant to represent the "common default behavior" people need in their projects.
+"""
+
+import argparse
+import logging
+import os
+import sys
+import weakref
+from collections import OrderedDict
+from typing import Optional
+import torch
+from fvcore.nn.precise_bn import get_bn_modules
+from omegaconf import OmegaConf
+from torch.nn.parallel import DistributedDataParallel
+
+import annotator.oneformer.detectron2.data.transforms as T
+from annotator.oneformer.detectron2.checkpoint import DetectionCheckpointer
+from annotator.oneformer.detectron2.config import CfgNode, LazyConfig
+from annotator.oneformer.detectron2.data import (
+ MetadataCatalog,
+ build_detection_test_loader,
+ build_detection_train_loader,
+)
+from annotator.oneformer.detectron2.evaluation import (
+ DatasetEvaluator,
+ inference_on_dataset,
+ print_csv_format,
+ verify_results,
+)
+from annotator.oneformer.detectron2.modeling import build_model
+from annotator.oneformer.detectron2.solver import build_lr_scheduler, build_optimizer
+from annotator.oneformer.detectron2.utils import comm
+from annotator.oneformer.detectron2.utils.collect_env import collect_env_info
+from annotator.oneformer.detectron2.utils.env import seed_all_rng
+from annotator.oneformer.detectron2.utils.events import CommonMetricPrinter, JSONWriter, TensorboardXWriter
+from annotator.oneformer.detectron2.utils.file_io import PathManager
+from annotator.oneformer.detectron2.utils.logger import setup_logger
+
+from . import hooks
+from .train_loop import AMPTrainer, SimpleTrainer, TrainerBase
+
+__all__ = [
+ "create_ddp_model",
+ "default_argument_parser",
+ "default_setup",
+ "default_writers",
+ "DefaultPredictor",
+ "DefaultTrainer",
+]
+
+
+def create_ddp_model(model, *, fp16_compression=False, **kwargs):
+ """
+ Create a DistributedDataParallel model if there are >1 processes.
+
+ Args:
+ model: a torch.nn.Module
+ fp16_compression: add fp16 compression hooks to the ddp object.
+ See more at https://pytorch.org/docs/stable/ddp_comm_hooks.html#torch.distributed.algorithms.ddp_comm_hooks.default_hooks.fp16_compress_hook
+ kwargs: other arguments of :module:`torch.nn.parallel.DistributedDataParallel`.
+ """ # noqa
+ if comm.get_world_size() == 1:
+ return model
+ if "device_ids" not in kwargs:
+ kwargs["device_ids"] = [comm.get_local_rank()]
+ ddp = DistributedDataParallel(model, **kwargs)
+ if fp16_compression:
+ from torch.distributed.algorithms.ddp_comm_hooks import default as comm_hooks
+
+ ddp.register_comm_hook(state=None, hook=comm_hooks.fp16_compress_hook)
+ return ddp
+
+
+def default_argument_parser(epilog=None):
+ """
+ Create a parser with some common arguments used by detectron2 users.
+
+ Args:
+ epilog (str): epilog passed to ArgumentParser describing the usage.
+
+ Returns:
+ argparse.ArgumentParser:
+ """
+ parser = argparse.ArgumentParser(
+ epilog=epilog
+ or f"""
+Examples:
+
+Run on single machine:
+ $ {sys.argv[0]} --num-gpus 8 --config-file cfg.yaml
+
+Change some config options:
+ $ {sys.argv[0]} --config-file cfg.yaml MODEL.WEIGHTS /path/to/weight.pth SOLVER.BASE_LR 0.001
+
+Run on multiple machines:
+ (machine0)$ {sys.argv[0]} --machine-rank 0 --num-machines 2 --dist-url [--other-flags]
+ (machine1)$ {sys.argv[0]} --machine-rank 1 --num-machines 2 --dist-url [--other-flags]
+""",
+ formatter_class=argparse.RawDescriptionHelpFormatter,
+ )
+ parser.add_argument("--config-file", default="", metavar="FILE", help="path to config file")
+ parser.add_argument(
+ "--resume",
+ action="store_true",
+ help="Whether to attempt to resume from the checkpoint directory. "
+ "See documentation of `DefaultTrainer.resume_or_load()` for what it means.",
+ )
+ parser.add_argument("--eval-only", action="store_true", help="perform evaluation only")
+ parser.add_argument("--num-gpus", type=int, default=1, help="number of gpus *per machine*")
+ parser.add_argument("--num-machines", type=int, default=1, help="total number of machines")
+ parser.add_argument(
+ "--machine-rank", type=int, default=0, help="the rank of this machine (unique per machine)"
+ )
+
+ # PyTorch still may leave orphan processes in multi-gpu training.
+ # Therefore we use a deterministic way to obtain port,
+ # so that users are aware of orphan processes by seeing the port occupied.
+ port = 2**15 + 2**14 + hash(os.getuid() if sys.platform != "win32" else 1) % 2**14
+ parser.add_argument(
+ "--dist-url",
+ default="tcp://127.0.0.1:{}".format(port),
+ help="initialization URL for pytorch distributed backend. See "
+ "https://pytorch.org/docs/stable/distributed.html for details.",
+ )
+ parser.add_argument(
+ "opts",
+ help="""
+Modify config options at the end of the command. For Yacs configs, use
+space-separated "PATH.KEY VALUE" pairs.
+For python-based LazyConfig, use "path.key=value".
+ """.strip(),
+ default=None,
+ nargs=argparse.REMAINDER,
+ )
+ return parser
+
+
+def _try_get_key(cfg, *keys, default=None):
+ """
+ Try select keys from cfg until the first key that exists. Otherwise return default.
+ """
+ if isinstance(cfg, CfgNode):
+ cfg = OmegaConf.create(cfg.dump())
+ for k in keys:
+ none = object()
+ p = OmegaConf.select(cfg, k, default=none)
+ if p is not none:
+ return p
+ return default
+
+
+def _highlight(code, filename):
+ try:
+ import pygments
+ except ImportError:
+ return code
+
+ from pygments.lexers import Python3Lexer, YamlLexer
+ from pygments.formatters import Terminal256Formatter
+
+ lexer = Python3Lexer() if filename.endswith(".py") else YamlLexer()
+ code = pygments.highlight(code, lexer, Terminal256Formatter(style="monokai"))
+ return code
+
+
+def default_setup(cfg, args):
+ """
+ Perform some basic common setups at the beginning of a job, including:
+
+ 1. Set up the detectron2 logger
+ 2. Log basic information about environment, cmdline arguments, and config
+ 3. Backup the config to the output directory
+
+ Args:
+ cfg (CfgNode or omegaconf.DictConfig): the full config to be used
+ args (argparse.NameSpace): the command line arguments to be logged
+ """
+ output_dir = _try_get_key(cfg, "OUTPUT_DIR", "output_dir", "train.output_dir")
+ if comm.is_main_process() and output_dir:
+ PathManager.mkdirs(output_dir)
+
+ rank = comm.get_rank()
+ setup_logger(output_dir, distributed_rank=rank, name="fvcore")
+ logger = setup_logger(output_dir, distributed_rank=rank)
+
+ logger.info("Rank of current process: {}. World size: {}".format(rank, comm.get_world_size()))
+ logger.info("Environment info:\n" + collect_env_info())
+
+ logger.info("Command line arguments: " + str(args))
+ if hasattr(args, "config_file") and args.config_file != "":
+ logger.info(
+ "Contents of args.config_file={}:\n{}".format(
+ args.config_file,
+ _highlight(PathManager.open(args.config_file, "r").read(), args.config_file),
+ )
+ )
+
+ if comm.is_main_process() and output_dir:
+ # Note: some of our scripts may expect the existence of
+ # config.yaml in output directory
+ path = os.path.join(output_dir, "config.yaml")
+ if isinstance(cfg, CfgNode):
+ logger.info("Running with full config:\n{}".format(_highlight(cfg.dump(), ".yaml")))
+ with PathManager.open(path, "w") as f:
+ f.write(cfg.dump())
+ else:
+ LazyConfig.save(cfg, path)
+ logger.info("Full config saved to {}".format(path))
+
+ # make sure each worker has a different, yet deterministic seed if specified
+ seed = _try_get_key(cfg, "SEED", "train.seed", default=-1)
+ seed_all_rng(None if seed < 0 else seed + rank)
+
+ # cudnn benchmark has large overhead. It shouldn't be used considering the small size of
+ # typical validation set.
+ if not (hasattr(args, "eval_only") and args.eval_only):
+ torch.backends.cudnn.benchmark = _try_get_key(
+ cfg, "CUDNN_BENCHMARK", "train.cudnn_benchmark", default=False
+ )
+
+
+def default_writers(output_dir: str, max_iter: Optional[int] = None):
+ """
+ Build a list of :class:`EventWriter` to be used.
+ It now consists of a :class:`CommonMetricPrinter`,
+ :class:`TensorboardXWriter` and :class:`JSONWriter`.
+
+ Args:
+ output_dir: directory to store JSON metrics and tensorboard events
+ max_iter: the total number of iterations
+
+ Returns:
+ list[EventWriter]: a list of :class:`EventWriter` objects.
+ """
+ PathManager.mkdirs(output_dir)
+ return [
+ # It may not always print what you want to see, since it prints "common" metrics only.
+ CommonMetricPrinter(max_iter),
+ JSONWriter(os.path.join(output_dir, "metrics.json")),
+ TensorboardXWriter(output_dir),
+ ]
+
+
+class DefaultPredictor:
+ """
+ Create a simple end-to-end predictor with the given config that runs on
+ single device for a single input image.
+
+ Compared to using the model directly, this class does the following additions:
+
+ 1. Load checkpoint from `cfg.MODEL.WEIGHTS`.
+ 2. Always take BGR image as the input and apply conversion defined by `cfg.INPUT.FORMAT`.
+ 3. Apply resizing defined by `cfg.INPUT.{MIN,MAX}_SIZE_TEST`.
+ 4. Take one input image and produce a single output, instead of a batch.
+
+ This is meant for simple demo purposes, so it does the above steps automatically.
+ This is not meant for benchmarks or running complicated inference logic.
+ If you'd like to do anything more complicated, please refer to its source code as
+ examples to build and use the model manually.
+
+ Attributes:
+ metadata (Metadata): the metadata of the underlying dataset, obtained from
+ cfg.DATASETS.TEST.
+
+ Examples:
+ ::
+ pred = DefaultPredictor(cfg)
+ inputs = cv2.imread("input.jpg")
+ outputs = pred(inputs)
+ """
+
+ def __init__(self, cfg):
+ self.cfg = cfg.clone() # cfg can be modified by model
+ self.model = build_model(self.cfg)
+ self.model.eval()
+ if len(cfg.DATASETS.TEST):
+ self.metadata = MetadataCatalog.get(cfg.DATASETS.TEST[0])
+
+ checkpointer = DetectionCheckpointer(self.model)
+ checkpointer.load(cfg.MODEL.WEIGHTS)
+
+ self.aug = T.ResizeShortestEdge(
+ [cfg.INPUT.MIN_SIZE_TEST, cfg.INPUT.MIN_SIZE_TEST], cfg.INPUT.MAX_SIZE_TEST
+ )
+
+ self.input_format = cfg.INPUT.FORMAT
+ assert self.input_format in ["RGB", "BGR"], self.input_format
+
+ def __call__(self, original_image):
+ """
+ Args:
+ original_image (np.ndarray): an image of shape (H, W, C) (in BGR order).
+
+ Returns:
+ predictions (dict):
+ the output of the model for one image only.
+ See :doc:`/tutorials/models` for details about the format.
+ """
+ with torch.no_grad(): # https://github.com/sphinx-doc/sphinx/issues/4258
+ # Apply pre-processing to image.
+ if self.input_format == "RGB":
+ # whether the model expects BGR inputs or RGB
+ original_image = original_image[:, :, ::-1]
+ height, width = original_image.shape[:2]
+ image = self.aug.get_transform(original_image).apply_image(original_image)
+ image = torch.as_tensor(image.astype("float32").transpose(2, 0, 1))
+
+ inputs = {"image": image, "height": height, "width": width}
+ predictions = self.model([inputs])[0]
+ return predictions
+
+
+class DefaultTrainer(TrainerBase):
+ """
+ A trainer with default training logic. It does the following:
+
+ 1. Create a :class:`SimpleTrainer` using model, optimizer, dataloader
+ defined by the given config. Create a LR scheduler defined by the config.
+ 2. Load the last checkpoint or `cfg.MODEL.WEIGHTS`, if exists, when
+ `resume_or_load` is called.
+ 3. Register a few common hooks defined by the config.
+
+ It is created to simplify the **standard model training workflow** and reduce code boilerplate
+ for users who only need the standard training workflow, with standard features.
+ It means this class makes *many assumptions* about your training logic that
+ may easily become invalid in a new research. In fact, any assumptions beyond those made in the
+ :class:`SimpleTrainer` are too much for research.
+
+ The code of this class has been annotated about restrictive assumptions it makes.
+ When they do not work for you, you're encouraged to:
+
+ 1. Overwrite methods of this class, OR:
+ 2. Use :class:`SimpleTrainer`, which only does minimal SGD training and
+ nothing else. You can then add your own hooks if needed. OR:
+ 3. Write your own training loop similar to `tools/plain_train_net.py`.
+
+ See the :doc:`/tutorials/training` tutorials for more details.
+
+ Note that the behavior of this class, like other functions/classes in
+ this file, is not stable, since it is meant to represent the "common default behavior".
+ It is only guaranteed to work well with the standard models and training workflow in detectron2.
+ To obtain more stable behavior, write your own training logic with other public APIs.
+
+ Examples:
+ ::
+ trainer = DefaultTrainer(cfg)
+ trainer.resume_or_load() # load last checkpoint or MODEL.WEIGHTS
+ trainer.train()
+
+ Attributes:
+ scheduler:
+ checkpointer (DetectionCheckpointer):
+ cfg (CfgNode):
+ """
+
+ def __init__(self, cfg):
+ """
+ Args:
+ cfg (CfgNode):
+ """
+ super().__init__()
+ logger = logging.getLogger("detectron2")
+ if not logger.isEnabledFor(logging.INFO): # setup_logger is not called for d2
+ setup_logger()
+ cfg = DefaultTrainer.auto_scale_workers(cfg, comm.get_world_size())
+
+ # Assume these objects must be constructed in this order.
+ model = self.build_model(cfg)
+ optimizer = self.build_optimizer(cfg, model)
+ data_loader = self.build_train_loader(cfg)
+
+ model = create_ddp_model(model, broadcast_buffers=False)
+ self._trainer = (AMPTrainer if cfg.SOLVER.AMP.ENABLED else SimpleTrainer)(
+ model, data_loader, optimizer
+ )
+
+ self.scheduler = self.build_lr_scheduler(cfg, optimizer)
+ self.checkpointer = DetectionCheckpointer(
+ # Assume you want to save checkpoints together with logs/statistics
+ model,
+ cfg.OUTPUT_DIR,
+ trainer=weakref.proxy(self),
+ )
+ self.start_iter = 0
+ self.max_iter = cfg.SOLVER.MAX_ITER
+ self.cfg = cfg
+
+ self.register_hooks(self.build_hooks())
+
+ def resume_or_load(self, resume=True):
+ """
+ If `resume==True` and `cfg.OUTPUT_DIR` contains the last checkpoint (defined by
+ a `last_checkpoint` file), resume from the file. Resuming means loading all
+ available states (eg. optimizer and scheduler) and update iteration counter
+ from the checkpoint. ``cfg.MODEL.WEIGHTS`` will not be used.
+
+ Otherwise, this is considered as an independent training. The method will load model
+ weights from the file `cfg.MODEL.WEIGHTS` (but will not load other states) and start
+ from iteration 0.
+
+ Args:
+ resume (bool): whether to do resume or not
+ """
+ self.checkpointer.resume_or_load(self.cfg.MODEL.WEIGHTS, resume=resume)
+ if resume and self.checkpointer.has_checkpoint():
+ # The checkpoint stores the training iteration that just finished, thus we start
+ # at the next iteration
+ self.start_iter = self.iter + 1
+
+ def build_hooks(self):
+ """
+ Build a list of default hooks, including timing, evaluation,
+ checkpointing, lr scheduling, precise BN, writing events.
+
+ Returns:
+ list[HookBase]:
+ """
+ cfg = self.cfg.clone()
+ cfg.defrost()
+ cfg.DATALOADER.NUM_WORKERS = 0 # save some memory and time for PreciseBN
+
+ ret = [
+ hooks.IterationTimer(),
+ hooks.LRScheduler(),
+ hooks.PreciseBN(
+ # Run at the same freq as (but before) evaluation.
+ cfg.TEST.EVAL_PERIOD,
+ self.model,
+ # Build a new data loader to not affect training
+ self.build_train_loader(cfg),
+ cfg.TEST.PRECISE_BN.NUM_ITER,
+ )
+ if cfg.TEST.PRECISE_BN.ENABLED and get_bn_modules(self.model)
+ else None,
+ ]
+
+ # Do PreciseBN before checkpointer, because it updates the model and need to
+ # be saved by checkpointer.
+ # This is not always the best: if checkpointing has a different frequency,
+ # some checkpoints may have more precise statistics than others.
+ if comm.is_main_process():
+ ret.append(hooks.PeriodicCheckpointer(self.checkpointer, cfg.SOLVER.CHECKPOINT_PERIOD))
+
+ def test_and_save_results():
+ self._last_eval_results = self.test(self.cfg, self.model)
+ return self._last_eval_results
+
+ # Do evaluation after checkpointer, because then if it fails,
+ # we can use the saved checkpoint to debug.
+ ret.append(hooks.EvalHook(cfg.TEST.EVAL_PERIOD, test_and_save_results))
+
+ if comm.is_main_process():
+ # Here the default print/log frequency of each writer is used.
+ # run writers in the end, so that evaluation metrics are written
+ ret.append(hooks.PeriodicWriter(self.build_writers(), period=20))
+ return ret
+
+ def build_writers(self):
+ """
+ Build a list of writers to be used using :func:`default_writers()`.
+ If you'd like a different list of writers, you can overwrite it in
+ your trainer.
+
+ Returns:
+ list[EventWriter]: a list of :class:`EventWriter` objects.
+ """
+ return default_writers(self.cfg.OUTPUT_DIR, self.max_iter)
+
+ def train(self):
+ """
+ Run training.
+
+ Returns:
+ OrderedDict of results, if evaluation is enabled. Otherwise None.
+ """
+ super().train(self.start_iter, self.max_iter)
+ if len(self.cfg.TEST.EXPECTED_RESULTS) and comm.is_main_process():
+ assert hasattr(
+ self, "_last_eval_results"
+ ), "No evaluation results obtained during training!"
+ verify_results(self.cfg, self._last_eval_results)
+ return self._last_eval_results
+
+ def run_step(self):
+ self._trainer.iter = self.iter
+ self._trainer.run_step()
+
+ def state_dict(self):
+ ret = super().state_dict()
+ ret["_trainer"] = self._trainer.state_dict()
+ return ret
+
+ def load_state_dict(self, state_dict):
+ super().load_state_dict(state_dict)
+ self._trainer.load_state_dict(state_dict["_trainer"])
+
+ @classmethod
+ def build_model(cls, cfg):
+ """
+ Returns:
+ torch.nn.Module:
+
+ It now calls :func:`detectron2.modeling.build_model`.
+ Overwrite it if you'd like a different model.
+ """
+ model = build_model(cfg)
+ logger = logging.getLogger(__name__)
+ logger.info("Model:\n{}".format(model))
+ return model
+
+ @classmethod
+ def build_optimizer(cls, cfg, model):
+ """
+ Returns:
+ torch.optim.Optimizer:
+
+ It now calls :func:`detectron2.solver.build_optimizer`.
+ Overwrite it if you'd like a different optimizer.
+ """
+ return build_optimizer(cfg, model)
+
+ @classmethod
+ def build_lr_scheduler(cls, cfg, optimizer):
+ """
+ It now calls :func:`detectron2.solver.build_lr_scheduler`.
+ Overwrite it if you'd like a different scheduler.
+ """
+ return build_lr_scheduler(cfg, optimizer)
+
+ @classmethod
+ def build_train_loader(cls, cfg):
+ """
+ Returns:
+ iterable
+
+ It now calls :func:`detectron2.data.build_detection_train_loader`.
+ Overwrite it if you'd like a different data loader.
+ """
+ return build_detection_train_loader(cfg)
+
+ @classmethod
+ def build_test_loader(cls, cfg, dataset_name):
+ """
+ Returns:
+ iterable
+
+ It now calls :func:`detectron2.data.build_detection_test_loader`.
+ Overwrite it if you'd like a different data loader.
+ """
+ return build_detection_test_loader(cfg, dataset_name)
+
+ @classmethod
+ def build_evaluator(cls, cfg, dataset_name):
+ """
+ Returns:
+ DatasetEvaluator or None
+
+ It is not implemented by default.
+ """
+ raise NotImplementedError(
+ """
+If you want DefaultTrainer to automatically run evaluation,
+please implement `build_evaluator()` in subclasses (see train_net.py for example).
+Alternatively, you can call evaluation functions yourself (see Colab balloon tutorial for example).
+"""
+ )
+
+ @classmethod
+ def test(cls, cfg, model, evaluators=None):
+ """
+ Evaluate the given model. The given model is expected to already contain
+ weights to evaluate.
+
+ Args:
+ cfg (CfgNode):
+ model (nn.Module):
+ evaluators (list[DatasetEvaluator] or None): if None, will call
+ :meth:`build_evaluator`. Otherwise, must have the same length as
+ ``cfg.DATASETS.TEST``.
+
+ Returns:
+ dict: a dict of result metrics
+ """
+ logger = logging.getLogger(__name__)
+ if isinstance(evaluators, DatasetEvaluator):
+ evaluators = [evaluators]
+ if evaluators is not None:
+ assert len(cfg.DATASETS.TEST) == len(evaluators), "{} != {}".format(
+ len(cfg.DATASETS.TEST), len(evaluators)
+ )
+
+ results = OrderedDict()
+ for idx, dataset_name in enumerate(cfg.DATASETS.TEST):
+ data_loader = cls.build_test_loader(cfg, dataset_name)
+ # When evaluators are passed in as arguments,
+ # implicitly assume that evaluators can be created before data_loader.
+ if evaluators is not None:
+ evaluator = evaluators[idx]
+ else:
+ try:
+ evaluator = cls.build_evaluator(cfg, dataset_name)
+ except NotImplementedError:
+ logger.warn(
+ "No evaluator found. Use `DefaultTrainer.test(evaluators=)`, "
+ "or implement its `build_evaluator` method."
+ )
+ results[dataset_name] = {}
+ continue
+ results_i = inference_on_dataset(model, data_loader, evaluator)
+ results[dataset_name] = results_i
+ if comm.is_main_process():
+ assert isinstance(
+ results_i, dict
+ ), "Evaluator must return a dict on the main process. Got {} instead.".format(
+ results_i
+ )
+ logger.info("Evaluation results for {} in csv format:".format(dataset_name))
+ print_csv_format(results_i)
+
+ if len(results) == 1:
+ results = list(results.values())[0]
+ return results
+
+ @staticmethod
+ def auto_scale_workers(cfg, num_workers: int):
+ """
+ When the config is defined for certain number of workers (according to
+ ``cfg.SOLVER.REFERENCE_WORLD_SIZE``) that's different from the number of
+ workers currently in use, returns a new cfg where the total batch size
+ is scaled so that the per-GPU batch size stays the same as the
+ original ``IMS_PER_BATCH // REFERENCE_WORLD_SIZE``.
+
+ Other config options are also scaled accordingly:
+ * training steps and warmup steps are scaled inverse proportionally.
+ * learning rate are scaled proportionally, following :paper:`ImageNet in 1h`.
+
+ For example, with the original config like the following:
+
+ .. code-block:: yaml
+
+ IMS_PER_BATCH: 16
+ BASE_LR: 0.1
+ REFERENCE_WORLD_SIZE: 8
+ MAX_ITER: 5000
+ STEPS: (4000,)
+ CHECKPOINT_PERIOD: 1000
+
+ When this config is used on 16 GPUs instead of the reference number 8,
+ calling this method will return a new config with:
+
+ .. code-block:: yaml
+
+ IMS_PER_BATCH: 32
+ BASE_LR: 0.2
+ REFERENCE_WORLD_SIZE: 16
+ MAX_ITER: 2500
+ STEPS: (2000,)
+ CHECKPOINT_PERIOD: 500
+
+ Note that both the original config and this new config can be trained on 16 GPUs.
+ It's up to user whether to enable this feature (by setting ``REFERENCE_WORLD_SIZE``).
+
+ Returns:
+ CfgNode: a new config. Same as original if ``cfg.SOLVER.REFERENCE_WORLD_SIZE==0``.
+ """
+ old_world_size = cfg.SOLVER.REFERENCE_WORLD_SIZE
+ if old_world_size == 0 or old_world_size == num_workers:
+ return cfg
+ cfg = cfg.clone()
+ frozen = cfg.is_frozen()
+ cfg.defrost()
+
+ assert (
+ cfg.SOLVER.IMS_PER_BATCH % old_world_size == 0
+ ), "Invalid REFERENCE_WORLD_SIZE in config!"
+ scale = num_workers / old_world_size
+ bs = cfg.SOLVER.IMS_PER_BATCH = int(round(cfg.SOLVER.IMS_PER_BATCH * scale))
+ lr = cfg.SOLVER.BASE_LR = cfg.SOLVER.BASE_LR * scale
+ max_iter = cfg.SOLVER.MAX_ITER = int(round(cfg.SOLVER.MAX_ITER / scale))
+ warmup_iter = cfg.SOLVER.WARMUP_ITERS = int(round(cfg.SOLVER.WARMUP_ITERS / scale))
+ cfg.SOLVER.STEPS = tuple(int(round(s / scale)) for s in cfg.SOLVER.STEPS)
+ cfg.TEST.EVAL_PERIOD = int(round(cfg.TEST.EVAL_PERIOD / scale))
+ cfg.SOLVER.CHECKPOINT_PERIOD = int(round(cfg.SOLVER.CHECKPOINT_PERIOD / scale))
+ cfg.SOLVER.REFERENCE_WORLD_SIZE = num_workers # maintain invariant
+ logger = logging.getLogger(__name__)
+ logger.info(
+ f"Auto-scaling the config to batch_size={bs}, learning_rate={lr}, "
+ f"max_iter={max_iter}, warmup={warmup_iter}."
+ )
+
+ if frozen:
+ cfg.freeze()
+ return cfg
+
+
+# Access basic attributes from the underlying trainer
+for _attr in ["model", "data_loader", "optimizer"]:
+ setattr(
+ DefaultTrainer,
+ _attr,
+ property(
+ # getter
+ lambda self, x=_attr: getattr(self._trainer, x),
+ # setter
+ lambda self, value, x=_attr: setattr(self._trainer, x, value),
+ ),
+ )
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/engine/hooks.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/engine/hooks.py
new file mode 100644
index 0000000000000000000000000000000000000000..7dd43ac77068c908bc13263f1697fa2e3332d7c9
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/engine/hooks.py
@@ -0,0 +1,690 @@
+# -*- coding: utf-8 -*-
+# Copyright (c) Facebook, Inc. and its affiliates.
+
+import datetime
+import itertools
+import logging
+import math
+import operator
+import os
+import tempfile
+import time
+import warnings
+from collections import Counter
+import torch
+from fvcore.common.checkpoint import Checkpointer
+from fvcore.common.checkpoint import PeriodicCheckpointer as _PeriodicCheckpointer
+from fvcore.common.param_scheduler import ParamScheduler
+from fvcore.common.timer import Timer
+from fvcore.nn.precise_bn import get_bn_modules, update_bn_stats
+
+import annotator.oneformer.detectron2.utils.comm as comm
+from annotator.oneformer.detectron2.evaluation.testing import flatten_results_dict
+from annotator.oneformer.detectron2.solver import LRMultiplier
+from annotator.oneformer.detectron2.solver import LRScheduler as _LRScheduler
+from annotator.oneformer.detectron2.utils.events import EventStorage, EventWriter
+from annotator.oneformer.detectron2.utils.file_io import PathManager
+
+from .train_loop import HookBase
+
+__all__ = [
+ "CallbackHook",
+ "IterationTimer",
+ "PeriodicWriter",
+ "PeriodicCheckpointer",
+ "BestCheckpointer",
+ "LRScheduler",
+ "AutogradProfiler",
+ "EvalHook",
+ "PreciseBN",
+ "TorchProfiler",
+ "TorchMemoryStats",
+]
+
+
+"""
+Implement some common hooks.
+"""
+
+
+class CallbackHook(HookBase):
+ """
+ Create a hook using callback functions provided by the user.
+ """
+
+ def __init__(self, *, before_train=None, after_train=None, before_step=None, after_step=None):
+ """
+ Each argument is a function that takes one argument: the trainer.
+ """
+ self._before_train = before_train
+ self._before_step = before_step
+ self._after_step = after_step
+ self._after_train = after_train
+
+ def before_train(self):
+ if self._before_train:
+ self._before_train(self.trainer)
+
+ def after_train(self):
+ if self._after_train:
+ self._after_train(self.trainer)
+ # The functions may be closures that hold reference to the trainer
+ # Therefore, delete them to avoid circular reference.
+ del self._before_train, self._after_train
+ del self._before_step, self._after_step
+
+ def before_step(self):
+ if self._before_step:
+ self._before_step(self.trainer)
+
+ def after_step(self):
+ if self._after_step:
+ self._after_step(self.trainer)
+
+
+class IterationTimer(HookBase):
+ """
+ Track the time spent for each iteration (each run_step call in the trainer).
+ Print a summary in the end of training.
+
+ This hook uses the time between the call to its :meth:`before_step`
+ and :meth:`after_step` methods.
+ Under the convention that :meth:`before_step` of all hooks should only
+ take negligible amount of time, the :class:`IterationTimer` hook should be
+ placed at the beginning of the list of hooks to obtain accurate timing.
+ """
+
+ def __init__(self, warmup_iter=3):
+ """
+ Args:
+ warmup_iter (int): the number of iterations at the beginning to exclude
+ from timing.
+ """
+ self._warmup_iter = warmup_iter
+ self._step_timer = Timer()
+ self._start_time = time.perf_counter()
+ self._total_timer = Timer()
+
+ def before_train(self):
+ self._start_time = time.perf_counter()
+ self._total_timer.reset()
+ self._total_timer.pause()
+
+ def after_train(self):
+ logger = logging.getLogger(__name__)
+ total_time = time.perf_counter() - self._start_time
+ total_time_minus_hooks = self._total_timer.seconds()
+ hook_time = total_time - total_time_minus_hooks
+
+ num_iter = self.trainer.storage.iter + 1 - self.trainer.start_iter - self._warmup_iter
+
+ if num_iter > 0 and total_time_minus_hooks > 0:
+ # Speed is meaningful only after warmup
+ # NOTE this format is parsed by grep in some scripts
+ logger.info(
+ "Overall training speed: {} iterations in {} ({:.4f} s / it)".format(
+ num_iter,
+ str(datetime.timedelta(seconds=int(total_time_minus_hooks))),
+ total_time_minus_hooks / num_iter,
+ )
+ )
+
+ logger.info(
+ "Total training time: {} ({} on hooks)".format(
+ str(datetime.timedelta(seconds=int(total_time))),
+ str(datetime.timedelta(seconds=int(hook_time))),
+ )
+ )
+
+ def before_step(self):
+ self._step_timer.reset()
+ self._total_timer.resume()
+
+ def after_step(self):
+ # +1 because we're in after_step, the current step is done
+ # but not yet counted
+ iter_done = self.trainer.storage.iter - self.trainer.start_iter + 1
+ if iter_done >= self._warmup_iter:
+ sec = self._step_timer.seconds()
+ self.trainer.storage.put_scalars(time=sec)
+ else:
+ self._start_time = time.perf_counter()
+ self._total_timer.reset()
+
+ self._total_timer.pause()
+
+
+class PeriodicWriter(HookBase):
+ """
+ Write events to EventStorage (by calling ``writer.write()``) periodically.
+
+ It is executed every ``period`` iterations and after the last iteration.
+ Note that ``period`` does not affect how data is smoothed by each writer.
+ """
+
+ def __init__(self, writers, period=20):
+ """
+ Args:
+ writers (list[EventWriter]): a list of EventWriter objects
+ period (int):
+ """
+ self._writers = writers
+ for w in writers:
+ assert isinstance(w, EventWriter), w
+ self._period = period
+
+ def after_step(self):
+ if (self.trainer.iter + 1) % self._period == 0 or (
+ self.trainer.iter == self.trainer.max_iter - 1
+ ):
+ for writer in self._writers:
+ writer.write()
+
+ def after_train(self):
+ for writer in self._writers:
+ # If any new data is found (e.g. produced by other after_train),
+ # write them before closing
+ writer.write()
+ writer.close()
+
+
+class PeriodicCheckpointer(_PeriodicCheckpointer, HookBase):
+ """
+ Same as :class:`detectron2.checkpoint.PeriodicCheckpointer`, but as a hook.
+
+ Note that when used as a hook,
+ it is unable to save additional data other than what's defined
+ by the given `checkpointer`.
+
+ It is executed every ``period`` iterations and after the last iteration.
+ """
+
+ def before_train(self):
+ self.max_iter = self.trainer.max_iter
+
+ def after_step(self):
+ # No way to use **kwargs
+ self.step(self.trainer.iter)
+
+
+class BestCheckpointer(HookBase):
+ """
+ Checkpoints best weights based off given metric.
+
+ This hook should be used in conjunction to and executed after the hook
+ that produces the metric, e.g. `EvalHook`.
+ """
+
+ def __init__(
+ self,
+ eval_period: int,
+ checkpointer: Checkpointer,
+ val_metric: str,
+ mode: str = "max",
+ file_prefix: str = "model_best",
+ ) -> None:
+ """
+ Args:
+ eval_period (int): the period `EvalHook` is set to run.
+ checkpointer: the checkpointer object used to save checkpoints.
+ val_metric (str): validation metric to track for best checkpoint, e.g. "bbox/AP50"
+ mode (str): one of {'max', 'min'}. controls whether the chosen val metric should be
+ maximized or minimized, e.g. for "bbox/AP50" it should be "max"
+ file_prefix (str): the prefix of checkpoint's filename, defaults to "model_best"
+ """
+ self._logger = logging.getLogger(__name__)
+ self._period = eval_period
+ self._val_metric = val_metric
+ assert mode in [
+ "max",
+ "min",
+ ], f'Mode "{mode}" to `BestCheckpointer` is unknown. It should be one of {"max", "min"}.'
+ if mode == "max":
+ self._compare = operator.gt
+ else:
+ self._compare = operator.lt
+ self._checkpointer = checkpointer
+ self._file_prefix = file_prefix
+ self.best_metric = None
+ self.best_iter = None
+
+ def _update_best(self, val, iteration):
+ if math.isnan(val) or math.isinf(val):
+ return False
+ self.best_metric = val
+ self.best_iter = iteration
+ return True
+
+ def _best_checking(self):
+ metric_tuple = self.trainer.storage.latest().get(self._val_metric)
+ if metric_tuple is None:
+ self._logger.warning(
+ f"Given val metric {self._val_metric} does not seem to be computed/stored."
+ "Will not be checkpointing based on it."
+ )
+ return
+ else:
+ latest_metric, metric_iter = metric_tuple
+
+ if self.best_metric is None:
+ if self._update_best(latest_metric, metric_iter):
+ additional_state = {"iteration": metric_iter}
+ self._checkpointer.save(f"{self._file_prefix}", **additional_state)
+ self._logger.info(
+ f"Saved first model at {self.best_metric:0.5f} @ {self.best_iter} steps"
+ )
+ elif self._compare(latest_metric, self.best_metric):
+ additional_state = {"iteration": metric_iter}
+ self._checkpointer.save(f"{self._file_prefix}", **additional_state)
+ self._logger.info(
+ f"Saved best model as latest eval score for {self._val_metric} is "
+ f"{latest_metric:0.5f}, better than last best score "
+ f"{self.best_metric:0.5f} @ iteration {self.best_iter}."
+ )
+ self._update_best(latest_metric, metric_iter)
+ else:
+ self._logger.info(
+ f"Not saving as latest eval score for {self._val_metric} is {latest_metric:0.5f}, "
+ f"not better than best score {self.best_metric:0.5f} @ iteration {self.best_iter}."
+ )
+
+ def after_step(self):
+ # same conditions as `EvalHook`
+ next_iter = self.trainer.iter + 1
+ if (
+ self._period > 0
+ and next_iter % self._period == 0
+ and next_iter != self.trainer.max_iter
+ ):
+ self._best_checking()
+
+ def after_train(self):
+ # same conditions as `EvalHook`
+ if self.trainer.iter + 1 >= self.trainer.max_iter:
+ self._best_checking()
+
+
+class LRScheduler(HookBase):
+ """
+ A hook which executes a torch builtin LR scheduler and summarizes the LR.
+ It is executed after every iteration.
+ """
+
+ def __init__(self, optimizer=None, scheduler=None):
+ """
+ Args:
+ optimizer (torch.optim.Optimizer):
+ scheduler (torch.optim.LRScheduler or fvcore.common.param_scheduler.ParamScheduler):
+ if a :class:`ParamScheduler` object, it defines the multiplier over the base LR
+ in the optimizer.
+
+ If any argument is not given, will try to obtain it from the trainer.
+ """
+ self._optimizer = optimizer
+ self._scheduler = scheduler
+
+ def before_train(self):
+ self._optimizer = self._optimizer or self.trainer.optimizer
+ if isinstance(self.scheduler, ParamScheduler):
+ self._scheduler = LRMultiplier(
+ self._optimizer,
+ self.scheduler,
+ self.trainer.max_iter,
+ last_iter=self.trainer.iter - 1,
+ )
+ self._best_param_group_id = LRScheduler.get_best_param_group_id(self._optimizer)
+
+ @staticmethod
+ def get_best_param_group_id(optimizer):
+ # NOTE: some heuristics on what LR to summarize
+ # summarize the param group with most parameters
+ largest_group = max(len(g["params"]) for g in optimizer.param_groups)
+
+ if largest_group == 1:
+ # If all groups have one parameter,
+ # then find the most common initial LR, and use it for summary
+ lr_count = Counter([g["lr"] for g in optimizer.param_groups])
+ lr = lr_count.most_common()[0][0]
+ for i, g in enumerate(optimizer.param_groups):
+ if g["lr"] == lr:
+ return i
+ else:
+ for i, g in enumerate(optimizer.param_groups):
+ if len(g["params"]) == largest_group:
+ return i
+
+ def after_step(self):
+ lr = self._optimizer.param_groups[self._best_param_group_id]["lr"]
+ self.trainer.storage.put_scalar("lr", lr, smoothing_hint=False)
+ self.scheduler.step()
+
+ @property
+ def scheduler(self):
+ return self._scheduler or self.trainer.scheduler
+
+ def state_dict(self):
+ if isinstance(self.scheduler, _LRScheduler):
+ return self.scheduler.state_dict()
+ return {}
+
+ def load_state_dict(self, state_dict):
+ if isinstance(self.scheduler, _LRScheduler):
+ logger = logging.getLogger(__name__)
+ logger.info("Loading scheduler from state_dict ...")
+ self.scheduler.load_state_dict(state_dict)
+
+
+class TorchProfiler(HookBase):
+ """
+ A hook which runs `torch.profiler.profile`.
+
+ Examples:
+ ::
+ hooks.TorchProfiler(
+ lambda trainer: 10 < trainer.iter < 20, self.cfg.OUTPUT_DIR
+ )
+
+ The above example will run the profiler for iteration 10~20 and dump
+ results to ``OUTPUT_DIR``. We did not profile the first few iterations
+ because they are typically slower than the rest.
+ The result files can be loaded in the ``chrome://tracing`` page in chrome browser,
+ and the tensorboard visualizations can be visualized using
+ ``tensorboard --logdir OUTPUT_DIR/log``
+ """
+
+ def __init__(self, enable_predicate, output_dir, *, activities=None, save_tensorboard=True):
+ """
+ Args:
+ enable_predicate (callable[trainer -> bool]): a function which takes a trainer,
+ and returns whether to enable the profiler.
+ It will be called once every step, and can be used to select which steps to profile.
+ output_dir (str): the output directory to dump tracing files.
+ activities (iterable): same as in `torch.profiler.profile`.
+ save_tensorboard (bool): whether to save tensorboard visualizations at (output_dir)/log/
+ """
+ self._enable_predicate = enable_predicate
+ self._activities = activities
+ self._output_dir = output_dir
+ self._save_tensorboard = save_tensorboard
+
+ def before_step(self):
+ if self._enable_predicate(self.trainer):
+ if self._save_tensorboard:
+ on_trace_ready = torch.profiler.tensorboard_trace_handler(
+ os.path.join(
+ self._output_dir,
+ "log",
+ "profiler-tensorboard-iter{}".format(self.trainer.iter),
+ ),
+ f"worker{comm.get_rank()}",
+ )
+ else:
+ on_trace_ready = None
+ self._profiler = torch.profiler.profile(
+ activities=self._activities,
+ on_trace_ready=on_trace_ready,
+ record_shapes=True,
+ profile_memory=True,
+ with_stack=True,
+ with_flops=True,
+ )
+ self._profiler.__enter__()
+ else:
+ self._profiler = None
+
+ def after_step(self):
+ if self._profiler is None:
+ return
+ self._profiler.__exit__(None, None, None)
+ if not self._save_tensorboard:
+ PathManager.mkdirs(self._output_dir)
+ out_file = os.path.join(
+ self._output_dir, "profiler-trace-iter{}.json".format(self.trainer.iter)
+ )
+ if "://" not in out_file:
+ self._profiler.export_chrome_trace(out_file)
+ else:
+ # Support non-posix filesystems
+ with tempfile.TemporaryDirectory(prefix="detectron2_profiler") as d:
+ tmp_file = os.path.join(d, "tmp.json")
+ self._profiler.export_chrome_trace(tmp_file)
+ with open(tmp_file) as f:
+ content = f.read()
+ with PathManager.open(out_file, "w") as f:
+ f.write(content)
+
+
+class AutogradProfiler(TorchProfiler):
+ """
+ A hook which runs `torch.autograd.profiler.profile`.
+
+ Examples:
+ ::
+ hooks.AutogradProfiler(
+ lambda trainer: 10 < trainer.iter < 20, self.cfg.OUTPUT_DIR
+ )
+
+ The above example will run the profiler for iteration 10~20 and dump
+ results to ``OUTPUT_DIR``. We did not profile the first few iterations
+ because they are typically slower than the rest.
+ The result files can be loaded in the ``chrome://tracing`` page in chrome browser.
+
+ Note:
+ When used together with NCCL on older version of GPUs,
+ autograd profiler may cause deadlock because it unnecessarily allocates
+ memory on every device it sees. The memory management calls, if
+ interleaved with NCCL calls, lead to deadlock on GPUs that do not
+ support ``cudaLaunchCooperativeKernelMultiDevice``.
+ """
+
+ def __init__(self, enable_predicate, output_dir, *, use_cuda=True):
+ """
+ Args:
+ enable_predicate (callable[trainer -> bool]): a function which takes a trainer,
+ and returns whether to enable the profiler.
+ It will be called once every step, and can be used to select which steps to profile.
+ output_dir (str): the output directory to dump tracing files.
+ use_cuda (bool): same as in `torch.autograd.profiler.profile`.
+ """
+ warnings.warn("AutogradProfiler has been deprecated in favor of TorchProfiler.")
+ self._enable_predicate = enable_predicate
+ self._use_cuda = use_cuda
+ self._output_dir = output_dir
+
+ def before_step(self):
+ if self._enable_predicate(self.trainer):
+ self._profiler = torch.autograd.profiler.profile(use_cuda=self._use_cuda)
+ self._profiler.__enter__()
+ else:
+ self._profiler = None
+
+
+class EvalHook(HookBase):
+ """
+ Run an evaluation function periodically, and at the end of training.
+
+ It is executed every ``eval_period`` iterations and after the last iteration.
+ """
+
+ def __init__(self, eval_period, eval_function, eval_after_train=True):
+ """
+ Args:
+ eval_period (int): the period to run `eval_function`. Set to 0 to
+ not evaluate periodically (but still evaluate after the last iteration
+ if `eval_after_train` is True).
+ eval_function (callable): a function which takes no arguments, and
+ returns a nested dict of evaluation metrics.
+ eval_after_train (bool): whether to evaluate after the last iteration
+
+ Note:
+ This hook must be enabled in all or none workers.
+ If you would like only certain workers to perform evaluation,
+ give other workers a no-op function (`eval_function=lambda: None`).
+ """
+ self._period = eval_period
+ self._func = eval_function
+ self._eval_after_train = eval_after_train
+
+ def _do_eval(self):
+ results = self._func()
+
+ if results:
+ assert isinstance(
+ results, dict
+ ), "Eval function must return a dict. Got {} instead.".format(results)
+
+ flattened_results = flatten_results_dict(results)
+ for k, v in flattened_results.items():
+ try:
+ v = float(v)
+ except Exception as e:
+ raise ValueError(
+ "[EvalHook] eval_function should return a nested dict of float. "
+ "Got '{}: {}' instead.".format(k, v)
+ ) from e
+ self.trainer.storage.put_scalars(**flattened_results, smoothing_hint=False)
+
+ # Evaluation may take different time among workers.
+ # A barrier make them start the next iteration together.
+ comm.synchronize()
+
+ def after_step(self):
+ next_iter = self.trainer.iter + 1
+ if self._period > 0 and next_iter % self._period == 0:
+ # do the last eval in after_train
+ if next_iter != self.trainer.max_iter:
+ self._do_eval()
+
+ def after_train(self):
+ # This condition is to prevent the eval from running after a failed training
+ if self._eval_after_train and self.trainer.iter + 1 >= self.trainer.max_iter:
+ self._do_eval()
+ # func is likely a closure that holds reference to the trainer
+ # therefore we clean it to avoid circular reference in the end
+ del self._func
+
+
+class PreciseBN(HookBase):
+ """
+ The standard implementation of BatchNorm uses EMA in inference, which is
+ sometimes suboptimal.
+ This class computes the true average of statistics rather than the moving average,
+ and put true averages to every BN layer in the given model.
+
+ It is executed every ``period`` iterations and after the last iteration.
+ """
+
+ def __init__(self, period, model, data_loader, num_iter):
+ """
+ Args:
+ period (int): the period this hook is run, or 0 to not run during training.
+ The hook will always run in the end of training.
+ model (nn.Module): a module whose all BN layers in training mode will be
+ updated by precise BN.
+ Note that user is responsible for ensuring the BN layers to be
+ updated are in training mode when this hook is triggered.
+ data_loader (iterable): it will produce data to be run by `model(data)`.
+ num_iter (int): number of iterations used to compute the precise
+ statistics.
+ """
+ self._logger = logging.getLogger(__name__)
+ if len(get_bn_modules(model)) == 0:
+ self._logger.info(
+ "PreciseBN is disabled because model does not contain BN layers in training mode."
+ )
+ self._disabled = True
+ return
+
+ self._model = model
+ self._data_loader = data_loader
+ self._num_iter = num_iter
+ self._period = period
+ self._disabled = False
+
+ self._data_iter = None
+
+ def after_step(self):
+ next_iter = self.trainer.iter + 1
+ is_final = next_iter == self.trainer.max_iter
+ if is_final or (self._period > 0 and next_iter % self._period == 0):
+ self.update_stats()
+
+ def update_stats(self):
+ """
+ Update the model with precise statistics. Users can manually call this method.
+ """
+ if self._disabled:
+ return
+
+ if self._data_iter is None:
+ self._data_iter = iter(self._data_loader)
+
+ def data_loader():
+ for num_iter in itertools.count(1):
+ if num_iter % 100 == 0:
+ self._logger.info(
+ "Running precise-BN ... {}/{} iterations.".format(num_iter, self._num_iter)
+ )
+ # This way we can reuse the same iterator
+ yield next(self._data_iter)
+
+ with EventStorage(): # capture events in a new storage to discard them
+ self._logger.info(
+ "Running precise-BN for {} iterations... ".format(self._num_iter)
+ + "Note that this could produce different statistics every time."
+ )
+ update_bn_stats(self._model, data_loader(), self._num_iter)
+
+
+class TorchMemoryStats(HookBase):
+ """
+ Writes pytorch's cuda memory statistics periodically.
+ """
+
+ def __init__(self, period=20, max_runs=10):
+ """
+ Args:
+ period (int): Output stats each 'period' iterations
+ max_runs (int): Stop the logging after 'max_runs'
+ """
+
+ self._logger = logging.getLogger(__name__)
+ self._period = period
+ self._max_runs = max_runs
+ self._runs = 0
+
+ def after_step(self):
+ if self._runs > self._max_runs:
+ return
+
+ if (self.trainer.iter + 1) % self._period == 0 or (
+ self.trainer.iter == self.trainer.max_iter - 1
+ ):
+ if torch.cuda.is_available():
+ max_reserved_mb = torch.cuda.max_memory_reserved() / 1024.0 / 1024.0
+ reserved_mb = torch.cuda.memory_reserved() / 1024.0 / 1024.0
+ max_allocated_mb = torch.cuda.max_memory_allocated() / 1024.0 / 1024.0
+ allocated_mb = torch.cuda.memory_allocated() / 1024.0 / 1024.0
+
+ self._logger.info(
+ (
+ " iter: {} "
+ " max_reserved_mem: {:.0f}MB "
+ " reserved_mem: {:.0f}MB "
+ " max_allocated_mem: {:.0f}MB "
+ " allocated_mem: {:.0f}MB "
+ ).format(
+ self.trainer.iter,
+ max_reserved_mb,
+ reserved_mb,
+ max_allocated_mb,
+ allocated_mb,
+ )
+ )
+
+ self._runs += 1
+ if self._runs == self._max_runs:
+ mem_summary = torch.cuda.memory_summary()
+ self._logger.info("\n" + mem_summary)
+
+ torch.cuda.reset_peak_memory_stats()
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/engine/launch.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/engine/launch.py
new file mode 100644
index 0000000000000000000000000000000000000000..0a2d6bcdb5f1906d3eedb04b5aa939f8269f0344
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/engine/launch.py
@@ -0,0 +1,123 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import logging
+from datetime import timedelta
+import torch
+import torch.distributed as dist
+import torch.multiprocessing as mp
+
+from annotator.oneformer.detectron2.utils import comm
+
+__all__ = ["DEFAULT_TIMEOUT", "launch"]
+
+DEFAULT_TIMEOUT = timedelta(minutes=30)
+
+
+def _find_free_port():
+ import socket
+
+ sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
+ # Binding to port 0 will cause the OS to find an available port for us
+ sock.bind(("", 0))
+ port = sock.getsockname()[1]
+ sock.close()
+ # NOTE: there is still a chance the port could be taken by other processes.
+ return port
+
+
+def launch(
+ main_func,
+ # Should be num_processes_per_machine, but kept for compatibility.
+ num_gpus_per_machine,
+ num_machines=1,
+ machine_rank=0,
+ dist_url=None,
+ args=(),
+ timeout=DEFAULT_TIMEOUT,
+):
+ """
+ Launch multi-process or distributed training.
+ This function must be called on all machines involved in the training.
+ It will spawn child processes (defined by ``num_gpus_per_machine``) on each machine.
+
+ Args:
+ main_func: a function that will be called by `main_func(*args)`
+ num_gpus_per_machine (int): number of processes per machine. When
+ using GPUs, this should be the number of GPUs.
+ num_machines (int): the total number of machines
+ machine_rank (int): the rank of this machine
+ dist_url (str): url to connect to for distributed jobs, including protocol
+ e.g. "tcp://127.0.0.1:8686".
+ Can be set to "auto" to automatically select a free port on localhost
+ timeout (timedelta): timeout of the distributed workers
+ args (tuple): arguments passed to main_func
+ """
+ world_size = num_machines * num_gpus_per_machine
+ if world_size > 1:
+ # https://github.com/pytorch/pytorch/pull/14391
+ # TODO prctl in spawned processes
+
+ if dist_url == "auto":
+ assert num_machines == 1, "dist_url=auto not supported in multi-machine jobs."
+ port = _find_free_port()
+ dist_url = f"tcp://127.0.0.1:{port}"
+ if num_machines > 1 and dist_url.startswith("file://"):
+ logger = logging.getLogger(__name__)
+ logger.warning(
+ "file:// is not a reliable init_method in multi-machine jobs. Prefer tcp://"
+ )
+
+ mp.start_processes(
+ _distributed_worker,
+ nprocs=num_gpus_per_machine,
+ args=(
+ main_func,
+ world_size,
+ num_gpus_per_machine,
+ machine_rank,
+ dist_url,
+ args,
+ timeout,
+ ),
+ daemon=False,
+ )
+ else:
+ main_func(*args)
+
+
+def _distributed_worker(
+ local_rank,
+ main_func,
+ world_size,
+ num_gpus_per_machine,
+ machine_rank,
+ dist_url,
+ args,
+ timeout=DEFAULT_TIMEOUT,
+):
+ has_gpu = torch.cuda.is_available()
+ if has_gpu:
+ assert num_gpus_per_machine <= torch.cuda.device_count()
+ global_rank = machine_rank * num_gpus_per_machine + local_rank
+ try:
+ dist.init_process_group(
+ backend="NCCL" if has_gpu else "GLOO",
+ init_method=dist_url,
+ world_size=world_size,
+ rank=global_rank,
+ timeout=timeout,
+ )
+ except Exception as e:
+ logger = logging.getLogger(__name__)
+ logger.error("Process group URL: {}".format(dist_url))
+ raise e
+
+ # Setup the local process group.
+ comm.create_local_process_group(num_gpus_per_machine)
+ if has_gpu:
+ torch.cuda.set_device(local_rank)
+
+ # synchronize is needed here to prevent a possible timeout after calling init_process_group
+ # See: https://github.com/facebookresearch/maskrcnn-benchmark/issues/172
+ comm.synchronize()
+
+ main_func(*args)
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/engine/train_loop.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/engine/train_loop.py
new file mode 100644
index 0000000000000000000000000000000000000000..0c24c5af94e8f9367a5d577a617ec426292d3f89
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/engine/train_loop.py
@@ -0,0 +1,469 @@
+# -*- coding: utf-8 -*-
+# Copyright (c) Facebook, Inc. and its affiliates.
+
+import logging
+import numpy as np
+import time
+import weakref
+from typing import List, Mapping, Optional
+import torch
+from torch.nn.parallel import DataParallel, DistributedDataParallel
+
+import annotator.oneformer.detectron2.utils.comm as comm
+from annotator.oneformer.detectron2.utils.events import EventStorage, get_event_storage
+from annotator.oneformer.detectron2.utils.logger import _log_api_usage
+
+__all__ = ["HookBase", "TrainerBase", "SimpleTrainer", "AMPTrainer"]
+
+
+class HookBase:
+ """
+ Base class for hooks that can be registered with :class:`TrainerBase`.
+
+ Each hook can implement 4 methods. The way they are called is demonstrated
+ in the following snippet:
+ ::
+ hook.before_train()
+ for iter in range(start_iter, max_iter):
+ hook.before_step()
+ trainer.run_step()
+ hook.after_step()
+ iter += 1
+ hook.after_train()
+
+ Notes:
+ 1. In the hook method, users can access ``self.trainer`` to access more
+ properties about the context (e.g., model, current iteration, or config
+ if using :class:`DefaultTrainer`).
+
+ 2. A hook that does something in :meth:`before_step` can often be
+ implemented equivalently in :meth:`after_step`.
+ If the hook takes non-trivial time, it is strongly recommended to
+ implement the hook in :meth:`after_step` instead of :meth:`before_step`.
+ The convention is that :meth:`before_step` should only take negligible time.
+
+ Following this convention will allow hooks that do care about the difference
+ between :meth:`before_step` and :meth:`after_step` (e.g., timer) to
+ function properly.
+
+ """
+
+ trainer: "TrainerBase" = None
+ """
+ A weak reference to the trainer object. Set by the trainer when the hook is registered.
+ """
+
+ def before_train(self):
+ """
+ Called before the first iteration.
+ """
+ pass
+
+ def after_train(self):
+ """
+ Called after the last iteration.
+ """
+ pass
+
+ def before_step(self):
+ """
+ Called before each iteration.
+ """
+ pass
+
+ def after_backward(self):
+ """
+ Called after the backward pass of each iteration.
+ """
+ pass
+
+ def after_step(self):
+ """
+ Called after each iteration.
+ """
+ pass
+
+ def state_dict(self):
+ """
+ Hooks are stateless by default, but can be made checkpointable by
+ implementing `state_dict` and `load_state_dict`.
+ """
+ return {}
+
+
+class TrainerBase:
+ """
+ Base class for iterative trainer with hooks.
+
+ The only assumption we made here is: the training runs in a loop.
+ A subclass can implement what the loop is.
+ We made no assumptions about the existence of dataloader, optimizer, model, etc.
+
+ Attributes:
+ iter(int): the current iteration.
+
+ start_iter(int): The iteration to start with.
+ By convention the minimum possible value is 0.
+
+ max_iter(int): The iteration to end training.
+
+ storage(EventStorage): An EventStorage that's opened during the course of training.
+ """
+
+ def __init__(self) -> None:
+ self._hooks: List[HookBase] = []
+ self.iter: int = 0
+ self.start_iter: int = 0
+ self.max_iter: int
+ self.storage: EventStorage
+ _log_api_usage("trainer." + self.__class__.__name__)
+
+ def register_hooks(self, hooks: List[Optional[HookBase]]) -> None:
+ """
+ Register hooks to the trainer. The hooks are executed in the order
+ they are registered.
+
+ Args:
+ hooks (list[Optional[HookBase]]): list of hooks
+ """
+ hooks = [h for h in hooks if h is not None]
+ for h in hooks:
+ assert isinstance(h, HookBase)
+ # To avoid circular reference, hooks and trainer cannot own each other.
+ # This normally does not matter, but will cause memory leak if the
+ # involved objects contain __del__:
+ # See http://engineering.hearsaysocial.com/2013/06/16/circular-references-in-python/
+ h.trainer = weakref.proxy(self)
+ self._hooks.extend(hooks)
+
+ def train(self, start_iter: int, max_iter: int):
+ """
+ Args:
+ start_iter, max_iter (int): See docs above
+ """
+ logger = logging.getLogger(__name__)
+ logger.info("Starting training from iteration {}".format(start_iter))
+
+ self.iter = self.start_iter = start_iter
+ self.max_iter = max_iter
+
+ with EventStorage(start_iter) as self.storage:
+ try:
+ self.before_train()
+ for self.iter in range(start_iter, max_iter):
+ self.before_step()
+ self.run_step()
+ self.after_step()
+ # self.iter == max_iter can be used by `after_train` to
+ # tell whether the training successfully finished or failed
+ # due to exceptions.
+ self.iter += 1
+ except Exception:
+ logger.exception("Exception during training:")
+ raise
+ finally:
+ self.after_train()
+
+ def before_train(self):
+ for h in self._hooks:
+ h.before_train()
+
+ def after_train(self):
+ self.storage.iter = self.iter
+ for h in self._hooks:
+ h.after_train()
+
+ def before_step(self):
+ # Maintain the invariant that storage.iter == trainer.iter
+ # for the entire execution of each step
+ self.storage.iter = self.iter
+
+ for h in self._hooks:
+ h.before_step()
+
+ def after_backward(self):
+ for h in self._hooks:
+ h.after_backward()
+
+ def after_step(self):
+ for h in self._hooks:
+ h.after_step()
+
+ def run_step(self):
+ raise NotImplementedError
+
+ def state_dict(self):
+ ret = {"iteration": self.iter}
+ hooks_state = {}
+ for h in self._hooks:
+ sd = h.state_dict()
+ if sd:
+ name = type(h).__qualname__
+ if name in hooks_state:
+ # TODO handle repetitive stateful hooks
+ continue
+ hooks_state[name] = sd
+ if hooks_state:
+ ret["hooks"] = hooks_state
+ return ret
+
+ def load_state_dict(self, state_dict):
+ logger = logging.getLogger(__name__)
+ self.iter = state_dict["iteration"]
+ for key, value in state_dict.get("hooks", {}).items():
+ for h in self._hooks:
+ try:
+ name = type(h).__qualname__
+ except AttributeError:
+ continue
+ if name == key:
+ h.load_state_dict(value)
+ break
+ else:
+ logger.warning(f"Cannot find the hook '{key}', its state_dict is ignored.")
+
+
+class SimpleTrainer(TrainerBase):
+ """
+ A simple trainer for the most common type of task:
+ single-cost single-optimizer single-data-source iterative optimization,
+ optionally using data-parallelism.
+ It assumes that every step, you:
+
+ 1. Compute the loss with a data from the data_loader.
+ 2. Compute the gradients with the above loss.
+ 3. Update the model with the optimizer.
+
+ All other tasks during training (checkpointing, logging, evaluation, LR schedule)
+ are maintained by hooks, which can be registered by :meth:`TrainerBase.register_hooks`.
+
+ If you want to do anything fancier than this,
+ either subclass TrainerBase and implement your own `run_step`,
+ or write your own training loop.
+ """
+
+ def __init__(self, model, data_loader, optimizer, gather_metric_period=1):
+ """
+ Args:
+ model: a torch Module. Takes a data from data_loader and returns a
+ dict of losses.
+ data_loader: an iterable. Contains data to be used to call model.
+ optimizer: a torch optimizer.
+ gather_metric_period: an int. Every gather_metric_period iterations
+ the metrics are gathered from all the ranks to rank 0 and logged.
+ """
+ super().__init__()
+
+ """
+ We set the model to training mode in the trainer.
+ However it's valid to train a model that's in eval mode.
+ If you want your model (or a submodule of it) to behave
+ like evaluation during training, you can overwrite its train() method.
+ """
+ model.train()
+
+ self.model = model
+ self.data_loader = data_loader
+ # to access the data loader iterator, call `self._data_loader_iter`
+ self._data_loader_iter_obj = None
+ self.optimizer = optimizer
+ self.gather_metric_period = gather_metric_period
+
+ def run_step(self):
+ """
+ Implement the standard training logic described above.
+ """
+ assert self.model.training, "[SimpleTrainer] model was changed to eval mode!"
+ start = time.perf_counter()
+ """
+ If you want to do something with the data, you can wrap the dataloader.
+ """
+ data = next(self._data_loader_iter)
+ data_time = time.perf_counter() - start
+
+ """
+ If you want to do something with the losses, you can wrap the model.
+ """
+ loss_dict = self.model(data)
+ if isinstance(loss_dict, torch.Tensor):
+ losses = loss_dict
+ loss_dict = {"total_loss": loss_dict}
+ else:
+ losses = sum(loss_dict.values())
+
+ """
+ If you need to accumulate gradients or do something similar, you can
+ wrap the optimizer with your custom `zero_grad()` method.
+ """
+ self.optimizer.zero_grad()
+ losses.backward()
+
+ self.after_backward()
+
+ self._write_metrics(loss_dict, data_time)
+
+ """
+ If you need gradient clipping/scaling or other processing, you can
+ wrap the optimizer with your custom `step()` method. But it is
+ suboptimal as explained in https://arxiv.org/abs/2006.15704 Sec 3.2.4
+ """
+ self.optimizer.step()
+
+ @property
+ def _data_loader_iter(self):
+ # only create the data loader iterator when it is used
+ if self._data_loader_iter_obj is None:
+ self._data_loader_iter_obj = iter(self.data_loader)
+ return self._data_loader_iter_obj
+
+ def reset_data_loader(self, data_loader_builder):
+ """
+ Delete and replace the current data loader with a new one, which will be created
+ by calling `data_loader_builder` (without argument).
+ """
+ del self.data_loader
+ data_loader = data_loader_builder()
+ self.data_loader = data_loader
+ self._data_loader_iter_obj = None
+
+ def _write_metrics(
+ self,
+ loss_dict: Mapping[str, torch.Tensor],
+ data_time: float,
+ prefix: str = "",
+ ) -> None:
+ if (self.iter + 1) % self.gather_metric_period == 0:
+ SimpleTrainer.write_metrics(loss_dict, data_time, prefix)
+
+ @staticmethod
+ def write_metrics(
+ loss_dict: Mapping[str, torch.Tensor],
+ data_time: float,
+ prefix: str = "",
+ ) -> None:
+ """
+ Args:
+ loss_dict (dict): dict of scalar losses
+ data_time (float): time taken by the dataloader iteration
+ prefix (str): prefix for logging keys
+ """
+ metrics_dict = {k: v.detach().cpu().item() for k, v in loss_dict.items()}
+ metrics_dict["data_time"] = data_time
+
+ # Gather metrics among all workers for logging
+ # This assumes we do DDP-style training, which is currently the only
+ # supported method in detectron2.
+ all_metrics_dict = comm.gather(metrics_dict)
+
+ if comm.is_main_process():
+ storage = get_event_storage()
+
+ # data_time among workers can have high variance. The actual latency
+ # caused by data_time is the maximum among workers.
+ data_time = np.max([x.pop("data_time") for x in all_metrics_dict])
+ storage.put_scalar("data_time", data_time)
+
+ # average the rest metrics
+ metrics_dict = {
+ k: np.mean([x[k] for x in all_metrics_dict]) for k in all_metrics_dict[0].keys()
+ }
+ total_losses_reduced = sum(metrics_dict.values())
+ if not np.isfinite(total_losses_reduced):
+ raise FloatingPointError(
+ f"Loss became infinite or NaN at iteration={storage.iter}!\n"
+ f"loss_dict = {metrics_dict}"
+ )
+
+ storage.put_scalar("{}total_loss".format(prefix), total_losses_reduced)
+ if len(metrics_dict) > 1:
+ storage.put_scalars(**metrics_dict)
+
+ def state_dict(self):
+ ret = super().state_dict()
+ ret["optimizer"] = self.optimizer.state_dict()
+ return ret
+
+ def load_state_dict(self, state_dict):
+ super().load_state_dict(state_dict)
+ self.optimizer.load_state_dict(state_dict["optimizer"])
+
+
+class AMPTrainer(SimpleTrainer):
+ """
+ Like :class:`SimpleTrainer`, but uses PyTorch's native automatic mixed precision
+ in the training loop.
+ """
+
+ def __init__(
+ self,
+ model,
+ data_loader,
+ optimizer,
+ gather_metric_period=1,
+ grad_scaler=None,
+ precision: torch.dtype = torch.float16,
+ log_grad_scaler: bool = False,
+ ):
+ """
+ Args:
+ model, data_loader, optimizer, gather_metric_period: same as in :class:`SimpleTrainer`.
+ grad_scaler: torch GradScaler to automatically scale gradients.
+ precision: torch.dtype as the target precision to cast to in computations
+ """
+ unsupported = "AMPTrainer does not support single-process multi-device training!"
+ if isinstance(model, DistributedDataParallel):
+ assert not (model.device_ids and len(model.device_ids) > 1), unsupported
+ assert not isinstance(model, DataParallel), unsupported
+
+ super().__init__(model, data_loader, optimizer, gather_metric_period)
+
+ if grad_scaler is None:
+ from torch.cuda.amp import GradScaler
+
+ grad_scaler = GradScaler()
+ self.grad_scaler = grad_scaler
+ self.precision = precision
+ self.log_grad_scaler = log_grad_scaler
+
+ def run_step(self):
+ """
+ Implement the AMP training logic.
+ """
+ assert self.model.training, "[AMPTrainer] model was changed to eval mode!"
+ assert torch.cuda.is_available(), "[AMPTrainer] CUDA is required for AMP training!"
+ from torch.cuda.amp import autocast
+
+ start = time.perf_counter()
+ data = next(self._data_loader_iter)
+ data_time = time.perf_counter() - start
+
+ with autocast(dtype=self.precision):
+ loss_dict = self.model(data)
+ if isinstance(loss_dict, torch.Tensor):
+ losses = loss_dict
+ loss_dict = {"total_loss": loss_dict}
+ else:
+ losses = sum(loss_dict.values())
+
+ self.optimizer.zero_grad()
+ self.grad_scaler.scale(losses).backward()
+
+ if self.log_grad_scaler:
+ storage = get_event_storage()
+ storage.put_scalar("[metric]grad_scaler", self.grad_scaler.get_scale())
+
+ self.after_backward()
+
+ self._write_metrics(loss_dict, data_time)
+
+ self.grad_scaler.step(self.optimizer)
+ self.grad_scaler.update()
+
+ def state_dict(self):
+ ret = super().state_dict()
+ ret["grad_scaler"] = self.grad_scaler.state_dict()
+ return ret
+
+ def load_state_dict(self, state_dict):
+ super().load_state_dict(state_dict)
+ self.grad_scaler.load_state_dict(state_dict["grad_scaler"])
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/__init__.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..d96609e8f2261a6800fe85fcf3e1eaeaa44455c6
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/__init__.py
@@ -0,0 +1,12 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+from .cityscapes_evaluation import CityscapesInstanceEvaluator, CityscapesSemSegEvaluator
+from .coco_evaluation import COCOEvaluator
+from .rotated_coco_evaluation import RotatedCOCOEvaluator
+from .evaluator import DatasetEvaluator, DatasetEvaluators, inference_context, inference_on_dataset
+from .lvis_evaluation import LVISEvaluator
+from .panoptic_evaluation import COCOPanopticEvaluator
+from .pascal_voc_evaluation import PascalVOCDetectionEvaluator
+from .sem_seg_evaluation import SemSegEvaluator
+from .testing import print_csv_format, verify_results
+
+__all__ = [k for k in globals().keys() if not k.startswith("_")]
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/cityscapes_evaluation.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/cityscapes_evaluation.py
new file mode 100644
index 0000000000000000000000000000000000000000..f5be637dc87b5ca8645563a4a921144f6c5fd877
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/cityscapes_evaluation.py
@@ -0,0 +1,197 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import glob
+import logging
+import numpy as np
+import os
+import tempfile
+from collections import OrderedDict
+import torch
+from PIL import Image
+
+from annotator.oneformer.detectron2.data import MetadataCatalog
+from annotator.oneformer.detectron2.utils import comm
+from annotator.oneformer.detectron2.utils.file_io import PathManager
+
+from .evaluator import DatasetEvaluator
+
+
+class CityscapesEvaluator(DatasetEvaluator):
+ """
+ Base class for evaluation using cityscapes API.
+ """
+
+ def __init__(self, dataset_name):
+ """
+ Args:
+ dataset_name (str): the name of the dataset.
+ It must have the following metadata associated with it:
+ "thing_classes", "gt_dir".
+ """
+ self._metadata = MetadataCatalog.get(dataset_name)
+ self._cpu_device = torch.device("cpu")
+ self._logger = logging.getLogger(__name__)
+
+ def reset(self):
+ self._working_dir = tempfile.TemporaryDirectory(prefix="cityscapes_eval_")
+ self._temp_dir = self._working_dir.name
+ # All workers will write to the same results directory
+ # TODO this does not work in distributed training
+ assert (
+ comm.get_local_size() == comm.get_world_size()
+ ), "CityscapesEvaluator currently do not work with multiple machines."
+ self._temp_dir = comm.all_gather(self._temp_dir)[0]
+ if self._temp_dir != self._working_dir.name:
+ self._working_dir.cleanup()
+ self._logger.info(
+ "Writing cityscapes results to temporary directory {} ...".format(self._temp_dir)
+ )
+
+
+class CityscapesInstanceEvaluator(CityscapesEvaluator):
+ """
+ Evaluate instance segmentation results on cityscapes dataset using cityscapes API.
+
+ Note:
+ * It does not work in multi-machine distributed training.
+ * It contains a synchronization, therefore has to be used on all ranks.
+ * Only the main process runs evaluation.
+ """
+
+ def process(self, inputs, outputs):
+ from cityscapesscripts.helpers.labels import name2label
+
+ for input, output in zip(inputs, outputs):
+ file_name = input["file_name"]
+ basename = os.path.splitext(os.path.basename(file_name))[0]
+ pred_txt = os.path.join(self._temp_dir, basename + "_pred.txt")
+
+ if "instances" in output:
+ output = output["instances"].to(self._cpu_device)
+ num_instances = len(output)
+ with open(pred_txt, "w") as fout:
+ for i in range(num_instances):
+ pred_class = output.pred_classes[i]
+ classes = self._metadata.thing_classes[pred_class]
+ class_id = name2label[classes].id
+ score = output.scores[i]
+ mask = output.pred_masks[i].numpy().astype("uint8")
+ png_filename = os.path.join(
+ self._temp_dir, basename + "_{}_{}.png".format(i, classes)
+ )
+
+ Image.fromarray(mask * 255).save(png_filename)
+ fout.write(
+ "{} {} {}\n".format(os.path.basename(png_filename), class_id, score)
+ )
+ else:
+ # Cityscapes requires a prediction file for every ground truth image.
+ with open(pred_txt, "w") as fout:
+ pass
+
+ def evaluate(self):
+ """
+ Returns:
+ dict: has a key "segm", whose value is a dict of "AP" and "AP50".
+ """
+ comm.synchronize()
+ if comm.get_rank() > 0:
+ return
+ import cityscapesscripts.evaluation.evalInstanceLevelSemanticLabeling as cityscapes_eval
+
+ self._logger.info("Evaluating results under {} ...".format(self._temp_dir))
+
+ # set some global states in cityscapes evaluation API, before evaluating
+ cityscapes_eval.args.predictionPath = os.path.abspath(self._temp_dir)
+ cityscapes_eval.args.predictionWalk = None
+ cityscapes_eval.args.JSONOutput = False
+ cityscapes_eval.args.colorized = False
+ cityscapes_eval.args.gtInstancesFile = os.path.join(self._temp_dir, "gtInstances.json")
+
+ # These lines are adopted from
+ # https://github.com/mcordts/cityscapesScripts/blob/master/cityscapesscripts/evaluation/evalInstanceLevelSemanticLabeling.py # noqa
+ gt_dir = PathManager.get_local_path(self._metadata.gt_dir)
+ groundTruthImgList = glob.glob(os.path.join(gt_dir, "*", "*_gtFine_instanceIds.png"))
+ assert len(
+ groundTruthImgList
+ ), "Cannot find any ground truth images to use for evaluation. Searched for: {}".format(
+ cityscapes_eval.args.groundTruthSearch
+ )
+ predictionImgList = []
+ for gt in groundTruthImgList:
+ predictionImgList.append(cityscapes_eval.getPrediction(gt, cityscapes_eval.args))
+ results = cityscapes_eval.evaluateImgLists(
+ predictionImgList, groundTruthImgList, cityscapes_eval.args
+ )["averages"]
+
+ ret = OrderedDict()
+ ret["segm"] = {"AP": results["allAp"] * 100, "AP50": results["allAp50%"] * 100}
+ self._working_dir.cleanup()
+ return ret
+
+
+class CityscapesSemSegEvaluator(CityscapesEvaluator):
+ """
+ Evaluate semantic segmentation results on cityscapes dataset using cityscapes API.
+
+ Note:
+ * It does not work in multi-machine distributed training.
+ * It contains a synchronization, therefore has to be used on all ranks.
+ * Only the main process runs evaluation.
+ """
+
+ def process(self, inputs, outputs):
+ from cityscapesscripts.helpers.labels import trainId2label
+
+ for input, output in zip(inputs, outputs):
+ file_name = input["file_name"]
+ basename = os.path.splitext(os.path.basename(file_name))[0]
+ pred_filename = os.path.join(self._temp_dir, basename + "_pred.png")
+
+ output = output["sem_seg"].argmax(dim=0).to(self._cpu_device).numpy()
+ pred = 255 * np.ones(output.shape, dtype=np.uint8)
+ for train_id, label in trainId2label.items():
+ if label.ignoreInEval:
+ continue
+ pred[output == train_id] = label.id
+ Image.fromarray(pred).save(pred_filename)
+
+ def evaluate(self):
+ comm.synchronize()
+ if comm.get_rank() > 0:
+ return
+ # Load the Cityscapes eval script *after* setting the required env var,
+ # since the script reads CITYSCAPES_DATASET into global variables at load time.
+ import cityscapesscripts.evaluation.evalPixelLevelSemanticLabeling as cityscapes_eval
+
+ self._logger.info("Evaluating results under {} ...".format(self._temp_dir))
+
+ # set some global states in cityscapes evaluation API, before evaluating
+ cityscapes_eval.args.predictionPath = os.path.abspath(self._temp_dir)
+ cityscapes_eval.args.predictionWalk = None
+ cityscapes_eval.args.JSONOutput = False
+ cityscapes_eval.args.colorized = False
+
+ # These lines are adopted from
+ # https://github.com/mcordts/cityscapesScripts/blob/master/cityscapesscripts/evaluation/evalPixelLevelSemanticLabeling.py # noqa
+ gt_dir = PathManager.get_local_path(self._metadata.gt_dir)
+ groundTruthImgList = glob.glob(os.path.join(gt_dir, "*", "*_gtFine_labelIds.png"))
+ assert len(
+ groundTruthImgList
+ ), "Cannot find any ground truth images to use for evaluation. Searched for: {}".format(
+ cityscapes_eval.args.groundTruthSearch
+ )
+ predictionImgList = []
+ for gt in groundTruthImgList:
+ predictionImgList.append(cityscapes_eval.getPrediction(cityscapes_eval.args, gt))
+ results = cityscapes_eval.evaluateImgLists(
+ predictionImgList, groundTruthImgList, cityscapes_eval.args
+ )
+ ret = OrderedDict()
+ ret["sem_seg"] = {
+ "IoU": 100.0 * results["averageScoreClasses"],
+ "iIoU": 100.0 * results["averageScoreInstClasses"],
+ "IoU_sup": 100.0 * results["averageScoreCategories"],
+ "iIoU_sup": 100.0 * results["averageScoreInstCategories"],
+ }
+ self._working_dir.cleanup()
+ return ret
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/coco_evaluation.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/coco_evaluation.py
new file mode 100644
index 0000000000000000000000000000000000000000..1eef5ce6f688a749cfa35a389f6599f10df79c22
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/coco_evaluation.py
@@ -0,0 +1,722 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import contextlib
+import copy
+import io
+import itertools
+import json
+import logging
+import numpy as np
+import os
+import pickle
+from collections import OrderedDict
+import pycocotools.mask as mask_util
+import torch
+from pycocotools.coco import COCO
+from pycocotools.cocoeval import COCOeval
+from tabulate import tabulate
+
+import annotator.oneformer.detectron2.utils.comm as comm
+from annotator.oneformer.detectron2.config import CfgNode
+from annotator.oneformer.detectron2.data import MetadataCatalog
+from annotator.oneformer.detectron2.data.datasets.coco import convert_to_coco_json
+from annotator.oneformer.detectron2.structures import Boxes, BoxMode, pairwise_iou
+from annotator.oneformer.detectron2.utils.file_io import PathManager
+from annotator.oneformer.detectron2.utils.logger import create_small_table
+
+from .evaluator import DatasetEvaluator
+
+try:
+ from annotator.oneformer.detectron2.evaluation.fast_eval_api import COCOeval_opt
+except ImportError:
+ COCOeval_opt = COCOeval
+
+
+class COCOEvaluator(DatasetEvaluator):
+ """
+ Evaluate AR for object proposals, AP for instance detection/segmentation, AP
+ for keypoint detection outputs using COCO's metrics.
+ See http://cocodataset.org/#detection-eval and
+ http://cocodataset.org/#keypoints-eval to understand its metrics.
+ The metrics range from 0 to 100 (instead of 0 to 1), where a -1 or NaN means
+ the metric cannot be computed (e.g. due to no predictions made).
+
+ In addition to COCO, this evaluator is able to support any bounding box detection,
+ instance segmentation, or keypoint detection dataset.
+ """
+
+ def __init__(
+ self,
+ dataset_name,
+ tasks=None,
+ distributed=True,
+ output_dir=None,
+ *,
+ max_dets_per_image=None,
+ use_fast_impl=True,
+ kpt_oks_sigmas=(),
+ allow_cached_coco=True,
+ ):
+ """
+ Args:
+ dataset_name (str): name of the dataset to be evaluated.
+ It must have either the following corresponding metadata:
+
+ "json_file": the path to the COCO format annotation
+
+ Or it must be in detectron2's standard dataset format
+ so it can be converted to COCO format automatically.
+ tasks (tuple[str]): tasks that can be evaluated under the given
+ configuration. A task is one of "bbox", "segm", "keypoints".
+ By default, will infer this automatically from predictions.
+ distributed (True): if True, will collect results from all ranks and run evaluation
+ in the main process.
+ Otherwise, will only evaluate the results in the current process.
+ output_dir (str): optional, an output directory to dump all
+ results predicted on the dataset. The dump contains two files:
+
+ 1. "instances_predictions.pth" a file that can be loaded with `torch.load` and
+ contains all the results in the format they are produced by the model.
+ 2. "coco_instances_results.json" a json file in COCO's result format.
+ max_dets_per_image (int): limit on the maximum number of detections per image.
+ By default in COCO, this limit is to 100, but this can be customized
+ to be greater, as is needed in evaluation metrics AP fixed and AP pool
+ (see https://arxiv.org/pdf/2102.01066.pdf)
+ This doesn't affect keypoint evaluation.
+ use_fast_impl (bool): use a fast but **unofficial** implementation to compute AP.
+ Although the results should be very close to the official implementation in COCO
+ API, it is still recommended to compute results with the official API for use in
+ papers. The faster implementation also uses more RAM.
+ kpt_oks_sigmas (list[float]): The sigmas used to calculate keypoint OKS.
+ See http://cocodataset.org/#keypoints-eval
+ When empty, it will use the defaults in COCO.
+ Otherwise it should be the same length as ROI_KEYPOINT_HEAD.NUM_KEYPOINTS.
+ allow_cached_coco (bool): Whether to use cached coco json from previous validation
+ runs. You should set this to False if you need to use different validation data.
+ Defaults to True.
+ """
+ self._logger = logging.getLogger(__name__)
+ self._distributed = distributed
+ self._output_dir = output_dir
+
+ if use_fast_impl and (COCOeval_opt is COCOeval):
+ self._logger.info("Fast COCO eval is not built. Falling back to official COCO eval.")
+ use_fast_impl = False
+ self._use_fast_impl = use_fast_impl
+
+ # COCOeval requires the limit on the number of detections per image (maxDets) to be a list
+ # with at least 3 elements. The default maxDets in COCOeval is [1, 10, 100], in which the
+ # 3rd element (100) is used as the limit on the number of detections per image when
+ # evaluating AP. COCOEvaluator expects an integer for max_dets_per_image, so for COCOeval,
+ # we reformat max_dets_per_image into [1, 10, max_dets_per_image], based on the defaults.
+ if max_dets_per_image is None:
+ max_dets_per_image = [1, 10, 100]
+ else:
+ max_dets_per_image = [1, 10, max_dets_per_image]
+ self._max_dets_per_image = max_dets_per_image
+
+ if tasks is not None and isinstance(tasks, CfgNode):
+ kpt_oks_sigmas = (
+ tasks.TEST.KEYPOINT_OKS_SIGMAS if not kpt_oks_sigmas else kpt_oks_sigmas
+ )
+ self._logger.warn(
+ "COCO Evaluator instantiated using config, this is deprecated behavior."
+ " Please pass in explicit arguments instead."
+ )
+ self._tasks = None # Infering it from predictions should be better
+ else:
+ self._tasks = tasks
+
+ self._cpu_device = torch.device("cpu")
+
+ self._metadata = MetadataCatalog.get(dataset_name)
+ if not hasattr(self._metadata, "json_file"):
+ if output_dir is None:
+ raise ValueError(
+ "output_dir must be provided to COCOEvaluator "
+ "for datasets not in COCO format."
+ )
+ self._logger.info(f"Trying to convert '{dataset_name}' to COCO format ...")
+
+ cache_path = os.path.join(output_dir, f"{dataset_name}_coco_format.json")
+ self._metadata.json_file = cache_path
+ convert_to_coco_json(dataset_name, cache_path, allow_cached=allow_cached_coco)
+
+ json_file = PathManager.get_local_path(self._metadata.json_file)
+ with contextlib.redirect_stdout(io.StringIO()):
+ self._coco_api = COCO(json_file)
+
+ # Test set json files do not contain annotations (evaluation must be
+ # performed using the COCO evaluation server).
+ self._do_evaluation = "annotations" in self._coco_api.dataset
+ if self._do_evaluation:
+ self._kpt_oks_sigmas = kpt_oks_sigmas
+
+ def reset(self):
+ self._predictions = []
+
+ def process(self, inputs, outputs):
+ """
+ Args:
+ inputs: the inputs to a COCO model (e.g., GeneralizedRCNN).
+ It is a list of dict. Each dict corresponds to an image and
+ contains keys like "height", "width", "file_name", "image_id".
+ outputs: the outputs of a COCO model. It is a list of dicts with key
+ "instances" that contains :class:`Instances`.
+ """
+ for input, output in zip(inputs, outputs):
+ prediction = {"image_id": input["image_id"]}
+
+ if "instances" in output:
+ instances = output["instances"].to(self._cpu_device)
+ prediction["instances"] = instances_to_coco_json(instances, input["image_id"])
+ if "proposals" in output:
+ prediction["proposals"] = output["proposals"].to(self._cpu_device)
+ if len(prediction) > 1:
+ self._predictions.append(prediction)
+
+ def evaluate(self, img_ids=None):
+ """
+ Args:
+ img_ids: a list of image IDs to evaluate on. Default to None for the whole dataset
+ """
+ if self._distributed:
+ comm.synchronize()
+ predictions = comm.gather(self._predictions, dst=0)
+ predictions = list(itertools.chain(*predictions))
+
+ if not comm.is_main_process():
+ return {}
+ else:
+ predictions = self._predictions
+
+ if len(predictions) == 0:
+ self._logger.warning("[COCOEvaluator] Did not receive valid predictions.")
+ return {}
+
+ if self._output_dir:
+ PathManager.mkdirs(self._output_dir)
+ file_path = os.path.join(self._output_dir, "instances_predictions.pth")
+ with PathManager.open(file_path, "wb") as f:
+ torch.save(predictions, f)
+
+ self._results = OrderedDict()
+ if "proposals" in predictions[0]:
+ self._eval_box_proposals(predictions)
+ if "instances" in predictions[0]:
+ self._eval_predictions(predictions, img_ids=img_ids)
+ # Copy so the caller can do whatever with results
+ return copy.deepcopy(self._results)
+
+ def _tasks_from_predictions(self, predictions):
+ """
+ Get COCO API "tasks" (i.e. iou_type) from COCO-format predictions.
+ """
+ tasks = {"bbox"}
+ for pred in predictions:
+ if "segmentation" in pred:
+ tasks.add("segm")
+ if "keypoints" in pred:
+ tasks.add("keypoints")
+ return sorted(tasks)
+
+ def _eval_predictions(self, predictions, img_ids=None):
+ """
+ Evaluate predictions. Fill self._results with the metrics of the tasks.
+ """
+ self._logger.info("Preparing results for COCO format ...")
+ coco_results = list(itertools.chain(*[x["instances"] for x in predictions]))
+ tasks = self._tasks or self._tasks_from_predictions(coco_results)
+
+ # unmap the category ids for COCO
+ if hasattr(self._metadata, "thing_dataset_id_to_contiguous_id"):
+ dataset_id_to_contiguous_id = self._metadata.thing_dataset_id_to_contiguous_id
+ all_contiguous_ids = list(dataset_id_to_contiguous_id.values())
+ num_classes = len(all_contiguous_ids)
+ assert min(all_contiguous_ids) == 0 and max(all_contiguous_ids) == num_classes - 1
+
+ reverse_id_mapping = {v: k for k, v in dataset_id_to_contiguous_id.items()}
+ for result in coco_results:
+ category_id = result["category_id"]
+ assert category_id < num_classes, (
+ f"A prediction has class={category_id}, "
+ f"but the dataset only has {num_classes} classes and "
+ f"predicted class id should be in [0, {num_classes - 1}]."
+ )
+ result["category_id"] = reverse_id_mapping[category_id]
+
+ if self._output_dir:
+ file_path = os.path.join(self._output_dir, "coco_instances_results.json")
+ self._logger.info("Saving results to {}".format(file_path))
+ with PathManager.open(file_path, "w") as f:
+ f.write(json.dumps(coco_results))
+ f.flush()
+
+ if not self._do_evaluation:
+ self._logger.info("Annotations are not available for evaluation.")
+ return
+
+ self._logger.info(
+ "Evaluating predictions with {} COCO API...".format(
+ "unofficial" if self._use_fast_impl else "official"
+ )
+ )
+ for task in sorted(tasks):
+ assert task in {"bbox", "segm", "keypoints"}, f"Got unknown task: {task}!"
+ coco_eval = (
+ _evaluate_predictions_on_coco(
+ self._coco_api,
+ coco_results,
+ task,
+ kpt_oks_sigmas=self._kpt_oks_sigmas,
+ cocoeval_fn=COCOeval_opt if self._use_fast_impl else COCOeval,
+ img_ids=img_ids,
+ max_dets_per_image=self._max_dets_per_image,
+ )
+ if len(coco_results) > 0
+ else None # cocoapi does not handle empty results very well
+ )
+
+ res = self._derive_coco_results(
+ coco_eval, task, class_names=self._metadata.get("thing_classes")
+ )
+ self._results[task] = res
+
+ def _eval_box_proposals(self, predictions):
+ """
+ Evaluate the box proposals in predictions.
+ Fill self._results with the metrics for "box_proposals" task.
+ """
+ if self._output_dir:
+ # Saving generated box proposals to file.
+ # Predicted box_proposals are in XYXY_ABS mode.
+ bbox_mode = BoxMode.XYXY_ABS.value
+ ids, boxes, objectness_logits = [], [], []
+ for prediction in predictions:
+ ids.append(prediction["image_id"])
+ boxes.append(prediction["proposals"].proposal_boxes.tensor.numpy())
+ objectness_logits.append(prediction["proposals"].objectness_logits.numpy())
+
+ proposal_data = {
+ "boxes": boxes,
+ "objectness_logits": objectness_logits,
+ "ids": ids,
+ "bbox_mode": bbox_mode,
+ }
+ with PathManager.open(os.path.join(self._output_dir, "box_proposals.pkl"), "wb") as f:
+ pickle.dump(proposal_data, f)
+
+ if not self._do_evaluation:
+ self._logger.info("Annotations are not available for evaluation.")
+ return
+
+ self._logger.info("Evaluating bbox proposals ...")
+ res = {}
+ areas = {"all": "", "small": "s", "medium": "m", "large": "l"}
+ for limit in [100, 1000]:
+ for area, suffix in areas.items():
+ stats = _evaluate_box_proposals(predictions, self._coco_api, area=area, limit=limit)
+ key = "AR{}@{:d}".format(suffix, limit)
+ res[key] = float(stats["ar"].item() * 100)
+ self._logger.info("Proposal metrics: \n" + create_small_table(res))
+ self._results["box_proposals"] = res
+
+ def _derive_coco_results(self, coco_eval, iou_type, class_names=None):
+ """
+ Derive the desired score numbers from summarized COCOeval.
+
+ Args:
+ coco_eval (None or COCOEval): None represents no predictions from model.
+ iou_type (str):
+ class_names (None or list[str]): if provided, will use it to predict
+ per-category AP.
+
+ Returns:
+ a dict of {metric name: score}
+ """
+
+ metrics = {
+ "bbox": ["AP", "AP50", "AP75", "APs", "APm", "APl"],
+ "segm": ["AP", "AP50", "AP75", "APs", "APm", "APl"],
+ "keypoints": ["AP", "AP50", "AP75", "APm", "APl"],
+ }[iou_type]
+
+ if coco_eval is None:
+ self._logger.warn("No predictions from the model!")
+ return {metric: float("nan") for metric in metrics}
+
+ # the standard metrics
+ results = {
+ metric: float(coco_eval.stats[idx] * 100 if coco_eval.stats[idx] >= 0 else "nan")
+ for idx, metric in enumerate(metrics)
+ }
+ self._logger.info(
+ "Evaluation results for {}: \n".format(iou_type) + create_small_table(results)
+ )
+ if not np.isfinite(sum(results.values())):
+ self._logger.info("Some metrics cannot be computed and is shown as NaN.")
+
+ if class_names is None or len(class_names) <= 1:
+ return results
+ # Compute per-category AP
+ # from https://github.com/facebookresearch/Detectron/blob/a6a835f5b8208c45d0dce217ce9bbda915f44df7/detectron/datasets/json_dataset_evaluator.py#L222-L252 # noqa
+ precisions = coco_eval.eval["precision"]
+ # precision has dims (iou, recall, cls, area range, max dets)
+ assert len(class_names) == precisions.shape[2]
+
+ results_per_category = []
+ for idx, name in enumerate(class_names):
+ # area range index 0: all area ranges
+ # max dets index -1: typically 100 per image
+ precision = precisions[:, :, idx, 0, -1]
+ precision = precision[precision > -1]
+ ap = np.mean(precision) if precision.size else float("nan")
+ results_per_category.append(("{}".format(name), float(ap * 100)))
+
+ # tabulate it
+ N_COLS = min(6, len(results_per_category) * 2)
+ results_flatten = list(itertools.chain(*results_per_category))
+ results_2d = itertools.zip_longest(*[results_flatten[i::N_COLS] for i in range(N_COLS)])
+ table = tabulate(
+ results_2d,
+ tablefmt="pipe",
+ floatfmt=".3f",
+ headers=["category", "AP"] * (N_COLS // 2),
+ numalign="left",
+ )
+ self._logger.info("Per-category {} AP: \n".format(iou_type) + table)
+
+ results.update({"AP-" + name: ap for name, ap in results_per_category})
+ return results
+
+
+def instances_to_coco_json(instances, img_id):
+ """
+ Dump an "Instances" object to a COCO-format json that's used for evaluation.
+
+ Args:
+ instances (Instances):
+ img_id (int): the image id
+
+ Returns:
+ list[dict]: list of json annotations in COCO format.
+ """
+ num_instance = len(instances)
+ if num_instance == 0:
+ return []
+
+ boxes = instances.pred_boxes.tensor.numpy()
+ boxes = BoxMode.convert(boxes, BoxMode.XYXY_ABS, BoxMode.XYWH_ABS)
+ boxes = boxes.tolist()
+ scores = instances.scores.tolist()
+ classes = instances.pred_classes.tolist()
+
+ has_mask = instances.has("pred_masks")
+ if has_mask:
+ # use RLE to encode the masks, because they are too large and takes memory
+ # since this evaluator stores outputs of the entire dataset
+ rles = [
+ mask_util.encode(np.array(mask[:, :, None], order="F", dtype="uint8"))[0]
+ for mask in instances.pred_masks
+ ]
+ for rle in rles:
+ # "counts" is an array encoded by mask_util as a byte-stream. Python3's
+ # json writer which always produces strings cannot serialize a bytestream
+ # unless you decode it. Thankfully, utf-8 works out (which is also what
+ # the pycocotools/_mask.pyx does).
+ rle["counts"] = rle["counts"].decode("utf-8")
+
+ has_keypoints = instances.has("pred_keypoints")
+ if has_keypoints:
+ keypoints = instances.pred_keypoints
+
+ results = []
+ for k in range(num_instance):
+ result = {
+ "image_id": img_id,
+ "category_id": classes[k],
+ "bbox": boxes[k],
+ "score": scores[k],
+ }
+ if has_mask:
+ result["segmentation"] = rles[k]
+ if has_keypoints:
+ # In COCO annotations,
+ # keypoints coordinates are pixel indices.
+ # However our predictions are floating point coordinates.
+ # Therefore we subtract 0.5 to be consistent with the annotation format.
+ # This is the inverse of data loading logic in `datasets/coco.py`.
+ keypoints[k][:, :2] -= 0.5
+ result["keypoints"] = keypoints[k].flatten().tolist()
+ results.append(result)
+ return results
+
+
+# inspired from Detectron:
+# https://github.com/facebookresearch/Detectron/blob/a6a835f5b8208c45d0dce217ce9bbda915f44df7/detectron/datasets/json_dataset_evaluator.py#L255 # noqa
+def _evaluate_box_proposals(dataset_predictions, coco_api, thresholds=None, area="all", limit=None):
+ """
+ Evaluate detection proposal recall metrics. This function is a much
+ faster alternative to the official COCO API recall evaluation code. However,
+ it produces slightly different results.
+ """
+ # Record max overlap value for each gt box
+ # Return vector of overlap values
+ areas = {
+ "all": 0,
+ "small": 1,
+ "medium": 2,
+ "large": 3,
+ "96-128": 4,
+ "128-256": 5,
+ "256-512": 6,
+ "512-inf": 7,
+ }
+ area_ranges = [
+ [0**2, 1e5**2], # all
+ [0**2, 32**2], # small
+ [32**2, 96**2], # medium
+ [96**2, 1e5**2], # large
+ [96**2, 128**2], # 96-128
+ [128**2, 256**2], # 128-256
+ [256**2, 512**2], # 256-512
+ [512**2, 1e5**2],
+ ] # 512-inf
+ assert area in areas, "Unknown area range: {}".format(area)
+ area_range = area_ranges[areas[area]]
+ gt_overlaps = []
+ num_pos = 0
+
+ for prediction_dict in dataset_predictions:
+ predictions = prediction_dict["proposals"]
+
+ # sort predictions in descending order
+ # TODO maybe remove this and make it explicit in the documentation
+ inds = predictions.objectness_logits.sort(descending=True)[1]
+ predictions = predictions[inds]
+
+ ann_ids = coco_api.getAnnIds(imgIds=prediction_dict["image_id"])
+ anno = coco_api.loadAnns(ann_ids)
+ gt_boxes = [
+ BoxMode.convert(obj["bbox"], BoxMode.XYWH_ABS, BoxMode.XYXY_ABS)
+ for obj in anno
+ if obj["iscrowd"] == 0
+ ]
+ gt_boxes = torch.as_tensor(gt_boxes).reshape(-1, 4) # guard against no boxes
+ gt_boxes = Boxes(gt_boxes)
+ gt_areas = torch.as_tensor([obj["area"] for obj in anno if obj["iscrowd"] == 0])
+
+ if len(gt_boxes) == 0 or len(predictions) == 0:
+ continue
+
+ valid_gt_inds = (gt_areas >= area_range[0]) & (gt_areas <= area_range[1])
+ gt_boxes = gt_boxes[valid_gt_inds]
+
+ num_pos += len(gt_boxes)
+
+ if len(gt_boxes) == 0:
+ continue
+
+ if limit is not None and len(predictions) > limit:
+ predictions = predictions[:limit]
+
+ overlaps = pairwise_iou(predictions.proposal_boxes, gt_boxes)
+
+ _gt_overlaps = torch.zeros(len(gt_boxes))
+ for j in range(min(len(predictions), len(gt_boxes))):
+ # find which proposal box maximally covers each gt box
+ # and get the iou amount of coverage for each gt box
+ max_overlaps, argmax_overlaps = overlaps.max(dim=0)
+
+ # find which gt box is 'best' covered (i.e. 'best' = most iou)
+ gt_ovr, gt_ind = max_overlaps.max(dim=0)
+ assert gt_ovr >= 0
+ # find the proposal box that covers the best covered gt box
+ box_ind = argmax_overlaps[gt_ind]
+ # record the iou coverage of this gt box
+ _gt_overlaps[j] = overlaps[box_ind, gt_ind]
+ assert _gt_overlaps[j] == gt_ovr
+ # mark the proposal box and the gt box as used
+ overlaps[box_ind, :] = -1
+ overlaps[:, gt_ind] = -1
+
+ # append recorded iou coverage level
+ gt_overlaps.append(_gt_overlaps)
+ gt_overlaps = (
+ torch.cat(gt_overlaps, dim=0) if len(gt_overlaps) else torch.zeros(0, dtype=torch.float32)
+ )
+ gt_overlaps, _ = torch.sort(gt_overlaps)
+
+ if thresholds is None:
+ step = 0.05
+ thresholds = torch.arange(0.5, 0.95 + 1e-5, step, dtype=torch.float32)
+ recalls = torch.zeros_like(thresholds)
+ # compute recall for each iou threshold
+ for i, t in enumerate(thresholds):
+ recalls[i] = (gt_overlaps >= t).float().sum() / float(num_pos)
+ # ar = 2 * np.trapz(recalls, thresholds)
+ ar = recalls.mean()
+ return {
+ "ar": ar,
+ "recalls": recalls,
+ "thresholds": thresholds,
+ "gt_overlaps": gt_overlaps,
+ "num_pos": num_pos,
+ }
+
+
+def _evaluate_predictions_on_coco(
+ coco_gt,
+ coco_results,
+ iou_type,
+ kpt_oks_sigmas=None,
+ cocoeval_fn=COCOeval_opt,
+ img_ids=None,
+ max_dets_per_image=None,
+):
+ """
+ Evaluate the coco results using COCOEval API.
+ """
+ assert len(coco_results) > 0
+
+ if iou_type == "segm":
+ coco_results = copy.deepcopy(coco_results)
+ # When evaluating mask AP, if the results contain bbox, cocoapi will
+ # use the box area as the area of the instance, instead of the mask area.
+ # This leads to a different definition of small/medium/large.
+ # We remove the bbox field to let mask AP use mask area.
+ for c in coco_results:
+ c.pop("bbox", None)
+
+ coco_dt = coco_gt.loadRes(coco_results)
+ coco_eval = cocoeval_fn(coco_gt, coco_dt, iou_type)
+ # For COCO, the default max_dets_per_image is [1, 10, 100].
+ if max_dets_per_image is None:
+ max_dets_per_image = [1, 10, 100] # Default from COCOEval
+ else:
+ assert (
+ len(max_dets_per_image) >= 3
+ ), "COCOeval requires maxDets (and max_dets_per_image) to have length at least 3"
+ # In the case that user supplies a custom input for max_dets_per_image,
+ # apply COCOevalMaxDets to evaluate AP with the custom input.
+ if max_dets_per_image[2] != 100:
+ coco_eval = COCOevalMaxDets(coco_gt, coco_dt, iou_type)
+ if iou_type != "keypoints":
+ coco_eval.params.maxDets = max_dets_per_image
+
+ if img_ids is not None:
+ coco_eval.params.imgIds = img_ids
+
+ if iou_type == "keypoints":
+ # Use the COCO default keypoint OKS sigmas unless overrides are specified
+ if kpt_oks_sigmas:
+ assert hasattr(coco_eval.params, "kpt_oks_sigmas"), "pycocotools is too old!"
+ coco_eval.params.kpt_oks_sigmas = np.array(kpt_oks_sigmas)
+ # COCOAPI requires every detection and every gt to have keypoints, so
+ # we just take the first entry from both
+ num_keypoints_dt = len(coco_results[0]["keypoints"]) // 3
+ num_keypoints_gt = len(next(iter(coco_gt.anns.values()))["keypoints"]) // 3
+ num_keypoints_oks = len(coco_eval.params.kpt_oks_sigmas)
+ assert num_keypoints_oks == num_keypoints_dt == num_keypoints_gt, (
+ f"[COCOEvaluator] Prediction contain {num_keypoints_dt} keypoints. "
+ f"Ground truth contains {num_keypoints_gt} keypoints. "
+ f"The length of cfg.TEST.KEYPOINT_OKS_SIGMAS is {num_keypoints_oks}. "
+ "They have to agree with each other. For meaning of OKS, please refer to "
+ "http://cocodataset.org/#keypoints-eval."
+ )
+
+ coco_eval.evaluate()
+ coco_eval.accumulate()
+ coco_eval.summarize()
+
+ return coco_eval
+
+
+class COCOevalMaxDets(COCOeval):
+ """
+ Modified version of COCOeval for evaluating AP with a custom
+ maxDets (by default for COCO, maxDets is 100)
+ """
+
+ def summarize(self):
+ """
+ Compute and display summary metrics for evaluation results given
+ a custom value for max_dets_per_image
+ """
+
+ def _summarize(ap=1, iouThr=None, areaRng="all", maxDets=100):
+ p = self.params
+ iStr = " {:<18} {} @[ IoU={:<9} | area={:>6s} | maxDets={:>3d} ] = {:0.3f}"
+ titleStr = "Average Precision" if ap == 1 else "Average Recall"
+ typeStr = "(AP)" if ap == 1 else "(AR)"
+ iouStr = (
+ "{:0.2f}:{:0.2f}".format(p.iouThrs[0], p.iouThrs[-1])
+ if iouThr is None
+ else "{:0.2f}".format(iouThr)
+ )
+
+ aind = [i for i, aRng in enumerate(p.areaRngLbl) if aRng == areaRng]
+ mind = [i for i, mDet in enumerate(p.maxDets) if mDet == maxDets]
+ if ap == 1:
+ # dimension of precision: [TxRxKxAxM]
+ s = self.eval["precision"]
+ # IoU
+ if iouThr is not None:
+ t = np.where(iouThr == p.iouThrs)[0]
+ s = s[t]
+ s = s[:, :, :, aind, mind]
+ else:
+ # dimension of recall: [TxKxAxM]
+ s = self.eval["recall"]
+ if iouThr is not None:
+ t = np.where(iouThr == p.iouThrs)[0]
+ s = s[t]
+ s = s[:, :, aind, mind]
+ if len(s[s > -1]) == 0:
+ mean_s = -1
+ else:
+ mean_s = np.mean(s[s > -1])
+ print(iStr.format(titleStr, typeStr, iouStr, areaRng, maxDets, mean_s))
+ return mean_s
+
+ def _summarizeDets():
+ stats = np.zeros((12,))
+ # Evaluate AP using the custom limit on maximum detections per image
+ stats[0] = _summarize(1, maxDets=self.params.maxDets[2])
+ stats[1] = _summarize(1, iouThr=0.5, maxDets=self.params.maxDets[2])
+ stats[2] = _summarize(1, iouThr=0.75, maxDets=self.params.maxDets[2])
+ stats[3] = _summarize(1, areaRng="small", maxDets=self.params.maxDets[2])
+ stats[4] = _summarize(1, areaRng="medium", maxDets=self.params.maxDets[2])
+ stats[5] = _summarize(1, areaRng="large", maxDets=self.params.maxDets[2])
+ stats[6] = _summarize(0, maxDets=self.params.maxDets[0])
+ stats[7] = _summarize(0, maxDets=self.params.maxDets[1])
+ stats[8] = _summarize(0, maxDets=self.params.maxDets[2])
+ stats[9] = _summarize(0, areaRng="small", maxDets=self.params.maxDets[2])
+ stats[10] = _summarize(0, areaRng="medium", maxDets=self.params.maxDets[2])
+ stats[11] = _summarize(0, areaRng="large", maxDets=self.params.maxDets[2])
+ return stats
+
+ def _summarizeKps():
+ stats = np.zeros((10,))
+ stats[0] = _summarize(1, maxDets=20)
+ stats[1] = _summarize(1, maxDets=20, iouThr=0.5)
+ stats[2] = _summarize(1, maxDets=20, iouThr=0.75)
+ stats[3] = _summarize(1, maxDets=20, areaRng="medium")
+ stats[4] = _summarize(1, maxDets=20, areaRng="large")
+ stats[5] = _summarize(0, maxDets=20)
+ stats[6] = _summarize(0, maxDets=20, iouThr=0.5)
+ stats[7] = _summarize(0, maxDets=20, iouThr=0.75)
+ stats[8] = _summarize(0, maxDets=20, areaRng="medium")
+ stats[9] = _summarize(0, maxDets=20, areaRng="large")
+ return stats
+
+ if not self.eval:
+ raise Exception("Please run accumulate() first")
+ iouType = self.params.iouType
+ if iouType == "segm" or iouType == "bbox":
+ summarize = _summarizeDets
+ elif iouType == "keypoints":
+ summarize = _summarizeKps
+ self.stats = summarize()
+
+ def __str__(self):
+ self.summarize()
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/evaluator.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/evaluator.py
new file mode 100644
index 0000000000000000000000000000000000000000..9cddc296432cbb6f11caf3c3be98833a50778ffb
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/evaluator.py
@@ -0,0 +1,224 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import datetime
+import logging
+import time
+from collections import OrderedDict, abc
+from contextlib import ExitStack, contextmanager
+from typing import List, Union
+import torch
+from torch import nn
+
+from annotator.oneformer.detectron2.utils.comm import get_world_size, is_main_process
+from annotator.oneformer.detectron2.utils.logger import log_every_n_seconds
+
+
+class DatasetEvaluator:
+ """
+ Base class for a dataset evaluator.
+
+ The function :func:`inference_on_dataset` runs the model over
+ all samples in the dataset, and have a DatasetEvaluator to process the inputs/outputs.
+
+ This class will accumulate information of the inputs/outputs (by :meth:`process`),
+ and produce evaluation results in the end (by :meth:`evaluate`).
+ """
+
+ def reset(self):
+ """
+ Preparation for a new round of evaluation.
+ Should be called before starting a round of evaluation.
+ """
+ pass
+
+ def process(self, inputs, outputs):
+ """
+ Process the pair of inputs and outputs.
+ If they contain batches, the pairs can be consumed one-by-one using `zip`:
+
+ .. code-block:: python
+
+ for input_, output in zip(inputs, outputs):
+ # do evaluation on single input/output pair
+ ...
+
+ Args:
+ inputs (list): the inputs that's used to call the model.
+ outputs (list): the return value of `model(inputs)`
+ """
+ pass
+
+ def evaluate(self):
+ """
+ Evaluate/summarize the performance, after processing all input/output pairs.
+
+ Returns:
+ dict:
+ A new evaluator class can return a dict of arbitrary format
+ as long as the user can process the results.
+ In our train_net.py, we expect the following format:
+
+ * key: the name of the task (e.g., bbox)
+ * value: a dict of {metric name: score}, e.g.: {"AP50": 80}
+ """
+ pass
+
+
+class DatasetEvaluators(DatasetEvaluator):
+ """
+ Wrapper class to combine multiple :class:`DatasetEvaluator` instances.
+
+ This class dispatches every evaluation call to
+ all of its :class:`DatasetEvaluator`.
+ """
+
+ def __init__(self, evaluators):
+ """
+ Args:
+ evaluators (list): the evaluators to combine.
+ """
+ super().__init__()
+ self._evaluators = evaluators
+
+ def reset(self):
+ for evaluator in self._evaluators:
+ evaluator.reset()
+
+ def process(self, inputs, outputs):
+ for evaluator in self._evaluators:
+ evaluator.process(inputs, outputs)
+
+ def evaluate(self):
+ results = OrderedDict()
+ for evaluator in self._evaluators:
+ result = evaluator.evaluate()
+ if is_main_process() and result is not None:
+ for k, v in result.items():
+ assert (
+ k not in results
+ ), "Different evaluators produce results with the same key {}".format(k)
+ results[k] = v
+ return results
+
+
+def inference_on_dataset(
+ model, data_loader, evaluator: Union[DatasetEvaluator, List[DatasetEvaluator], None]
+):
+ """
+ Run model on the data_loader and evaluate the metrics with evaluator.
+ Also benchmark the inference speed of `model.__call__` accurately.
+ The model will be used in eval mode.
+
+ Args:
+ model (callable): a callable which takes an object from
+ `data_loader` and returns some outputs.
+
+ If it's an nn.Module, it will be temporarily set to `eval` mode.
+ If you wish to evaluate a model in `training` mode instead, you can
+ wrap the given model and override its behavior of `.eval()` and `.train()`.
+ data_loader: an iterable object with a length.
+ The elements it generates will be the inputs to the model.
+ evaluator: the evaluator(s) to run. Use `None` if you only want to benchmark,
+ but don't want to do any evaluation.
+
+ Returns:
+ The return value of `evaluator.evaluate()`
+ """
+ num_devices = get_world_size()
+ logger = logging.getLogger(__name__)
+ logger.info("Start inference on {} batches".format(len(data_loader)))
+
+ total = len(data_loader) # inference data loader must have a fixed length
+ if evaluator is None:
+ # create a no-op evaluator
+ evaluator = DatasetEvaluators([])
+ if isinstance(evaluator, abc.MutableSequence):
+ evaluator = DatasetEvaluators(evaluator)
+ evaluator.reset()
+
+ num_warmup = min(5, total - 1)
+ start_time = time.perf_counter()
+ total_data_time = 0
+ total_compute_time = 0
+ total_eval_time = 0
+ with ExitStack() as stack:
+ if isinstance(model, nn.Module):
+ stack.enter_context(inference_context(model))
+ stack.enter_context(torch.no_grad())
+
+ start_data_time = time.perf_counter()
+ for idx, inputs in enumerate(data_loader):
+ total_data_time += time.perf_counter() - start_data_time
+ if idx == num_warmup:
+ start_time = time.perf_counter()
+ total_data_time = 0
+ total_compute_time = 0
+ total_eval_time = 0
+
+ start_compute_time = time.perf_counter()
+ outputs = model(inputs)
+ if torch.cuda.is_available():
+ torch.cuda.synchronize()
+ total_compute_time += time.perf_counter() - start_compute_time
+
+ start_eval_time = time.perf_counter()
+ evaluator.process(inputs, outputs)
+ total_eval_time += time.perf_counter() - start_eval_time
+
+ iters_after_start = idx + 1 - num_warmup * int(idx >= num_warmup)
+ data_seconds_per_iter = total_data_time / iters_after_start
+ compute_seconds_per_iter = total_compute_time / iters_after_start
+ eval_seconds_per_iter = total_eval_time / iters_after_start
+ total_seconds_per_iter = (time.perf_counter() - start_time) / iters_after_start
+ if idx >= num_warmup * 2 or compute_seconds_per_iter > 5:
+ eta = datetime.timedelta(seconds=int(total_seconds_per_iter * (total - idx - 1)))
+ log_every_n_seconds(
+ logging.INFO,
+ (
+ f"Inference done {idx + 1}/{total}. "
+ f"Dataloading: {data_seconds_per_iter:.4f} s/iter. "
+ f"Inference: {compute_seconds_per_iter:.4f} s/iter. "
+ f"Eval: {eval_seconds_per_iter:.4f} s/iter. "
+ f"Total: {total_seconds_per_iter:.4f} s/iter. "
+ f"ETA={eta}"
+ ),
+ n=5,
+ )
+ start_data_time = time.perf_counter()
+
+ # Measure the time only for this worker (before the synchronization barrier)
+ total_time = time.perf_counter() - start_time
+ total_time_str = str(datetime.timedelta(seconds=total_time))
+ # NOTE this format is parsed by grep
+ logger.info(
+ "Total inference time: {} ({:.6f} s / iter per device, on {} devices)".format(
+ total_time_str, total_time / (total - num_warmup), num_devices
+ )
+ )
+ total_compute_time_str = str(datetime.timedelta(seconds=int(total_compute_time)))
+ logger.info(
+ "Total inference pure compute time: {} ({:.6f} s / iter per device, on {} devices)".format(
+ total_compute_time_str, total_compute_time / (total - num_warmup), num_devices
+ )
+ )
+
+ results = evaluator.evaluate()
+ # An evaluator may return None when not in main process.
+ # Replace it by an empty dict instead to make it easier for downstream code to handle
+ if results is None:
+ results = {}
+ return results
+
+
+@contextmanager
+def inference_context(model):
+ """
+ A context where the model is temporarily changed to eval mode,
+ and restored to previous mode afterwards.
+
+ Args:
+ model: a torch Module
+ """
+ training_mode = model.training
+ model.eval()
+ yield
+ model.train(training_mode)
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/fast_eval_api.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/fast_eval_api.py
new file mode 100644
index 0000000000000000000000000000000000000000..75458b1cf8c26500da9b6e60cb6224a3c26d6dd2
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/fast_eval_api.py
@@ -0,0 +1,121 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import copy
+import logging
+import numpy as np
+import time
+from pycocotools.cocoeval import COCOeval
+
+from annotator.oneformer.detectron2 import _C
+
+logger = logging.getLogger(__name__)
+
+
+class COCOeval_opt(COCOeval):
+ """
+ This is a slightly modified version of the original COCO API, where the functions evaluateImg()
+ and accumulate() are implemented in C++ to speedup evaluation
+ """
+
+ def evaluate(self):
+ """
+ Run per image evaluation on given images and store results in self.evalImgs_cpp, a
+ datastructure that isn't readable from Python but is used by a c++ implementation of
+ accumulate(). Unlike the original COCO PythonAPI, we don't populate the datastructure
+ self.evalImgs because this datastructure is a computational bottleneck.
+ :return: None
+ """
+ tic = time.time()
+
+ p = self.params
+ # add backward compatibility if useSegm is specified in params
+ if p.useSegm is not None:
+ p.iouType = "segm" if p.useSegm == 1 else "bbox"
+ logger.info("Evaluate annotation type *{}*".format(p.iouType))
+ p.imgIds = list(np.unique(p.imgIds))
+ if p.useCats:
+ p.catIds = list(np.unique(p.catIds))
+ p.maxDets = sorted(p.maxDets)
+ self.params = p
+
+ self._prepare() # bottleneck
+
+ # loop through images, area range, max detection number
+ catIds = p.catIds if p.useCats else [-1]
+
+ if p.iouType == "segm" or p.iouType == "bbox":
+ computeIoU = self.computeIoU
+ elif p.iouType == "keypoints":
+ computeIoU = self.computeOks
+ self.ious = {
+ (imgId, catId): computeIoU(imgId, catId) for imgId in p.imgIds for catId in catIds
+ } # bottleneck
+
+ maxDet = p.maxDets[-1]
+
+ # <<<< Beginning of code differences with original COCO API
+ def convert_instances_to_cpp(instances, is_det=False):
+ # Convert annotations for a list of instances in an image to a format that's fast
+ # to access in C++
+ instances_cpp = []
+ for instance in instances:
+ instance_cpp = _C.InstanceAnnotation(
+ int(instance["id"]),
+ instance["score"] if is_det else instance.get("score", 0.0),
+ instance["area"],
+ bool(instance.get("iscrowd", 0)),
+ bool(instance.get("ignore", 0)),
+ )
+ instances_cpp.append(instance_cpp)
+ return instances_cpp
+
+ # Convert GT annotations, detections, and IOUs to a format that's fast to access in C++
+ ground_truth_instances = [
+ [convert_instances_to_cpp(self._gts[imgId, catId]) for catId in p.catIds]
+ for imgId in p.imgIds
+ ]
+ detected_instances = [
+ [convert_instances_to_cpp(self._dts[imgId, catId], is_det=True) for catId in p.catIds]
+ for imgId in p.imgIds
+ ]
+ ious = [[self.ious[imgId, catId] for catId in catIds] for imgId in p.imgIds]
+
+ if not p.useCats:
+ # For each image, flatten per-category lists into a single list
+ ground_truth_instances = [[[o for c in i for o in c]] for i in ground_truth_instances]
+ detected_instances = [[[o for c in i for o in c]] for i in detected_instances]
+
+ # Call C++ implementation of self.evaluateImgs()
+ self._evalImgs_cpp = _C.COCOevalEvaluateImages(
+ p.areaRng, maxDet, p.iouThrs, ious, ground_truth_instances, detected_instances
+ )
+ self._evalImgs = None
+
+ self._paramsEval = copy.deepcopy(self.params)
+ toc = time.time()
+ logger.info("COCOeval_opt.evaluate() finished in {:0.2f} seconds.".format(toc - tic))
+ # >>>> End of code differences with original COCO API
+
+ def accumulate(self):
+ """
+ Accumulate per image evaluation results and store the result in self.eval. Does not
+ support changing parameter settings from those used by self.evaluate()
+ """
+ logger.info("Accumulating evaluation results...")
+ tic = time.time()
+ assert hasattr(
+ self, "_evalImgs_cpp"
+ ), "evaluate() must be called before accmulate() is called."
+
+ self.eval = _C.COCOevalAccumulate(self._paramsEval, self._evalImgs_cpp)
+
+ # recall is num_iou_thresholds X num_categories X num_area_ranges X num_max_detections
+ self.eval["recall"] = np.array(self.eval["recall"]).reshape(
+ self.eval["counts"][:1] + self.eval["counts"][2:]
+ )
+
+ # precision and scores are num_iou_thresholds X num_recall_thresholds X num_categories X
+ # num_area_ranges X num_max_detections
+ self.eval["precision"] = np.array(self.eval["precision"]).reshape(self.eval["counts"])
+ self.eval["scores"] = np.array(self.eval["scores"]).reshape(self.eval["counts"])
+ toc = time.time()
+ logger.info("COCOeval_opt.accumulate() finished in {:0.2f} seconds.".format(toc - tic))
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/lvis_evaluation.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/lvis_evaluation.py
new file mode 100644
index 0000000000000000000000000000000000000000..7d712ef262789edb85392cb54577c3a6b15e223e
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/lvis_evaluation.py
@@ -0,0 +1,380 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import copy
+import itertools
+import json
+import logging
+import os
+import pickle
+from collections import OrderedDict
+import torch
+
+import annotator.oneformer.detectron2.utils.comm as comm
+from annotator.oneformer.detectron2.config import CfgNode
+from annotator.oneformer.detectron2.data import MetadataCatalog
+from annotator.oneformer.detectron2.structures import Boxes, BoxMode, pairwise_iou
+from annotator.oneformer.detectron2.utils.file_io import PathManager
+from annotator.oneformer.detectron2.utils.logger import create_small_table
+
+from .coco_evaluation import instances_to_coco_json
+from .evaluator import DatasetEvaluator
+
+
+class LVISEvaluator(DatasetEvaluator):
+ """
+ Evaluate object proposal and instance detection/segmentation outputs using
+ LVIS's metrics and evaluation API.
+ """
+
+ def __init__(
+ self,
+ dataset_name,
+ tasks=None,
+ distributed=True,
+ output_dir=None,
+ *,
+ max_dets_per_image=None,
+ ):
+ """
+ Args:
+ dataset_name (str): name of the dataset to be evaluated.
+ It must have the following corresponding metadata:
+ "json_file": the path to the LVIS format annotation
+ tasks (tuple[str]): tasks that can be evaluated under the given
+ configuration. A task is one of "bbox", "segm".
+ By default, will infer this automatically from predictions.
+ distributed (True): if True, will collect results from all ranks for evaluation.
+ Otherwise, will evaluate the results in the current process.
+ output_dir (str): optional, an output directory to dump results.
+ max_dets_per_image (None or int): limit on maximum detections per image in evaluating AP
+ This limit, by default of the LVIS dataset, is 300.
+ """
+ from lvis import LVIS
+
+ self._logger = logging.getLogger(__name__)
+
+ if tasks is not None and isinstance(tasks, CfgNode):
+ self._logger.warn(
+ "COCO Evaluator instantiated using config, this is deprecated behavior."
+ " Please pass in explicit arguments instead."
+ )
+ self._tasks = None # Infering it from predictions should be better
+ else:
+ self._tasks = tasks
+
+ self._distributed = distributed
+ self._output_dir = output_dir
+ self._max_dets_per_image = max_dets_per_image
+
+ self._cpu_device = torch.device("cpu")
+
+ self._metadata = MetadataCatalog.get(dataset_name)
+ json_file = PathManager.get_local_path(self._metadata.json_file)
+ self._lvis_api = LVIS(json_file)
+ # Test set json files do not contain annotations (evaluation must be
+ # performed using the LVIS evaluation server).
+ self._do_evaluation = len(self._lvis_api.get_ann_ids()) > 0
+
+ def reset(self):
+ self._predictions = []
+
+ def process(self, inputs, outputs):
+ """
+ Args:
+ inputs: the inputs to a LVIS model (e.g., GeneralizedRCNN).
+ It is a list of dict. Each dict corresponds to an image and
+ contains keys like "height", "width", "file_name", "image_id".
+ outputs: the outputs of a LVIS model. It is a list of dicts with key
+ "instances" that contains :class:`Instances`.
+ """
+ for input, output in zip(inputs, outputs):
+ prediction = {"image_id": input["image_id"]}
+
+ if "instances" in output:
+ instances = output["instances"].to(self._cpu_device)
+ prediction["instances"] = instances_to_coco_json(instances, input["image_id"])
+ if "proposals" in output:
+ prediction["proposals"] = output["proposals"].to(self._cpu_device)
+ self._predictions.append(prediction)
+
+ def evaluate(self):
+ if self._distributed:
+ comm.synchronize()
+ predictions = comm.gather(self._predictions, dst=0)
+ predictions = list(itertools.chain(*predictions))
+
+ if not comm.is_main_process():
+ return
+ else:
+ predictions = self._predictions
+
+ if len(predictions) == 0:
+ self._logger.warning("[LVISEvaluator] Did not receive valid predictions.")
+ return {}
+
+ if self._output_dir:
+ PathManager.mkdirs(self._output_dir)
+ file_path = os.path.join(self._output_dir, "instances_predictions.pth")
+ with PathManager.open(file_path, "wb") as f:
+ torch.save(predictions, f)
+
+ self._results = OrderedDict()
+ if "proposals" in predictions[0]:
+ self._eval_box_proposals(predictions)
+ if "instances" in predictions[0]:
+ self._eval_predictions(predictions)
+ # Copy so the caller can do whatever with results
+ return copy.deepcopy(self._results)
+
+ def _tasks_from_predictions(self, predictions):
+ for pred in predictions:
+ if "segmentation" in pred:
+ return ("bbox", "segm")
+ return ("bbox",)
+
+ def _eval_predictions(self, predictions):
+ """
+ Evaluate predictions. Fill self._results with the metrics of the tasks.
+
+ Args:
+ predictions (list[dict]): list of outputs from the model
+ """
+ self._logger.info("Preparing results in the LVIS format ...")
+ lvis_results = list(itertools.chain(*[x["instances"] for x in predictions]))
+ tasks = self._tasks or self._tasks_from_predictions(lvis_results)
+
+ # LVIS evaluator can be used to evaluate results for COCO dataset categories.
+ # In this case `_metadata` variable will have a field with COCO-specific category mapping.
+ if hasattr(self._metadata, "thing_dataset_id_to_contiguous_id"):
+ reverse_id_mapping = {
+ v: k for k, v in self._metadata.thing_dataset_id_to_contiguous_id.items()
+ }
+ for result in lvis_results:
+ result["category_id"] = reverse_id_mapping[result["category_id"]]
+ else:
+ # unmap the category ids for LVIS (from 0-indexed to 1-indexed)
+ for result in lvis_results:
+ result["category_id"] += 1
+
+ if self._output_dir:
+ file_path = os.path.join(self._output_dir, "lvis_instances_results.json")
+ self._logger.info("Saving results to {}".format(file_path))
+ with PathManager.open(file_path, "w") as f:
+ f.write(json.dumps(lvis_results))
+ f.flush()
+
+ if not self._do_evaluation:
+ self._logger.info("Annotations are not available for evaluation.")
+ return
+
+ self._logger.info("Evaluating predictions ...")
+ for task in sorted(tasks):
+ res = _evaluate_predictions_on_lvis(
+ self._lvis_api,
+ lvis_results,
+ task,
+ max_dets_per_image=self._max_dets_per_image,
+ class_names=self._metadata.get("thing_classes"),
+ )
+ self._results[task] = res
+
+ def _eval_box_proposals(self, predictions):
+ """
+ Evaluate the box proposals in predictions.
+ Fill self._results with the metrics for "box_proposals" task.
+ """
+ if self._output_dir:
+ # Saving generated box proposals to file.
+ # Predicted box_proposals are in XYXY_ABS mode.
+ bbox_mode = BoxMode.XYXY_ABS.value
+ ids, boxes, objectness_logits = [], [], []
+ for prediction in predictions:
+ ids.append(prediction["image_id"])
+ boxes.append(prediction["proposals"].proposal_boxes.tensor.numpy())
+ objectness_logits.append(prediction["proposals"].objectness_logits.numpy())
+
+ proposal_data = {
+ "boxes": boxes,
+ "objectness_logits": objectness_logits,
+ "ids": ids,
+ "bbox_mode": bbox_mode,
+ }
+ with PathManager.open(os.path.join(self._output_dir, "box_proposals.pkl"), "wb") as f:
+ pickle.dump(proposal_data, f)
+
+ if not self._do_evaluation:
+ self._logger.info("Annotations are not available for evaluation.")
+ return
+
+ self._logger.info("Evaluating bbox proposals ...")
+ res = {}
+ areas = {"all": "", "small": "s", "medium": "m", "large": "l"}
+ for limit in [100, 1000]:
+ for area, suffix in areas.items():
+ stats = _evaluate_box_proposals(predictions, self._lvis_api, area=area, limit=limit)
+ key = "AR{}@{:d}".format(suffix, limit)
+ res[key] = float(stats["ar"].item() * 100)
+ self._logger.info("Proposal metrics: \n" + create_small_table(res))
+ self._results["box_proposals"] = res
+
+
+# inspired from Detectron:
+# https://github.com/facebookresearch/Detectron/blob/a6a835f5b8208c45d0dce217ce9bbda915f44df7/detectron/datasets/json_dataset_evaluator.py#L255 # noqa
+def _evaluate_box_proposals(dataset_predictions, lvis_api, thresholds=None, area="all", limit=None):
+ """
+ Evaluate detection proposal recall metrics. This function is a much
+ faster alternative to the official LVIS API recall evaluation code. However,
+ it produces slightly different results.
+ """
+ # Record max overlap value for each gt box
+ # Return vector of overlap values
+ areas = {
+ "all": 0,
+ "small": 1,
+ "medium": 2,
+ "large": 3,
+ "96-128": 4,
+ "128-256": 5,
+ "256-512": 6,
+ "512-inf": 7,
+ }
+ area_ranges = [
+ [0**2, 1e5**2], # all
+ [0**2, 32**2], # small
+ [32**2, 96**2], # medium
+ [96**2, 1e5**2], # large
+ [96**2, 128**2], # 96-128
+ [128**2, 256**2], # 128-256
+ [256**2, 512**2], # 256-512
+ [512**2, 1e5**2],
+ ] # 512-inf
+ assert area in areas, "Unknown area range: {}".format(area)
+ area_range = area_ranges[areas[area]]
+ gt_overlaps = []
+ num_pos = 0
+
+ for prediction_dict in dataset_predictions:
+ predictions = prediction_dict["proposals"]
+
+ # sort predictions in descending order
+ # TODO maybe remove this and make it explicit in the documentation
+ inds = predictions.objectness_logits.sort(descending=True)[1]
+ predictions = predictions[inds]
+
+ ann_ids = lvis_api.get_ann_ids(img_ids=[prediction_dict["image_id"]])
+ anno = lvis_api.load_anns(ann_ids)
+ gt_boxes = [
+ BoxMode.convert(obj["bbox"], BoxMode.XYWH_ABS, BoxMode.XYXY_ABS) for obj in anno
+ ]
+ gt_boxes = torch.as_tensor(gt_boxes).reshape(-1, 4) # guard against no boxes
+ gt_boxes = Boxes(gt_boxes)
+ gt_areas = torch.as_tensor([obj["area"] for obj in anno])
+
+ if len(gt_boxes) == 0 or len(predictions) == 0:
+ continue
+
+ valid_gt_inds = (gt_areas >= area_range[0]) & (gt_areas <= area_range[1])
+ gt_boxes = gt_boxes[valid_gt_inds]
+
+ num_pos += len(gt_boxes)
+
+ if len(gt_boxes) == 0:
+ continue
+
+ if limit is not None and len(predictions) > limit:
+ predictions = predictions[:limit]
+
+ overlaps = pairwise_iou(predictions.proposal_boxes, gt_boxes)
+
+ _gt_overlaps = torch.zeros(len(gt_boxes))
+ for j in range(min(len(predictions), len(gt_boxes))):
+ # find which proposal box maximally covers each gt box
+ # and get the iou amount of coverage for each gt box
+ max_overlaps, argmax_overlaps = overlaps.max(dim=0)
+
+ # find which gt box is 'best' covered (i.e. 'best' = most iou)
+ gt_ovr, gt_ind = max_overlaps.max(dim=0)
+ assert gt_ovr >= 0
+ # find the proposal box that covers the best covered gt box
+ box_ind = argmax_overlaps[gt_ind]
+ # record the iou coverage of this gt box
+ _gt_overlaps[j] = overlaps[box_ind, gt_ind]
+ assert _gt_overlaps[j] == gt_ovr
+ # mark the proposal box and the gt box as used
+ overlaps[box_ind, :] = -1
+ overlaps[:, gt_ind] = -1
+
+ # append recorded iou coverage level
+ gt_overlaps.append(_gt_overlaps)
+ gt_overlaps = (
+ torch.cat(gt_overlaps, dim=0) if len(gt_overlaps) else torch.zeros(0, dtype=torch.float32)
+ )
+ gt_overlaps, _ = torch.sort(gt_overlaps)
+
+ if thresholds is None:
+ step = 0.05
+ thresholds = torch.arange(0.5, 0.95 + 1e-5, step, dtype=torch.float32)
+ recalls = torch.zeros_like(thresholds)
+ # compute recall for each iou threshold
+ for i, t in enumerate(thresholds):
+ recalls[i] = (gt_overlaps >= t).float().sum() / float(num_pos)
+ # ar = 2 * np.trapz(recalls, thresholds)
+ ar = recalls.mean()
+ return {
+ "ar": ar,
+ "recalls": recalls,
+ "thresholds": thresholds,
+ "gt_overlaps": gt_overlaps,
+ "num_pos": num_pos,
+ }
+
+
+def _evaluate_predictions_on_lvis(
+ lvis_gt, lvis_results, iou_type, max_dets_per_image=None, class_names=None
+):
+ """
+ Args:
+ iou_type (str):
+ max_dets_per_image (None or int): limit on maximum detections per image in evaluating AP
+ This limit, by default of the LVIS dataset, is 300.
+ class_names (None or list[str]): if provided, will use it to predict
+ per-category AP.
+
+ Returns:
+ a dict of {metric name: score}
+ """
+ metrics = {
+ "bbox": ["AP", "AP50", "AP75", "APs", "APm", "APl", "APr", "APc", "APf"],
+ "segm": ["AP", "AP50", "AP75", "APs", "APm", "APl", "APr", "APc", "APf"],
+ }[iou_type]
+
+ logger = logging.getLogger(__name__)
+
+ if len(lvis_results) == 0: # TODO: check if needed
+ logger.warn("No predictions from the model!")
+ return {metric: float("nan") for metric in metrics}
+
+ if iou_type == "segm":
+ lvis_results = copy.deepcopy(lvis_results)
+ # When evaluating mask AP, if the results contain bbox, LVIS API will
+ # use the box area as the area of the instance, instead of the mask area.
+ # This leads to a different definition of small/medium/large.
+ # We remove the bbox field to let mask AP use mask area.
+ for c in lvis_results:
+ c.pop("bbox", None)
+
+ if max_dets_per_image is None:
+ max_dets_per_image = 300 # Default for LVIS dataset
+
+ from lvis import LVISEval, LVISResults
+
+ logger.info(f"Evaluating with max detections per image = {max_dets_per_image}")
+ lvis_results = LVISResults(lvis_gt, lvis_results, max_dets=max_dets_per_image)
+ lvis_eval = LVISEval(lvis_gt, lvis_results, iou_type)
+ lvis_eval.run()
+ lvis_eval.print_results()
+
+ # Pull the standard metrics from the LVIS results
+ results = lvis_eval.get_results()
+ results = {metric: float(results[metric] * 100) for metric in metrics}
+ logger.info("Evaluation results for {}: \n".format(iou_type) + create_small_table(results))
+ return results
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/panoptic_evaluation.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/panoptic_evaluation.py
new file mode 100644
index 0000000000000000000000000000000000000000..bf77fe061291f44381f8417e82e8b2bc7c5a60c6
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/panoptic_evaluation.py
@@ -0,0 +1,199 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import contextlib
+import io
+import itertools
+import json
+import logging
+import numpy as np
+import os
+import tempfile
+from collections import OrderedDict
+from typing import Optional
+from PIL import Image
+from tabulate import tabulate
+
+from annotator.oneformer.detectron2.data import MetadataCatalog
+from annotator.oneformer.detectron2.utils import comm
+from annotator.oneformer.detectron2.utils.file_io import PathManager
+
+from .evaluator import DatasetEvaluator
+
+logger = logging.getLogger(__name__)
+
+
+class COCOPanopticEvaluator(DatasetEvaluator):
+ """
+ Evaluate Panoptic Quality metrics on COCO using PanopticAPI.
+ It saves panoptic segmentation prediction in `output_dir`
+
+ It contains a synchronize call and has to be called from all workers.
+ """
+
+ def __init__(self, dataset_name: str, output_dir: Optional[str] = None):
+ """
+ Args:
+ dataset_name: name of the dataset
+ output_dir: output directory to save results for evaluation.
+ """
+ self._metadata = MetadataCatalog.get(dataset_name)
+ self._thing_contiguous_id_to_dataset_id = {
+ v: k for k, v in self._metadata.thing_dataset_id_to_contiguous_id.items()
+ }
+ self._stuff_contiguous_id_to_dataset_id = {
+ v: k for k, v in self._metadata.stuff_dataset_id_to_contiguous_id.items()
+ }
+
+ self._output_dir = output_dir
+ if self._output_dir is not None:
+ PathManager.mkdirs(self._output_dir)
+
+ def reset(self):
+ self._predictions = []
+
+ def _convert_category_id(self, segment_info):
+ isthing = segment_info.pop("isthing", None)
+ if isthing is None:
+ # the model produces panoptic category id directly. No more conversion needed
+ return segment_info
+ if isthing is True:
+ segment_info["category_id"] = self._thing_contiguous_id_to_dataset_id[
+ segment_info["category_id"]
+ ]
+ else:
+ segment_info["category_id"] = self._stuff_contiguous_id_to_dataset_id[
+ segment_info["category_id"]
+ ]
+ return segment_info
+
+ def process(self, inputs, outputs):
+ from panopticapi.utils import id2rgb
+
+ for input, output in zip(inputs, outputs):
+ panoptic_img, segments_info = output["panoptic_seg"]
+ panoptic_img = panoptic_img.cpu().numpy()
+ if segments_info is None:
+ # If "segments_info" is None, we assume "panoptic_img" is a
+ # H*W int32 image storing the panoptic_id in the format of
+ # category_id * label_divisor + instance_id. We reserve -1 for
+ # VOID label, and add 1 to panoptic_img since the official
+ # evaluation script uses 0 for VOID label.
+ label_divisor = self._metadata.label_divisor
+ segments_info = []
+ for panoptic_label in np.unique(panoptic_img):
+ if panoptic_label == -1:
+ # VOID region.
+ continue
+ pred_class = panoptic_label // label_divisor
+ isthing = (
+ pred_class in self._metadata.thing_dataset_id_to_contiguous_id.values()
+ )
+ segments_info.append(
+ {
+ "id": int(panoptic_label) + 1,
+ "category_id": int(pred_class),
+ "isthing": bool(isthing),
+ }
+ )
+ # Official evaluation script uses 0 for VOID label.
+ panoptic_img += 1
+
+ file_name = os.path.basename(input["file_name"])
+ file_name_png = os.path.splitext(file_name)[0] + ".png"
+ with io.BytesIO() as out:
+ Image.fromarray(id2rgb(panoptic_img)).save(out, format="PNG")
+ segments_info = [self._convert_category_id(x) for x in segments_info]
+ self._predictions.append(
+ {
+ "image_id": input["image_id"],
+ "file_name": file_name_png,
+ "png_string": out.getvalue(),
+ "segments_info": segments_info,
+ }
+ )
+
+ def evaluate(self):
+ comm.synchronize()
+
+ self._predictions = comm.gather(self._predictions)
+ self._predictions = list(itertools.chain(*self._predictions))
+ if not comm.is_main_process():
+ return
+
+ # PanopticApi requires local files
+ gt_json = PathManager.get_local_path(self._metadata.panoptic_json)
+ gt_folder = PathManager.get_local_path(self._metadata.panoptic_root)
+
+ with tempfile.TemporaryDirectory(prefix="panoptic_eval") as pred_dir:
+ logger.info("Writing all panoptic predictions to {} ...".format(pred_dir))
+ for p in self._predictions:
+ with open(os.path.join(pred_dir, p["file_name"]), "wb") as f:
+ f.write(p.pop("png_string"))
+
+ with open(gt_json, "r") as f:
+ json_data = json.load(f)
+ json_data["annotations"] = self._predictions
+
+ output_dir = self._output_dir or pred_dir
+ predictions_json = os.path.join(output_dir, "predictions.json")
+ with PathManager.open(predictions_json, "w") as f:
+ f.write(json.dumps(json_data))
+
+ from panopticapi.evaluation import pq_compute
+
+ with contextlib.redirect_stdout(io.StringIO()):
+ pq_res = pq_compute(
+ gt_json,
+ PathManager.get_local_path(predictions_json),
+ gt_folder=gt_folder,
+ pred_folder=pred_dir,
+ )
+
+ res = {}
+ res["PQ"] = 100 * pq_res["All"]["pq"]
+ res["SQ"] = 100 * pq_res["All"]["sq"]
+ res["RQ"] = 100 * pq_res["All"]["rq"]
+ res["PQ_th"] = 100 * pq_res["Things"]["pq"]
+ res["SQ_th"] = 100 * pq_res["Things"]["sq"]
+ res["RQ_th"] = 100 * pq_res["Things"]["rq"]
+ res["PQ_st"] = 100 * pq_res["Stuff"]["pq"]
+ res["SQ_st"] = 100 * pq_res["Stuff"]["sq"]
+ res["RQ_st"] = 100 * pq_res["Stuff"]["rq"]
+
+ results = OrderedDict({"panoptic_seg": res})
+ _print_panoptic_results(pq_res)
+
+ return results
+
+
+def _print_panoptic_results(pq_res):
+ headers = ["", "PQ", "SQ", "RQ", "#categories"]
+ data = []
+ for name in ["All", "Things", "Stuff"]:
+ row = [name] + [pq_res[name][k] * 100 for k in ["pq", "sq", "rq"]] + [pq_res[name]["n"]]
+ data.append(row)
+ table = tabulate(
+ data, headers=headers, tablefmt="pipe", floatfmt=".3f", stralign="center", numalign="center"
+ )
+ logger.info("Panoptic Evaluation Results:\n" + table)
+
+
+if __name__ == "__main__":
+ from annotator.oneformer.detectron2.utils.logger import setup_logger
+
+ logger = setup_logger()
+ import argparse
+
+ parser = argparse.ArgumentParser()
+ parser.add_argument("--gt-json")
+ parser.add_argument("--gt-dir")
+ parser.add_argument("--pred-json")
+ parser.add_argument("--pred-dir")
+ args = parser.parse_args()
+
+ from panopticapi.evaluation import pq_compute
+
+ with contextlib.redirect_stdout(io.StringIO()):
+ pq_res = pq_compute(
+ args.gt_json, args.pred_json, gt_folder=args.gt_dir, pred_folder=args.pred_dir
+ )
+ _print_panoptic_results(pq_res)
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/pascal_voc_evaluation.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/pascal_voc_evaluation.py
new file mode 100644
index 0000000000000000000000000000000000000000..b2963e5dc5b6ed471f0c37056b35a350ea4cf020
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/pascal_voc_evaluation.py
@@ -0,0 +1,300 @@
+# -*- coding: utf-8 -*-
+# Copyright (c) Facebook, Inc. and its affiliates.
+
+import logging
+import numpy as np
+import os
+import tempfile
+import xml.etree.ElementTree as ET
+from collections import OrderedDict, defaultdict
+from functools import lru_cache
+import torch
+
+from annotator.oneformer.detectron2.data import MetadataCatalog
+from annotator.oneformer.detectron2.utils import comm
+from annotator.oneformer.detectron2.utils.file_io import PathManager
+
+from .evaluator import DatasetEvaluator
+
+
+class PascalVOCDetectionEvaluator(DatasetEvaluator):
+ """
+ Evaluate Pascal VOC style AP for Pascal VOC dataset.
+ It contains a synchronization, therefore has to be called from all ranks.
+
+ Note that the concept of AP can be implemented in different ways and may not
+ produce identical results. This class mimics the implementation of the official
+ Pascal VOC Matlab API, and should produce similar but not identical results to the
+ official API.
+ """
+
+ def __init__(self, dataset_name):
+ """
+ Args:
+ dataset_name (str): name of the dataset, e.g., "voc_2007_test"
+ """
+ self._dataset_name = dataset_name
+ meta = MetadataCatalog.get(dataset_name)
+
+ # Too many tiny files, download all to local for speed.
+ annotation_dir_local = PathManager.get_local_path(
+ os.path.join(meta.dirname, "Annotations/")
+ )
+ self._anno_file_template = os.path.join(annotation_dir_local, "{}.xml")
+ self._image_set_path = os.path.join(meta.dirname, "ImageSets", "Main", meta.split + ".txt")
+ self._class_names = meta.thing_classes
+ assert meta.year in [2007, 2012], meta.year
+ self._is_2007 = meta.year == 2007
+ self._cpu_device = torch.device("cpu")
+ self._logger = logging.getLogger(__name__)
+
+ def reset(self):
+ self._predictions = defaultdict(list) # class name -> list of prediction strings
+
+ def process(self, inputs, outputs):
+ for input, output in zip(inputs, outputs):
+ image_id = input["image_id"]
+ instances = output["instances"].to(self._cpu_device)
+ boxes = instances.pred_boxes.tensor.numpy()
+ scores = instances.scores.tolist()
+ classes = instances.pred_classes.tolist()
+ for box, score, cls in zip(boxes, scores, classes):
+ xmin, ymin, xmax, ymax = box
+ # The inverse of data loading logic in `datasets/pascal_voc.py`
+ xmin += 1
+ ymin += 1
+ self._predictions[cls].append(
+ f"{image_id} {score:.3f} {xmin:.1f} {ymin:.1f} {xmax:.1f} {ymax:.1f}"
+ )
+
+ def evaluate(self):
+ """
+ Returns:
+ dict: has a key "segm", whose value is a dict of "AP", "AP50", and "AP75".
+ """
+ all_predictions = comm.gather(self._predictions, dst=0)
+ if not comm.is_main_process():
+ return
+ predictions = defaultdict(list)
+ for predictions_per_rank in all_predictions:
+ for clsid, lines in predictions_per_rank.items():
+ predictions[clsid].extend(lines)
+ del all_predictions
+
+ self._logger.info(
+ "Evaluating {} using {} metric. "
+ "Note that results do not use the official Matlab API.".format(
+ self._dataset_name, 2007 if self._is_2007 else 2012
+ )
+ )
+
+ with tempfile.TemporaryDirectory(prefix="pascal_voc_eval_") as dirname:
+ res_file_template = os.path.join(dirname, "{}.txt")
+
+ aps = defaultdict(list) # iou -> ap per class
+ for cls_id, cls_name in enumerate(self._class_names):
+ lines = predictions.get(cls_id, [""])
+
+ with open(res_file_template.format(cls_name), "w") as f:
+ f.write("\n".join(lines))
+
+ for thresh in range(50, 100, 5):
+ rec, prec, ap = voc_eval(
+ res_file_template,
+ self._anno_file_template,
+ self._image_set_path,
+ cls_name,
+ ovthresh=thresh / 100.0,
+ use_07_metric=self._is_2007,
+ )
+ aps[thresh].append(ap * 100)
+
+ ret = OrderedDict()
+ mAP = {iou: np.mean(x) for iou, x in aps.items()}
+ ret["bbox"] = {"AP": np.mean(list(mAP.values())), "AP50": mAP[50], "AP75": mAP[75]}
+ return ret
+
+
+##############################################################################
+#
+# Below code is modified from
+# https://github.com/rbgirshick/py-faster-rcnn/blob/master/lib/datasets/voc_eval.py
+# --------------------------------------------------------
+# Fast/er R-CNN
+# Licensed under The MIT License [see LICENSE for details]
+# Written by Bharath Hariharan
+# --------------------------------------------------------
+
+"""Python implementation of the PASCAL VOC devkit's AP evaluation code."""
+
+
+@lru_cache(maxsize=None)
+def parse_rec(filename):
+ """Parse a PASCAL VOC xml file."""
+ with PathManager.open(filename) as f:
+ tree = ET.parse(f)
+ objects = []
+ for obj in tree.findall("object"):
+ obj_struct = {}
+ obj_struct["name"] = obj.find("name").text
+ obj_struct["pose"] = obj.find("pose").text
+ obj_struct["truncated"] = int(obj.find("truncated").text)
+ obj_struct["difficult"] = int(obj.find("difficult").text)
+ bbox = obj.find("bndbox")
+ obj_struct["bbox"] = [
+ int(bbox.find("xmin").text),
+ int(bbox.find("ymin").text),
+ int(bbox.find("xmax").text),
+ int(bbox.find("ymax").text),
+ ]
+ objects.append(obj_struct)
+
+ return objects
+
+
+def voc_ap(rec, prec, use_07_metric=False):
+ """Compute VOC AP given precision and recall. If use_07_metric is true, uses
+ the VOC 07 11-point method (default:False).
+ """
+ if use_07_metric:
+ # 11 point metric
+ ap = 0.0
+ for t in np.arange(0.0, 1.1, 0.1):
+ if np.sum(rec >= t) == 0:
+ p = 0
+ else:
+ p = np.max(prec[rec >= t])
+ ap = ap + p / 11.0
+ else:
+ # correct AP calculation
+ # first append sentinel values at the end
+ mrec = np.concatenate(([0.0], rec, [1.0]))
+ mpre = np.concatenate(([0.0], prec, [0.0]))
+
+ # compute the precision envelope
+ for i in range(mpre.size - 1, 0, -1):
+ mpre[i - 1] = np.maximum(mpre[i - 1], mpre[i])
+
+ # to calculate area under PR curve, look for points
+ # where X axis (recall) changes value
+ i = np.where(mrec[1:] != mrec[:-1])[0]
+
+ # and sum (\Delta recall) * prec
+ ap = np.sum((mrec[i + 1] - mrec[i]) * mpre[i + 1])
+ return ap
+
+
+def voc_eval(detpath, annopath, imagesetfile, classname, ovthresh=0.5, use_07_metric=False):
+ """rec, prec, ap = voc_eval(detpath,
+ annopath,
+ imagesetfile,
+ classname,
+ [ovthresh],
+ [use_07_metric])
+
+ Top level function that does the PASCAL VOC evaluation.
+
+ detpath: Path to detections
+ detpath.format(classname) should produce the detection results file.
+ annopath: Path to annotations
+ annopath.format(imagename) should be the xml annotations file.
+ imagesetfile: Text file containing the list of images, one image per line.
+ classname: Category name (duh)
+ [ovthresh]: Overlap threshold (default = 0.5)
+ [use_07_metric]: Whether to use VOC07's 11 point AP computation
+ (default False)
+ """
+ # assumes detections are in detpath.format(classname)
+ # assumes annotations are in annopath.format(imagename)
+ # assumes imagesetfile is a text file with each line an image name
+
+ # first load gt
+ # read list of images
+ with PathManager.open(imagesetfile, "r") as f:
+ lines = f.readlines()
+ imagenames = [x.strip() for x in lines]
+
+ # load annots
+ recs = {}
+ for imagename in imagenames:
+ recs[imagename] = parse_rec(annopath.format(imagename))
+
+ # extract gt objects for this class
+ class_recs = {}
+ npos = 0
+ for imagename in imagenames:
+ R = [obj for obj in recs[imagename] if obj["name"] == classname]
+ bbox = np.array([x["bbox"] for x in R])
+ difficult = np.array([x["difficult"] for x in R]).astype(bool)
+ # difficult = np.array([False for x in R]).astype(bool) # treat all "difficult" as GT
+ det = [False] * len(R)
+ npos = npos + sum(~difficult)
+ class_recs[imagename] = {"bbox": bbox, "difficult": difficult, "det": det}
+
+ # read dets
+ detfile = detpath.format(classname)
+ with open(detfile, "r") as f:
+ lines = f.readlines()
+
+ splitlines = [x.strip().split(" ") for x in lines]
+ image_ids = [x[0] for x in splitlines]
+ confidence = np.array([float(x[1]) for x in splitlines])
+ BB = np.array([[float(z) for z in x[2:]] for x in splitlines]).reshape(-1, 4)
+
+ # sort by confidence
+ sorted_ind = np.argsort(-confidence)
+ BB = BB[sorted_ind, :]
+ image_ids = [image_ids[x] for x in sorted_ind]
+
+ # go down dets and mark TPs and FPs
+ nd = len(image_ids)
+ tp = np.zeros(nd)
+ fp = np.zeros(nd)
+ for d in range(nd):
+ R = class_recs[image_ids[d]]
+ bb = BB[d, :].astype(float)
+ ovmax = -np.inf
+ BBGT = R["bbox"].astype(float)
+
+ if BBGT.size > 0:
+ # compute overlaps
+ # intersection
+ ixmin = np.maximum(BBGT[:, 0], bb[0])
+ iymin = np.maximum(BBGT[:, 1], bb[1])
+ ixmax = np.minimum(BBGT[:, 2], bb[2])
+ iymax = np.minimum(BBGT[:, 3], bb[3])
+ iw = np.maximum(ixmax - ixmin + 1.0, 0.0)
+ ih = np.maximum(iymax - iymin + 1.0, 0.0)
+ inters = iw * ih
+
+ # union
+ uni = (
+ (bb[2] - bb[0] + 1.0) * (bb[3] - bb[1] + 1.0)
+ + (BBGT[:, 2] - BBGT[:, 0] + 1.0) * (BBGT[:, 3] - BBGT[:, 1] + 1.0)
+ - inters
+ )
+
+ overlaps = inters / uni
+ ovmax = np.max(overlaps)
+ jmax = np.argmax(overlaps)
+
+ if ovmax > ovthresh:
+ if not R["difficult"][jmax]:
+ if not R["det"][jmax]:
+ tp[d] = 1.0
+ R["det"][jmax] = 1
+ else:
+ fp[d] = 1.0
+ else:
+ fp[d] = 1.0
+
+ # compute precision recall
+ fp = np.cumsum(fp)
+ tp = np.cumsum(tp)
+ rec = tp / float(npos)
+ # avoid divide by zero in case the first detection matches a difficult
+ # ground truth
+ prec = tp / np.maximum(tp + fp, np.finfo(np.float64).eps)
+ ap = voc_ap(rec, prec, use_07_metric)
+
+ return rec, prec, ap
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/rotated_coco_evaluation.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/rotated_coco_evaluation.py
new file mode 100644
index 0000000000000000000000000000000000000000..4cf954d751dfe25367ce6059626b7118b34bb45a
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/rotated_coco_evaluation.py
@@ -0,0 +1,207 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import itertools
+import json
+import numpy as np
+import os
+import torch
+from pycocotools.cocoeval import COCOeval, maskUtils
+
+from annotator.oneformer.detectron2.structures import BoxMode, RotatedBoxes, pairwise_iou_rotated
+from annotator.oneformer.detectron2.utils.file_io import PathManager
+
+from .coco_evaluation import COCOEvaluator
+
+
+class RotatedCOCOeval(COCOeval):
+ @staticmethod
+ def is_rotated(box_list):
+ if type(box_list) == np.ndarray:
+ return box_list.shape[1] == 5
+ elif type(box_list) == list:
+ if box_list == []: # cannot decide the box_dim
+ return False
+ return np.all(
+ np.array(
+ [
+ (len(obj) == 5) and ((type(obj) == list) or (type(obj) == np.ndarray))
+ for obj in box_list
+ ]
+ )
+ )
+ return False
+
+ @staticmethod
+ def boxlist_to_tensor(boxlist, output_box_dim):
+ if type(boxlist) == np.ndarray:
+ box_tensor = torch.from_numpy(boxlist)
+ elif type(boxlist) == list:
+ if boxlist == []:
+ return torch.zeros((0, output_box_dim), dtype=torch.float32)
+ else:
+ box_tensor = torch.FloatTensor(boxlist)
+ else:
+ raise Exception("Unrecognized boxlist type")
+
+ input_box_dim = box_tensor.shape[1]
+ if input_box_dim != output_box_dim:
+ if input_box_dim == 4 and output_box_dim == 5:
+ box_tensor = BoxMode.convert(box_tensor, BoxMode.XYWH_ABS, BoxMode.XYWHA_ABS)
+ else:
+ raise Exception(
+ "Unable to convert from {}-dim box to {}-dim box".format(
+ input_box_dim, output_box_dim
+ )
+ )
+ return box_tensor
+
+ def compute_iou_dt_gt(self, dt, gt, is_crowd):
+ if self.is_rotated(dt) or self.is_rotated(gt):
+ # TODO: take is_crowd into consideration
+ assert all(c == 0 for c in is_crowd)
+ dt = RotatedBoxes(self.boxlist_to_tensor(dt, output_box_dim=5))
+ gt = RotatedBoxes(self.boxlist_to_tensor(gt, output_box_dim=5))
+ return pairwise_iou_rotated(dt, gt)
+ else:
+ # This is the same as the classical COCO evaluation
+ return maskUtils.iou(dt, gt, is_crowd)
+
+ def computeIoU(self, imgId, catId):
+ p = self.params
+ if p.useCats:
+ gt = self._gts[imgId, catId]
+ dt = self._dts[imgId, catId]
+ else:
+ gt = [_ for cId in p.catIds for _ in self._gts[imgId, cId]]
+ dt = [_ for cId in p.catIds for _ in self._dts[imgId, cId]]
+ if len(gt) == 0 and len(dt) == 0:
+ return []
+ inds = np.argsort([-d["score"] for d in dt], kind="mergesort")
+ dt = [dt[i] for i in inds]
+ if len(dt) > p.maxDets[-1]:
+ dt = dt[0 : p.maxDets[-1]]
+
+ assert p.iouType == "bbox", "unsupported iouType for iou computation"
+
+ g = [g["bbox"] for g in gt]
+ d = [d["bbox"] for d in dt]
+
+ # compute iou between each dt and gt region
+ iscrowd = [int(o["iscrowd"]) for o in gt]
+
+ # Note: this function is copied from cocoeval.py in cocoapi
+ # and the major difference is here.
+ ious = self.compute_iou_dt_gt(d, g, iscrowd)
+ return ious
+
+
+class RotatedCOCOEvaluator(COCOEvaluator):
+ """
+ Evaluate object proposal/instance detection outputs using COCO-like metrics and APIs,
+ with rotated boxes support.
+ Note: this uses IOU only and does not consider angle differences.
+ """
+
+ def process(self, inputs, outputs):
+ """
+ Args:
+ inputs: the inputs to a COCO model (e.g., GeneralizedRCNN).
+ It is a list of dict. Each dict corresponds to an image and
+ contains keys like "height", "width", "file_name", "image_id".
+ outputs: the outputs of a COCO model. It is a list of dicts with key
+ "instances" that contains :class:`Instances`.
+ """
+ for input, output in zip(inputs, outputs):
+ prediction = {"image_id": input["image_id"]}
+
+ if "instances" in output:
+ instances = output["instances"].to(self._cpu_device)
+
+ prediction["instances"] = self.instances_to_json(instances, input["image_id"])
+ if "proposals" in output:
+ prediction["proposals"] = output["proposals"].to(self._cpu_device)
+ self._predictions.append(prediction)
+
+ def instances_to_json(self, instances, img_id):
+ num_instance = len(instances)
+ if num_instance == 0:
+ return []
+
+ boxes = instances.pred_boxes.tensor.numpy()
+ if boxes.shape[1] == 4:
+ boxes = BoxMode.convert(boxes, BoxMode.XYXY_ABS, BoxMode.XYWH_ABS)
+ boxes = boxes.tolist()
+ scores = instances.scores.tolist()
+ classes = instances.pred_classes.tolist()
+
+ results = []
+ for k in range(num_instance):
+ result = {
+ "image_id": img_id,
+ "category_id": classes[k],
+ "bbox": boxes[k],
+ "score": scores[k],
+ }
+
+ results.append(result)
+ return results
+
+ def _eval_predictions(self, predictions, img_ids=None): # img_ids: unused
+ """
+ Evaluate predictions on the given tasks.
+ Fill self._results with the metrics of the tasks.
+ """
+ self._logger.info("Preparing results for COCO format ...")
+ coco_results = list(itertools.chain(*[x["instances"] for x in predictions]))
+
+ # unmap the category ids for COCO
+ if hasattr(self._metadata, "thing_dataset_id_to_contiguous_id"):
+ reverse_id_mapping = {
+ v: k for k, v in self._metadata.thing_dataset_id_to_contiguous_id.items()
+ }
+ for result in coco_results:
+ result["category_id"] = reverse_id_mapping[result["category_id"]]
+
+ if self._output_dir:
+ file_path = os.path.join(self._output_dir, "coco_instances_results.json")
+ self._logger.info("Saving results to {}".format(file_path))
+ with PathManager.open(file_path, "w") as f:
+ f.write(json.dumps(coco_results))
+ f.flush()
+
+ if not self._do_evaluation:
+ self._logger.info("Annotations are not available for evaluation.")
+ return
+
+ self._logger.info("Evaluating predictions ...")
+
+ assert self._tasks is None or set(self._tasks) == {
+ "bbox"
+ }, "[RotatedCOCOEvaluator] Only bbox evaluation is supported"
+ coco_eval = (
+ self._evaluate_predictions_on_coco(self._coco_api, coco_results)
+ if len(coco_results) > 0
+ else None # cocoapi does not handle empty results very well
+ )
+
+ task = "bbox"
+ res = self._derive_coco_results(
+ coco_eval, task, class_names=self._metadata.get("thing_classes")
+ )
+ self._results[task] = res
+
+ def _evaluate_predictions_on_coco(self, coco_gt, coco_results):
+ """
+ Evaluate the coco results using COCOEval API.
+ """
+ assert len(coco_results) > 0
+
+ coco_dt = coco_gt.loadRes(coco_results)
+
+ # Only bbox is supported for now
+ coco_eval = RotatedCOCOeval(coco_gt, coco_dt, iouType="bbox")
+
+ coco_eval.evaluate()
+ coco_eval.accumulate()
+ coco_eval.summarize()
+
+ return coco_eval
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/sem_seg_evaluation.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/sem_seg_evaluation.py
new file mode 100644
index 0000000000000000000000000000000000000000..f8bc0e901954fc0eefca6386bcf8ad31e0e66277
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/sem_seg_evaluation.py
@@ -0,0 +1,265 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import itertools
+import json
+import logging
+import numpy as np
+import os
+from collections import OrderedDict
+from typing import Optional, Union
+import pycocotools.mask as mask_util
+import torch
+from PIL import Image
+
+from annotator.oneformer.detectron2.data import DatasetCatalog, MetadataCatalog
+from annotator.oneformer.detectron2.utils.comm import all_gather, is_main_process, synchronize
+from annotator.oneformer.detectron2.utils.file_io import PathManager
+
+from .evaluator import DatasetEvaluator
+
+_CV2_IMPORTED = True
+try:
+ import cv2 # noqa
+except ImportError:
+ # OpenCV is an optional dependency at the moment
+ _CV2_IMPORTED = False
+
+
+def load_image_into_numpy_array(
+ filename: str,
+ copy: bool = False,
+ dtype: Optional[Union[np.dtype, str]] = None,
+) -> np.ndarray:
+ with PathManager.open(filename, "rb") as f:
+ array = np.array(Image.open(f), copy=copy, dtype=dtype)
+ return array
+
+
+class SemSegEvaluator(DatasetEvaluator):
+ """
+ Evaluate semantic segmentation metrics.
+ """
+
+ def __init__(
+ self,
+ dataset_name,
+ distributed=True,
+ output_dir=None,
+ *,
+ sem_seg_loading_fn=load_image_into_numpy_array,
+ num_classes=None,
+ ignore_label=None,
+ ):
+ """
+ Args:
+ dataset_name (str): name of the dataset to be evaluated.
+ distributed (bool): if True, will collect results from all ranks for evaluation.
+ Otherwise, will evaluate the results in the current process.
+ output_dir (str): an output directory to dump results.
+ sem_seg_loading_fn: function to read sem seg file and load into numpy array.
+ Default provided, but projects can customize.
+ num_classes, ignore_label: deprecated argument
+ """
+ self._logger = logging.getLogger(__name__)
+ if num_classes is not None:
+ self._logger.warn(
+ "SemSegEvaluator(num_classes) is deprecated! It should be obtained from metadata."
+ )
+ if ignore_label is not None:
+ self._logger.warn(
+ "SemSegEvaluator(ignore_label) is deprecated! It should be obtained from metadata."
+ )
+ self._dataset_name = dataset_name
+ self._distributed = distributed
+ self._output_dir = output_dir
+
+ self._cpu_device = torch.device("cpu")
+
+ self.input_file_to_gt_file = {
+ dataset_record["file_name"]: dataset_record["sem_seg_file_name"]
+ for dataset_record in DatasetCatalog.get(dataset_name)
+ }
+
+ meta = MetadataCatalog.get(dataset_name)
+ # Dict that maps contiguous training ids to COCO category ids
+ try:
+ c2d = meta.stuff_dataset_id_to_contiguous_id
+ self._contiguous_id_to_dataset_id = {v: k for k, v in c2d.items()}
+ except AttributeError:
+ self._contiguous_id_to_dataset_id = None
+ self._class_names = meta.stuff_classes
+ self.sem_seg_loading_fn = sem_seg_loading_fn
+ self._num_classes = len(meta.stuff_classes)
+ if num_classes is not None:
+ assert self._num_classes == num_classes, f"{self._num_classes} != {num_classes}"
+ self._ignore_label = ignore_label if ignore_label is not None else meta.ignore_label
+
+ # This is because cv2.erode did not work for int datatype. Only works for uint8.
+ self._compute_boundary_iou = True
+ if not _CV2_IMPORTED:
+ self._compute_boundary_iou = False
+ self._logger.warn(
+ """Boundary IoU calculation requires OpenCV. B-IoU metrics are
+ not going to be computed because OpenCV is not available to import."""
+ )
+ if self._num_classes >= np.iinfo(np.uint8).max:
+ self._compute_boundary_iou = False
+ self._logger.warn(
+ f"""SemSegEvaluator(num_classes) is more than supported value for Boundary IoU calculation!
+ B-IoU metrics are not going to be computed. Max allowed value (exclusive)
+ for num_classes for calculating Boundary IoU is {np.iinfo(np.uint8).max}.
+ The number of classes of dataset {self._dataset_name} is {self._num_classes}"""
+ )
+
+ def reset(self):
+ self._conf_matrix = np.zeros((self._num_classes + 1, self._num_classes + 1), dtype=np.int64)
+ self._b_conf_matrix = np.zeros(
+ (self._num_classes + 1, self._num_classes + 1), dtype=np.int64
+ )
+ self._predictions = []
+
+ def process(self, inputs, outputs):
+ """
+ Args:
+ inputs: the inputs to a model.
+ It is a list of dicts. Each dict corresponds to an image and
+ contains keys like "height", "width", "file_name".
+ outputs: the outputs of a model. It is either list of semantic segmentation predictions
+ (Tensor [H, W]) or list of dicts with key "sem_seg" that contains semantic
+ segmentation prediction in the same format.
+ """
+ for input, output in zip(inputs, outputs):
+ output = output["sem_seg"].argmax(dim=0).to(self._cpu_device)
+ pred = np.array(output, dtype=np.int)
+ gt_filename = self.input_file_to_gt_file[input["file_name"]]
+ gt = self.sem_seg_loading_fn(gt_filename, dtype=np.int)
+
+ gt[gt == self._ignore_label] = self._num_classes
+
+ self._conf_matrix += np.bincount(
+ (self._num_classes + 1) * pred.reshape(-1) + gt.reshape(-1),
+ minlength=self._conf_matrix.size,
+ ).reshape(self._conf_matrix.shape)
+
+ if self._compute_boundary_iou:
+ b_gt = self._mask_to_boundary(gt.astype(np.uint8))
+ b_pred = self._mask_to_boundary(pred.astype(np.uint8))
+
+ self._b_conf_matrix += np.bincount(
+ (self._num_classes + 1) * b_pred.reshape(-1) + b_gt.reshape(-1),
+ minlength=self._conf_matrix.size,
+ ).reshape(self._conf_matrix.shape)
+
+ self._predictions.extend(self.encode_json_sem_seg(pred, input["file_name"]))
+
+ def evaluate(self):
+ """
+ Evaluates standard semantic segmentation metrics (http://cocodataset.org/#stuff-eval):
+
+ * Mean intersection-over-union averaged across classes (mIoU)
+ * Frequency Weighted IoU (fwIoU)
+ * Mean pixel accuracy averaged across classes (mACC)
+ * Pixel Accuracy (pACC)
+ """
+ if self._distributed:
+ synchronize()
+ conf_matrix_list = all_gather(self._conf_matrix)
+ b_conf_matrix_list = all_gather(self._b_conf_matrix)
+ self._predictions = all_gather(self._predictions)
+ self._predictions = list(itertools.chain(*self._predictions))
+ if not is_main_process():
+ return
+
+ self._conf_matrix = np.zeros_like(self._conf_matrix)
+ for conf_matrix in conf_matrix_list:
+ self._conf_matrix += conf_matrix
+
+ self._b_conf_matrix = np.zeros_like(self._b_conf_matrix)
+ for b_conf_matrix in b_conf_matrix_list:
+ self._b_conf_matrix += b_conf_matrix
+
+ if self._output_dir:
+ PathManager.mkdirs(self._output_dir)
+ file_path = os.path.join(self._output_dir, "sem_seg_predictions.json")
+ with PathManager.open(file_path, "w") as f:
+ f.write(json.dumps(self._predictions))
+
+ acc = np.full(self._num_classes, np.nan, dtype=np.float)
+ iou = np.full(self._num_classes, np.nan, dtype=np.float)
+ tp = self._conf_matrix.diagonal()[:-1].astype(np.float)
+ pos_gt = np.sum(self._conf_matrix[:-1, :-1], axis=0).astype(np.float)
+ class_weights = pos_gt / np.sum(pos_gt)
+ pos_pred = np.sum(self._conf_matrix[:-1, :-1], axis=1).astype(np.float)
+ acc_valid = pos_gt > 0
+ acc[acc_valid] = tp[acc_valid] / pos_gt[acc_valid]
+ union = pos_gt + pos_pred - tp
+ iou_valid = np.logical_and(acc_valid, union > 0)
+ iou[iou_valid] = tp[iou_valid] / union[iou_valid]
+ macc = np.sum(acc[acc_valid]) / np.sum(acc_valid)
+ miou = np.sum(iou[iou_valid]) / np.sum(iou_valid)
+ fiou = np.sum(iou[iou_valid] * class_weights[iou_valid])
+ pacc = np.sum(tp) / np.sum(pos_gt)
+
+ if self._compute_boundary_iou:
+ b_iou = np.full(self._num_classes, np.nan, dtype=np.float)
+ b_tp = self._b_conf_matrix.diagonal()[:-1].astype(np.float)
+ b_pos_gt = np.sum(self._b_conf_matrix[:-1, :-1], axis=0).astype(np.float)
+ b_pos_pred = np.sum(self._b_conf_matrix[:-1, :-1], axis=1).astype(np.float)
+ b_union = b_pos_gt + b_pos_pred - b_tp
+ b_iou_valid = b_union > 0
+ b_iou[b_iou_valid] = b_tp[b_iou_valid] / b_union[b_iou_valid]
+
+ res = {}
+ res["mIoU"] = 100 * miou
+ res["fwIoU"] = 100 * fiou
+ for i, name in enumerate(self._class_names):
+ res[f"IoU-{name}"] = 100 * iou[i]
+ if self._compute_boundary_iou:
+ res[f"BoundaryIoU-{name}"] = 100 * b_iou[i]
+ res[f"min(IoU, B-Iou)-{name}"] = 100 * min(iou[i], b_iou[i])
+ res["mACC"] = 100 * macc
+ res["pACC"] = 100 * pacc
+ for i, name in enumerate(self._class_names):
+ res[f"ACC-{name}"] = 100 * acc[i]
+
+ if self._output_dir:
+ file_path = os.path.join(self._output_dir, "sem_seg_evaluation.pth")
+ with PathManager.open(file_path, "wb") as f:
+ torch.save(res, f)
+ results = OrderedDict({"sem_seg": res})
+ self._logger.info(results)
+ return results
+
+ def encode_json_sem_seg(self, sem_seg, input_file_name):
+ """
+ Convert semantic segmentation to COCO stuff format with segments encoded as RLEs.
+ See http://cocodataset.org/#format-results
+ """
+ json_list = []
+ for label in np.unique(sem_seg):
+ if self._contiguous_id_to_dataset_id is not None:
+ assert (
+ label in self._contiguous_id_to_dataset_id
+ ), "Label {} is not in the metadata info for {}".format(label, self._dataset_name)
+ dataset_id = self._contiguous_id_to_dataset_id[label]
+ else:
+ dataset_id = int(label)
+ mask = (sem_seg == label).astype(np.uint8)
+ mask_rle = mask_util.encode(np.array(mask[:, :, None], order="F"))[0]
+ mask_rle["counts"] = mask_rle["counts"].decode("utf-8")
+ json_list.append(
+ {"file_name": input_file_name, "category_id": dataset_id, "segmentation": mask_rle}
+ )
+ return json_list
+
+ def _mask_to_boundary(self, mask: np.ndarray, dilation_ratio=0.02):
+ assert mask.ndim == 2, "mask_to_boundary expects a 2-dimensional image"
+ h, w = mask.shape
+ diag_len = np.sqrt(h**2 + w**2)
+ dilation = max(1, int(round(dilation_ratio * diag_len)))
+ kernel = np.ones((3, 3), dtype=np.uint8)
+
+ padded_mask = cv2.copyMakeBorder(mask, 1, 1, 1, 1, cv2.BORDER_CONSTANT, value=0)
+ eroded_mask_with_padding = cv2.erode(padded_mask, kernel, iterations=dilation)
+ eroded_mask = eroded_mask_with_padding[1:-1, 1:-1]
+ boundary = mask - eroded_mask
+ return boundary
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/testing.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/testing.py
new file mode 100644
index 0000000000000000000000000000000000000000..9e5ae625bb0593fc20739dd3ea549157e4df4f3d
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/evaluation/testing.py
@@ -0,0 +1,85 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import logging
+import numpy as np
+import pprint
+import sys
+from collections.abc import Mapping
+
+
+def print_csv_format(results):
+ """
+ Print main metrics in a format similar to Detectron,
+ so that they are easy to copypaste into a spreadsheet.
+
+ Args:
+ results (OrderedDict[dict]): task_name -> {metric -> score}
+ unordered dict can also be printed, but in arbitrary order
+ """
+ assert isinstance(results, Mapping) or not len(results), results
+ logger = logging.getLogger(__name__)
+ for task, res in results.items():
+ if isinstance(res, Mapping):
+ # Don't print "AP-category" metrics since they are usually not tracked.
+ important_res = [(k, v) for k, v in res.items() if "-" not in k]
+ logger.info("copypaste: Task: {}".format(task))
+ logger.info("copypaste: " + ",".join([k[0] for k in important_res]))
+ logger.info("copypaste: " + ",".join(["{0:.4f}".format(k[1]) for k in important_res]))
+ else:
+ logger.info(f"copypaste: {task}={res}")
+
+
+def verify_results(cfg, results):
+ """
+ Args:
+ results (OrderedDict[dict]): task_name -> {metric -> score}
+
+ Returns:
+ bool: whether the verification succeeds or not
+ """
+ expected_results = cfg.TEST.EXPECTED_RESULTS
+ if not len(expected_results):
+ return True
+
+ ok = True
+ for task, metric, expected, tolerance in expected_results:
+ actual = results[task].get(metric, None)
+ if actual is None:
+ ok = False
+ continue
+ if not np.isfinite(actual):
+ ok = False
+ continue
+ diff = abs(actual - expected)
+ if diff > tolerance:
+ ok = False
+
+ logger = logging.getLogger(__name__)
+ if not ok:
+ logger.error("Result verification failed!")
+ logger.error("Expected Results: " + str(expected_results))
+ logger.error("Actual Results: " + pprint.pformat(results))
+
+ sys.exit(1)
+ else:
+ logger.info("Results verification passed.")
+ return ok
+
+
+def flatten_results_dict(results):
+ """
+ Expand a hierarchical dict of scalars into a flat dict of scalars.
+ If results[k1][k2][k3] = v, the returned dict will have the entry
+ {"k1/k2/k3": v}.
+
+ Args:
+ results (dict):
+ """
+ r = {}
+ for k, v in results.items():
+ if isinstance(v, Mapping):
+ v = flatten_results_dict(v)
+ for kk, vv in v.items():
+ r[k + "/" + kk] = vv
+ else:
+ r[k] = v
+ return r
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/README.md b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..c86ff62516f4e8e4b1a6c1f33f11192933cf3861
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/README.md
@@ -0,0 +1,15 @@
+
+This directory contains code to prepare a detectron2 model for deployment.
+Currently it supports exporting a detectron2 model to TorchScript, ONNX, or (deprecated) Caffe2 format.
+
+Please see [documentation](https://detectron2.readthedocs.io/tutorials/deployment.html) for its usage.
+
+
+### Acknowledgements
+
+Thanks to Mobile Vision team at Facebook for developing the Caffe2 conversion tools.
+
+Thanks to Computing Platform Department - PAI team at Alibaba Group (@bddpqq, @chenbohua3) who
+help export Detectron2 models to TorchScript.
+
+Thanks to ONNX Converter team at Microsoft who help export Detectron2 models to ONNX.
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/__init__.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..5a58758f64aae6071fa688be4400622ce6036efa
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/__init__.py
@@ -0,0 +1,30 @@
+# -*- coding: utf-8 -*-
+
+import warnings
+
+from .flatten import TracingAdapter
+from .torchscript import dump_torchscript_IR, scripting_with_instances
+
+try:
+ from caffe2.proto import caffe2_pb2 as _tmp
+ from caffe2.python import core
+
+ # caffe2 is optional
+except ImportError:
+ pass
+else:
+ from .api import *
+
+
+# TODO: Update ONNX Opset version and run tests when a newer PyTorch is supported
+STABLE_ONNX_OPSET_VERSION = 11
+
+
+def add_export_config(cfg):
+ warnings.warn(
+ "add_export_config has been deprecated and behaves as no-op function.", DeprecationWarning
+ )
+ return cfg
+
+
+__all__ = [k for k in globals().keys() if not k.startswith("_")]
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/api.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/api.py
new file mode 100644
index 0000000000000000000000000000000000000000..cf1a27a4806ca83d97f5cd8c27726ec29f4e7e50
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/api.py
@@ -0,0 +1,230 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import copy
+import logging
+import os
+import torch
+from caffe2.proto import caffe2_pb2
+from torch import nn
+
+from annotator.oneformer.detectron2.config import CfgNode
+from annotator.oneformer.detectron2.utils.file_io import PathManager
+
+from .caffe2_inference import ProtobufDetectionModel
+from .caffe2_modeling import META_ARCH_CAFFE2_EXPORT_TYPE_MAP, convert_batched_inputs_to_c2_format
+from .shared import get_pb_arg_vali, get_pb_arg_vals, save_graph
+
+__all__ = [
+ "Caffe2Model",
+ "Caffe2Tracer",
+]
+
+
+class Caffe2Tracer:
+ """
+ Make a detectron2 model traceable with Caffe2 operators.
+ This class creates a traceable version of a detectron2 model which:
+
+ 1. Rewrite parts of the model using ops in Caffe2. Note that some ops do
+ not have GPU implementation in Caffe2.
+ 2. Remove post-processing and only produce raw layer outputs
+
+ After making a traceable model, the class provide methods to export such a
+ model to different deployment formats.
+ Exported graph produced by this class take two input tensors:
+
+ 1. (1, C, H, W) float "data" which is an image (usually in [0, 255]).
+ (H, W) often has to be padded to multiple of 32 (depend on the model
+ architecture).
+ 2. 1x3 float "im_info", each row of which is (height, width, 1.0).
+ Height and width are true image shapes before padding.
+
+ The class currently only supports models using builtin meta architectures.
+ Batch inference is not supported, and contributions are welcome.
+ """
+
+ def __init__(self, cfg: CfgNode, model: nn.Module, inputs):
+ """
+ Args:
+ cfg (CfgNode): a detectron2 config used to construct caffe2-compatible model.
+ model (nn.Module): An original pytorch model. Must be among a few official models
+ in detectron2 that can be converted to become caffe2-compatible automatically.
+ Weights have to be already loaded to this model.
+ inputs: sample inputs that the given model takes for inference.
+ Will be used to trace the model. For most models, random inputs with
+ no detected objects will not work as they lead to wrong traces.
+ """
+ assert isinstance(cfg, CfgNode), cfg
+ assert isinstance(model, torch.nn.Module), type(model)
+
+ # TODO make it support custom models, by passing in c2 model directly
+ C2MetaArch = META_ARCH_CAFFE2_EXPORT_TYPE_MAP[cfg.MODEL.META_ARCHITECTURE]
+ self.traceable_model = C2MetaArch(cfg, copy.deepcopy(model))
+ self.inputs = inputs
+ self.traceable_inputs = self.traceable_model.get_caffe2_inputs(inputs)
+
+ def export_caffe2(self):
+ """
+ Export the model to Caffe2's protobuf format.
+ The returned object can be saved with its :meth:`.save_protobuf()` method.
+ The result can be loaded and executed using Caffe2 runtime.
+
+ Returns:
+ :class:`Caffe2Model`
+ """
+ from .caffe2_export import export_caffe2_detection_model
+
+ predict_net, init_net = export_caffe2_detection_model(
+ self.traceable_model, self.traceable_inputs
+ )
+ return Caffe2Model(predict_net, init_net)
+
+ def export_onnx(self):
+ """
+ Export the model to ONNX format.
+ Note that the exported model contains custom ops only available in caffe2, therefore it
+ cannot be directly executed by other runtime (such as onnxruntime or TensorRT).
+ Post-processing or transformation passes may be applied on the model to accommodate
+ different runtimes, but we currently do not provide support for them.
+
+ Returns:
+ onnx.ModelProto: an onnx model.
+ """
+ from .caffe2_export import export_onnx_model as export_onnx_model_impl
+
+ return export_onnx_model_impl(self.traceable_model, (self.traceable_inputs,))
+
+ def export_torchscript(self):
+ """
+ Export the model to a ``torch.jit.TracedModule`` by tracing.
+ The returned object can be saved to a file by ``.save()``.
+
+ Returns:
+ torch.jit.TracedModule: a torch TracedModule
+ """
+ logger = logging.getLogger(__name__)
+ logger.info("Tracing the model with torch.jit.trace ...")
+ with torch.no_grad():
+ return torch.jit.trace(self.traceable_model, (self.traceable_inputs,))
+
+
+class Caffe2Model(nn.Module):
+ """
+ A wrapper around the traced model in Caffe2's protobuf format.
+ The exported graph has different inputs/outputs from the original Pytorch
+ model, as explained in :class:`Caffe2Tracer`. This class wraps around the
+ exported graph to simulate the same interface as the original Pytorch model.
+ It also provides functions to save/load models in Caffe2's format.'
+
+ Examples:
+ ::
+ c2_model = Caffe2Tracer(cfg, torch_model, inputs).export_caffe2()
+ inputs = [{"image": img_tensor_CHW}]
+ outputs = c2_model(inputs)
+ orig_outputs = torch_model(inputs)
+ """
+
+ def __init__(self, predict_net, init_net):
+ super().__init__()
+ self.eval() # always in eval mode
+ self._predict_net = predict_net
+ self._init_net = init_net
+ self._predictor = None
+
+ __init__.__HIDE_SPHINX_DOC__ = True
+
+ @property
+ def predict_net(self):
+ """
+ caffe2.core.Net: the underlying caffe2 predict net
+ """
+ return self._predict_net
+
+ @property
+ def init_net(self):
+ """
+ caffe2.core.Net: the underlying caffe2 init net
+ """
+ return self._init_net
+
+ def save_protobuf(self, output_dir):
+ """
+ Save the model as caffe2's protobuf format.
+ It saves the following files:
+
+ * "model.pb": definition of the graph. Can be visualized with
+ tools like `netron `_.
+ * "model_init.pb": model parameters
+ * "model.pbtxt": human-readable definition of the graph. Not
+ needed for deployment.
+
+ Args:
+ output_dir (str): the output directory to save protobuf files.
+ """
+ logger = logging.getLogger(__name__)
+ logger.info("Saving model to {} ...".format(output_dir))
+ if not PathManager.exists(output_dir):
+ PathManager.mkdirs(output_dir)
+
+ with PathManager.open(os.path.join(output_dir, "model.pb"), "wb") as f:
+ f.write(self._predict_net.SerializeToString())
+ with PathManager.open(os.path.join(output_dir, "model.pbtxt"), "w") as f:
+ f.write(str(self._predict_net))
+ with PathManager.open(os.path.join(output_dir, "model_init.pb"), "wb") as f:
+ f.write(self._init_net.SerializeToString())
+
+ def save_graph(self, output_file, inputs=None):
+ """
+ Save the graph as SVG format.
+
+ Args:
+ output_file (str): a SVG file
+ inputs: optional inputs given to the model.
+ If given, the inputs will be used to run the graph to record
+ shape of every tensor. The shape information will be
+ saved together with the graph.
+ """
+ from .caffe2_export import run_and_save_graph
+
+ if inputs is None:
+ save_graph(self._predict_net, output_file, op_only=False)
+ else:
+ size_divisibility = get_pb_arg_vali(self._predict_net, "size_divisibility", 0)
+ device = get_pb_arg_vals(self._predict_net, "device", b"cpu").decode("ascii")
+ inputs = convert_batched_inputs_to_c2_format(inputs, size_divisibility, device)
+ inputs = [x.cpu().numpy() for x in inputs]
+ run_and_save_graph(self._predict_net, self._init_net, inputs, output_file)
+
+ @staticmethod
+ def load_protobuf(dir):
+ """
+ Args:
+ dir (str): a directory used to save Caffe2Model with
+ :meth:`save_protobuf`.
+ The files "model.pb" and "model_init.pb" are needed.
+
+ Returns:
+ Caffe2Model: the caffe2 model loaded from this directory.
+ """
+ predict_net = caffe2_pb2.NetDef()
+ with PathManager.open(os.path.join(dir, "model.pb"), "rb") as f:
+ predict_net.ParseFromString(f.read())
+
+ init_net = caffe2_pb2.NetDef()
+ with PathManager.open(os.path.join(dir, "model_init.pb"), "rb") as f:
+ init_net.ParseFromString(f.read())
+
+ return Caffe2Model(predict_net, init_net)
+
+ def __call__(self, inputs):
+ """
+ An interface that wraps around a Caffe2 model and mimics detectron2's models'
+ input/output format. See details about the format at :doc:`/tutorials/models`.
+ This is used to compare the outputs of caffe2 model with its original torch model.
+
+ Due to the extra conversion between Pytorch/Caffe2, this method is not meant for
+ benchmark. Because of the conversion, this method also has dependency
+ on detectron2 in order to convert to detectron2's output format.
+ """
+ if self._predictor is None:
+ self._predictor = ProtobufDetectionModel(self._predict_net, self._init_net)
+ return self._predictor(inputs)
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/c10.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/c10.py
new file mode 100644
index 0000000000000000000000000000000000000000..fde3fb71189e6f1061e83b878bfdd16add7d8350
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/c10.py
@@ -0,0 +1,557 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+
+import math
+from typing import Dict
+import torch
+import torch.nn.functional as F
+
+from annotator.oneformer.detectron2.layers import ShapeSpec, cat
+from annotator.oneformer.detectron2.layers.roi_align_rotated import ROIAlignRotated
+from annotator.oneformer.detectron2.modeling import poolers
+from annotator.oneformer.detectron2.modeling.proposal_generator import rpn
+from annotator.oneformer.detectron2.modeling.roi_heads.mask_head import mask_rcnn_inference
+from annotator.oneformer.detectron2.structures import Boxes, ImageList, Instances, Keypoints, RotatedBoxes
+
+from .shared import alias, to_device
+
+
+"""
+This file contains caffe2-compatible implementation of several detectron2 components.
+"""
+
+
+class Caffe2Boxes(Boxes):
+ """
+ Representing a list of detectron2.structures.Boxes from minibatch, each box
+ is represented by a 5d vector (batch index + 4 coordinates), or a 6d vector
+ (batch index + 5 coordinates) for RotatedBoxes.
+ """
+
+ def __init__(self, tensor):
+ assert isinstance(tensor, torch.Tensor)
+ assert tensor.dim() == 2 and tensor.size(-1) in [4, 5, 6], tensor.size()
+ # TODO: make tensor immutable when dim is Nx5 for Boxes,
+ # and Nx6 for RotatedBoxes?
+ self.tensor = tensor
+
+
+# TODO clean up this class, maybe just extend Instances
+class InstancesList(object):
+ """
+ Tensor representation of a list of Instances object for a batch of images.
+
+ When dealing with a batch of images with Caffe2 ops, a list of bboxes
+ (instances) are usually represented by single Tensor with size
+ (sigma(Ni), 5) or (sigma(Ni), 4) plus a batch split Tensor. This class is
+ for providing common functions to convert between these two representations.
+ """
+
+ def __init__(self, im_info, indices, extra_fields=None):
+ # [N, 3] -> (H, W, Scale)
+ self.im_info = im_info
+ # [N,] -> indice of batch to which the instance belongs
+ self.indices = indices
+ # [N, ...]
+ self.batch_extra_fields = extra_fields or {}
+
+ self.image_size = self.im_info
+
+ def get_fields(self):
+ """like `get_fields` in the Instances object,
+ but return each field in tensor representations"""
+ ret = {}
+ for k, v in self.batch_extra_fields.items():
+ # if isinstance(v, torch.Tensor):
+ # tensor_rep = v
+ # elif isinstance(v, (Boxes, Keypoints)):
+ # tensor_rep = v.tensor
+ # else:
+ # raise ValueError("Can't find tensor representation for: {}".format())
+ ret[k] = v
+ return ret
+
+ def has(self, name):
+ return name in self.batch_extra_fields
+
+ def set(self, name, value):
+ # len(tensor) is a bad practice that generates ONNX constants during tracing.
+ # Although not a problem for the `assert` statement below, torch ONNX exporter
+ # still raises a misleading warning as it does not this call comes from `assert`
+ if isinstance(value, Boxes):
+ data_len = value.tensor.shape[0]
+ elif isinstance(value, torch.Tensor):
+ data_len = value.shape[0]
+ else:
+ data_len = len(value)
+ if len(self.batch_extra_fields):
+ assert (
+ len(self) == data_len
+ ), "Adding a field of length {} to a Instances of length {}".format(data_len, len(self))
+ self.batch_extra_fields[name] = value
+
+ def __getattr__(self, name):
+ if name not in self.batch_extra_fields:
+ raise AttributeError("Cannot find field '{}' in the given Instances!".format(name))
+ return self.batch_extra_fields[name]
+
+ def __len__(self):
+ return len(self.indices)
+
+ def flatten(self):
+ ret = []
+ for _, v in self.batch_extra_fields.items():
+ if isinstance(v, (Boxes, Keypoints)):
+ ret.append(v.tensor)
+ else:
+ ret.append(v)
+ return ret
+
+ @staticmethod
+ def to_d2_instances_list(instances_list):
+ """
+ Convert InstancesList to List[Instances]. The input `instances_list` can
+ also be a List[Instances], in this case this method is a non-op.
+ """
+ if not isinstance(instances_list, InstancesList):
+ assert all(isinstance(x, Instances) for x in instances_list)
+ return instances_list
+
+ ret = []
+ for i, info in enumerate(instances_list.im_info):
+ instances = Instances(torch.Size([int(info[0].item()), int(info[1].item())]))
+
+ ids = instances_list.indices == i
+ for k, v in instances_list.batch_extra_fields.items():
+ if isinstance(v, torch.Tensor):
+ instances.set(k, v[ids])
+ continue
+ elif isinstance(v, Boxes):
+ instances.set(k, v[ids, -4:])
+ continue
+
+ target_type, tensor_source = v
+ assert isinstance(tensor_source, torch.Tensor)
+ assert tensor_source.shape[0] == instances_list.indices.shape[0]
+ tensor_source = tensor_source[ids]
+
+ if issubclass(target_type, Boxes):
+ instances.set(k, Boxes(tensor_source[:, -4:]))
+ elif issubclass(target_type, Keypoints):
+ instances.set(k, Keypoints(tensor_source))
+ elif issubclass(target_type, torch.Tensor):
+ instances.set(k, tensor_source)
+ else:
+ raise ValueError("Can't handle targe type: {}".format(target_type))
+
+ ret.append(instances)
+ return ret
+
+
+class Caffe2Compatible(object):
+ """
+ A model can inherit this class to indicate that it can be traced and deployed with caffe2.
+ """
+
+ def _get_tensor_mode(self):
+ return self._tensor_mode
+
+ def _set_tensor_mode(self, v):
+ self._tensor_mode = v
+
+ tensor_mode = property(_get_tensor_mode, _set_tensor_mode)
+ """
+ If true, the model expects C2-style tensor only inputs/outputs format.
+ """
+
+
+class Caffe2RPN(Caffe2Compatible, rpn.RPN):
+ @classmethod
+ def from_config(cls, cfg, input_shape: Dict[str, ShapeSpec]):
+ ret = super(Caffe2Compatible, cls).from_config(cfg, input_shape)
+ assert tuple(cfg.MODEL.RPN.BBOX_REG_WEIGHTS) == (1.0, 1.0, 1.0, 1.0) or tuple(
+ cfg.MODEL.RPN.BBOX_REG_WEIGHTS
+ ) == (1.0, 1.0, 1.0, 1.0, 1.0)
+ return ret
+
+ def _generate_proposals(
+ self, images, objectness_logits_pred, anchor_deltas_pred, gt_instances=None
+ ):
+ assert isinstance(images, ImageList)
+ if self.tensor_mode:
+ im_info = images.image_sizes
+ else:
+ im_info = torch.tensor([[im_sz[0], im_sz[1], 1.0] for im_sz in images.image_sizes]).to(
+ images.tensor.device
+ )
+ assert isinstance(im_info, torch.Tensor)
+
+ rpn_rois_list = []
+ rpn_roi_probs_list = []
+ for scores, bbox_deltas, cell_anchors_tensor, feat_stride in zip(
+ objectness_logits_pred,
+ anchor_deltas_pred,
+ [b for (n, b) in self.anchor_generator.cell_anchors.named_buffers()],
+ self.anchor_generator.strides,
+ ):
+ scores = scores.detach()
+ bbox_deltas = bbox_deltas.detach()
+
+ rpn_rois, rpn_roi_probs = torch.ops._caffe2.GenerateProposals(
+ scores,
+ bbox_deltas,
+ im_info,
+ cell_anchors_tensor,
+ spatial_scale=1.0 / feat_stride,
+ pre_nms_topN=self.pre_nms_topk[self.training],
+ post_nms_topN=self.post_nms_topk[self.training],
+ nms_thresh=self.nms_thresh,
+ min_size=self.min_box_size,
+ # correct_transform_coords=True, # deprecated argument
+ angle_bound_on=True, # Default
+ angle_bound_lo=-180,
+ angle_bound_hi=180,
+ clip_angle_thresh=1.0, # Default
+ legacy_plus_one=False,
+ )
+ rpn_rois_list.append(rpn_rois)
+ rpn_roi_probs_list.append(rpn_roi_probs)
+
+ # For FPN in D2, in RPN all proposals from different levels are concated
+ # together, ranked and picked by top post_nms_topk. Then in ROIPooler
+ # it calculates level_assignments and calls the RoIAlign from
+ # the corresponding level.
+
+ if len(objectness_logits_pred) == 1:
+ rpn_rois = rpn_rois_list[0]
+ rpn_roi_probs = rpn_roi_probs_list[0]
+ else:
+ assert len(rpn_rois_list) == len(rpn_roi_probs_list)
+ rpn_post_nms_topN = self.post_nms_topk[self.training]
+
+ device = rpn_rois_list[0].device
+ input_list = [to_device(x, "cpu") for x in (rpn_rois_list + rpn_roi_probs_list)]
+
+ # TODO remove this after confirming rpn_max_level/rpn_min_level
+ # is not needed in CollectRpnProposals.
+ feature_strides = list(self.anchor_generator.strides)
+ rpn_min_level = int(math.log2(feature_strides[0]))
+ rpn_max_level = int(math.log2(feature_strides[-1]))
+ assert (rpn_max_level - rpn_min_level + 1) == len(
+ rpn_rois_list
+ ), "CollectRpnProposals requires continuous levels"
+
+ rpn_rois = torch.ops._caffe2.CollectRpnProposals(
+ input_list,
+ # NOTE: in current implementation, rpn_max_level and rpn_min_level
+ # are not needed, only the subtraction of two matters and it
+ # can be infer from the number of inputs. Keep them now for
+ # consistency.
+ rpn_max_level=2 + len(rpn_rois_list) - 1,
+ rpn_min_level=2,
+ rpn_post_nms_topN=rpn_post_nms_topN,
+ )
+ rpn_rois = to_device(rpn_rois, device)
+ rpn_roi_probs = []
+
+ proposals = self.c2_postprocess(im_info, rpn_rois, rpn_roi_probs, self.tensor_mode)
+ return proposals, {}
+
+ def forward(self, images, features, gt_instances=None):
+ assert not self.training
+ features = [features[f] for f in self.in_features]
+ objectness_logits_pred, anchor_deltas_pred = self.rpn_head(features)
+ return self._generate_proposals(
+ images,
+ objectness_logits_pred,
+ anchor_deltas_pred,
+ gt_instances,
+ )
+
+ @staticmethod
+ def c2_postprocess(im_info, rpn_rois, rpn_roi_probs, tensor_mode):
+ proposals = InstancesList(
+ im_info=im_info,
+ indices=rpn_rois[:, 0],
+ extra_fields={
+ "proposal_boxes": Caffe2Boxes(rpn_rois),
+ "objectness_logits": (torch.Tensor, rpn_roi_probs),
+ },
+ )
+ if not tensor_mode:
+ proposals = InstancesList.to_d2_instances_list(proposals)
+ else:
+ proposals = [proposals]
+ return proposals
+
+
+class Caffe2ROIPooler(Caffe2Compatible, poolers.ROIPooler):
+ @staticmethod
+ def c2_preprocess(box_lists):
+ assert all(isinstance(x, Boxes) for x in box_lists)
+ if all(isinstance(x, Caffe2Boxes) for x in box_lists):
+ # input is pure-tensor based
+ assert len(box_lists) == 1
+ pooler_fmt_boxes = box_lists[0].tensor
+ else:
+ pooler_fmt_boxes = poolers.convert_boxes_to_pooler_format(box_lists)
+ return pooler_fmt_boxes
+
+ def forward(self, x, box_lists):
+ assert not self.training
+
+ pooler_fmt_boxes = self.c2_preprocess(box_lists)
+ num_level_assignments = len(self.level_poolers)
+
+ if num_level_assignments == 1:
+ if isinstance(self.level_poolers[0], ROIAlignRotated):
+ c2_roi_align = torch.ops._caffe2.RoIAlignRotated
+ aligned = True
+ else:
+ c2_roi_align = torch.ops._caffe2.RoIAlign
+ aligned = self.level_poolers[0].aligned
+
+ x0 = x[0]
+ if x0.is_quantized:
+ x0 = x0.dequantize()
+
+ out = c2_roi_align(
+ x0,
+ pooler_fmt_boxes,
+ order="NCHW",
+ spatial_scale=float(self.level_poolers[0].spatial_scale),
+ pooled_h=int(self.output_size[0]),
+ pooled_w=int(self.output_size[1]),
+ sampling_ratio=int(self.level_poolers[0].sampling_ratio),
+ aligned=aligned,
+ )
+ return out
+
+ device = pooler_fmt_boxes.device
+ assert (
+ self.max_level - self.min_level + 1 == 4
+ ), "Currently DistributeFpnProposals only support 4 levels"
+ fpn_outputs = torch.ops._caffe2.DistributeFpnProposals(
+ to_device(pooler_fmt_boxes, "cpu"),
+ roi_canonical_scale=self.canonical_box_size,
+ roi_canonical_level=self.canonical_level,
+ roi_max_level=self.max_level,
+ roi_min_level=self.min_level,
+ legacy_plus_one=False,
+ )
+ fpn_outputs = [to_device(x, device) for x in fpn_outputs]
+
+ rois_fpn_list = fpn_outputs[:-1]
+ rois_idx_restore_int32 = fpn_outputs[-1]
+
+ roi_feat_fpn_list = []
+ for roi_fpn, x_level, pooler in zip(rois_fpn_list, x, self.level_poolers):
+ if isinstance(pooler, ROIAlignRotated):
+ c2_roi_align = torch.ops._caffe2.RoIAlignRotated
+ aligned = True
+ else:
+ c2_roi_align = torch.ops._caffe2.RoIAlign
+ aligned = bool(pooler.aligned)
+
+ if x_level.is_quantized:
+ x_level = x_level.dequantize()
+
+ roi_feat_fpn = c2_roi_align(
+ x_level,
+ roi_fpn,
+ order="NCHW",
+ spatial_scale=float(pooler.spatial_scale),
+ pooled_h=int(self.output_size[0]),
+ pooled_w=int(self.output_size[1]),
+ sampling_ratio=int(pooler.sampling_ratio),
+ aligned=aligned,
+ )
+ roi_feat_fpn_list.append(roi_feat_fpn)
+
+ roi_feat_shuffled = cat(roi_feat_fpn_list, dim=0)
+ assert roi_feat_shuffled.numel() > 0 and rois_idx_restore_int32.numel() > 0, (
+ "Caffe2 export requires tracing with a model checkpoint + input that can produce valid"
+ " detections. But no detections were obtained with the given checkpoint and input!"
+ )
+ roi_feat = torch.ops._caffe2.BatchPermutation(roi_feat_shuffled, rois_idx_restore_int32)
+ return roi_feat
+
+
+class Caffe2FastRCNNOutputsInference:
+ def __init__(self, tensor_mode):
+ self.tensor_mode = tensor_mode # whether the output is caffe2 tensor mode
+
+ def __call__(self, box_predictor, predictions, proposals):
+ """equivalent to FastRCNNOutputLayers.inference"""
+ num_classes = box_predictor.num_classes
+ score_thresh = box_predictor.test_score_thresh
+ nms_thresh = box_predictor.test_nms_thresh
+ topk_per_image = box_predictor.test_topk_per_image
+ is_rotated = len(box_predictor.box2box_transform.weights) == 5
+
+ if is_rotated:
+ box_dim = 5
+ assert box_predictor.box2box_transform.weights[4] == 1, (
+ "The weights for Rotated BBoxTransform in C2 have only 4 dimensions,"
+ + " thus enforcing the angle weight to be 1 for now"
+ )
+ box2box_transform_weights = box_predictor.box2box_transform.weights[:4]
+ else:
+ box_dim = 4
+ box2box_transform_weights = box_predictor.box2box_transform.weights
+
+ class_logits, box_regression = predictions
+ if num_classes + 1 == class_logits.shape[1]:
+ class_prob = F.softmax(class_logits, -1)
+ else:
+ assert num_classes == class_logits.shape[1]
+ class_prob = F.sigmoid(class_logits)
+ # BoxWithNMSLimit will infer num_classes from the shape of the class_prob
+ # So append a zero column as placeholder for the background class
+ class_prob = torch.cat((class_prob, torch.zeros(class_prob.shape[0], 1)), dim=1)
+
+ assert box_regression.shape[1] % box_dim == 0
+ cls_agnostic_bbox_reg = box_regression.shape[1] // box_dim == 1
+
+ input_tensor_mode = proposals[0].proposal_boxes.tensor.shape[1] == box_dim + 1
+
+ proposal_boxes = proposals[0].proposal_boxes
+ if isinstance(proposal_boxes, Caffe2Boxes):
+ rois = Caffe2Boxes.cat([p.proposal_boxes for p in proposals])
+ elif isinstance(proposal_boxes, RotatedBoxes):
+ rois = RotatedBoxes.cat([p.proposal_boxes for p in proposals])
+ elif isinstance(proposal_boxes, Boxes):
+ rois = Boxes.cat([p.proposal_boxes for p in proposals])
+ else:
+ raise NotImplementedError(
+ 'Expected proposals[0].proposal_boxes to be type "Boxes", '
+ f"instead got {type(proposal_boxes)}"
+ )
+
+ device, dtype = rois.tensor.device, rois.tensor.dtype
+ if input_tensor_mode:
+ im_info = proposals[0].image_size
+ rois = rois.tensor
+ else:
+ im_info = torch.tensor(
+ [[sz[0], sz[1], 1.0] for sz in [x.image_size for x in proposals]]
+ )
+ batch_ids = cat(
+ [
+ torch.full((b, 1), i, dtype=dtype, device=device)
+ for i, b in enumerate(len(p) for p in proposals)
+ ],
+ dim=0,
+ )
+ rois = torch.cat([batch_ids, rois.tensor], dim=1)
+
+ roi_pred_bbox, roi_batch_splits = torch.ops._caffe2.BBoxTransform(
+ to_device(rois, "cpu"),
+ to_device(box_regression, "cpu"),
+ to_device(im_info, "cpu"),
+ weights=box2box_transform_weights,
+ apply_scale=True,
+ rotated=is_rotated,
+ angle_bound_on=True,
+ angle_bound_lo=-180,
+ angle_bound_hi=180,
+ clip_angle_thresh=1.0,
+ legacy_plus_one=False,
+ )
+ roi_pred_bbox = to_device(roi_pred_bbox, device)
+ roi_batch_splits = to_device(roi_batch_splits, device)
+
+ nms_outputs = torch.ops._caffe2.BoxWithNMSLimit(
+ to_device(class_prob, "cpu"),
+ to_device(roi_pred_bbox, "cpu"),
+ to_device(roi_batch_splits, "cpu"),
+ score_thresh=float(score_thresh),
+ nms=float(nms_thresh),
+ detections_per_im=int(topk_per_image),
+ soft_nms_enabled=False,
+ soft_nms_method="linear",
+ soft_nms_sigma=0.5,
+ soft_nms_min_score_thres=0.001,
+ rotated=is_rotated,
+ cls_agnostic_bbox_reg=cls_agnostic_bbox_reg,
+ input_boxes_include_bg_cls=False,
+ output_classes_include_bg_cls=False,
+ legacy_plus_one=False,
+ )
+ roi_score_nms = to_device(nms_outputs[0], device)
+ roi_bbox_nms = to_device(nms_outputs[1], device)
+ roi_class_nms = to_device(nms_outputs[2], device)
+ roi_batch_splits_nms = to_device(nms_outputs[3], device)
+ roi_keeps_nms = to_device(nms_outputs[4], device)
+ roi_keeps_size_nms = to_device(nms_outputs[5], device)
+ if not self.tensor_mode:
+ roi_class_nms = roi_class_nms.to(torch.int64)
+
+ roi_batch_ids = cat(
+ [
+ torch.full((b, 1), i, dtype=dtype, device=device)
+ for i, b in enumerate(int(x.item()) for x in roi_batch_splits_nms)
+ ],
+ dim=0,
+ )
+
+ roi_class_nms = alias(roi_class_nms, "class_nms")
+ roi_score_nms = alias(roi_score_nms, "score_nms")
+ roi_bbox_nms = alias(roi_bbox_nms, "bbox_nms")
+ roi_batch_splits_nms = alias(roi_batch_splits_nms, "batch_splits_nms")
+ roi_keeps_nms = alias(roi_keeps_nms, "keeps_nms")
+ roi_keeps_size_nms = alias(roi_keeps_size_nms, "keeps_size_nms")
+
+ results = InstancesList(
+ im_info=im_info,
+ indices=roi_batch_ids[:, 0],
+ extra_fields={
+ "pred_boxes": Caffe2Boxes(roi_bbox_nms),
+ "scores": roi_score_nms,
+ "pred_classes": roi_class_nms,
+ },
+ )
+
+ if not self.tensor_mode:
+ results = InstancesList.to_d2_instances_list(results)
+ batch_splits = roi_batch_splits_nms.int().tolist()
+ kept_indices = list(roi_keeps_nms.to(torch.int64).split(batch_splits))
+ else:
+ results = [results]
+ kept_indices = [roi_keeps_nms]
+
+ return results, kept_indices
+
+
+class Caffe2MaskRCNNInference:
+ def __call__(self, pred_mask_logits, pred_instances):
+ """equivalent to mask_head.mask_rcnn_inference"""
+ if all(isinstance(x, InstancesList) for x in pred_instances):
+ assert len(pred_instances) == 1
+ mask_probs_pred = pred_mask_logits.sigmoid()
+ mask_probs_pred = alias(mask_probs_pred, "mask_fcn_probs")
+ pred_instances[0].set("pred_masks", mask_probs_pred)
+ else:
+ mask_rcnn_inference(pred_mask_logits, pred_instances)
+
+
+class Caffe2KeypointRCNNInference:
+ def __init__(self, use_heatmap_max_keypoint):
+ self.use_heatmap_max_keypoint = use_heatmap_max_keypoint
+
+ def __call__(self, pred_keypoint_logits, pred_instances):
+ # just return the keypoint heatmap for now,
+ # there will be option to call HeatmapMaxKeypointOp
+ output = alias(pred_keypoint_logits, "kps_score")
+ if all(isinstance(x, InstancesList) for x in pred_instances):
+ assert len(pred_instances) == 1
+ if self.use_heatmap_max_keypoint:
+ device = output.device
+ output = torch.ops._caffe2.HeatmapMaxKeypoint(
+ to_device(output, "cpu"),
+ pred_instances[0].pred_boxes.tensor,
+ should_output_softmax=True, # worth make it configerable?
+ )
+ output = to_device(output, device)
+ output = alias(output, "keypoints_out")
+ pred_instances[0].set("pred_keypoints", output)
+ return pred_keypoint_logits
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/caffe2_export.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/caffe2_export.py
new file mode 100644
index 0000000000000000000000000000000000000000..d609c27c7deb396352967dbcbc79b1e00f2a2de1
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/caffe2_export.py
@@ -0,0 +1,203 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+
+import copy
+import io
+import logging
+import numpy as np
+from typing import List
+import onnx
+import onnx.optimizer
+import torch
+from caffe2.proto import caffe2_pb2
+from caffe2.python import core
+from caffe2.python.onnx.backend import Caffe2Backend
+from tabulate import tabulate
+from termcolor import colored
+from torch.onnx import OperatorExportTypes
+
+from .shared import (
+ ScopedWS,
+ construct_init_net_from_params,
+ fuse_alias_placeholder,
+ fuse_copy_between_cpu_and_gpu,
+ get_params_from_init_net,
+ group_norm_replace_aten_with_caffe2,
+ infer_device_type,
+ remove_dead_end_ops,
+ remove_reshape_for_fc,
+ save_graph,
+)
+
+logger = logging.getLogger(__name__)
+
+
+def export_onnx_model(model, inputs):
+ """
+ Trace and export a model to onnx format.
+
+ Args:
+ model (nn.Module):
+ inputs (tuple[args]): the model will be called by `model(*inputs)`
+
+ Returns:
+ an onnx model
+ """
+ assert isinstance(model, torch.nn.Module)
+
+ # make sure all modules are in eval mode, onnx may change the training state
+ # of the module if the states are not consistent
+ def _check_eval(module):
+ assert not module.training
+
+ model.apply(_check_eval)
+
+ # Export the model to ONNX
+ with torch.no_grad():
+ with io.BytesIO() as f:
+ torch.onnx.export(
+ model,
+ inputs,
+ f,
+ operator_export_type=OperatorExportTypes.ONNX_ATEN_FALLBACK,
+ # verbose=True, # NOTE: uncomment this for debugging
+ # export_params=True,
+ )
+ onnx_model = onnx.load_from_string(f.getvalue())
+
+ return onnx_model
+
+
+def _op_stats(net_def):
+ type_count = {}
+ for t in [op.type for op in net_def.op]:
+ type_count[t] = type_count.get(t, 0) + 1
+ type_count_list = sorted(type_count.items(), key=lambda kv: kv[0]) # alphabet
+ type_count_list = sorted(type_count_list, key=lambda kv: -kv[1]) # count
+ return "\n".join("{:>4}x {}".format(count, name) for name, count in type_count_list)
+
+
+def _assign_device_option(
+ predict_net: caffe2_pb2.NetDef, init_net: caffe2_pb2.NetDef, tensor_inputs: List[torch.Tensor]
+):
+ """
+ ONNX exported network doesn't have concept of device, assign necessary
+ device option for each op in order to make it runable on GPU runtime.
+ """
+
+ def _get_device_type(torch_tensor):
+ assert torch_tensor.device.type in ["cpu", "cuda"]
+ assert torch_tensor.device.index == 0
+ return torch_tensor.device.type
+
+ def _assign_op_device_option(net_proto, net_ssa, blob_device_types):
+ for op, ssa_i in zip(net_proto.op, net_ssa):
+ if op.type in ["CopyCPUToGPU", "CopyGPUToCPU"]:
+ op.device_option.CopyFrom(core.DeviceOption(caffe2_pb2.CUDA, 0))
+ else:
+ devices = [blob_device_types[b] for b in ssa_i[0] + ssa_i[1]]
+ assert all(d == devices[0] for d in devices)
+ if devices[0] == "cuda":
+ op.device_option.CopyFrom(core.DeviceOption(caffe2_pb2.CUDA, 0))
+
+ # update ops in predict_net
+ predict_net_input_device_types = {
+ (name, 0): _get_device_type(tensor)
+ for name, tensor in zip(predict_net.external_input, tensor_inputs)
+ }
+ predict_net_device_types = infer_device_type(
+ predict_net, known_status=predict_net_input_device_types, device_name_style="pytorch"
+ )
+ predict_net_ssa, _ = core.get_ssa(predict_net)
+ _assign_op_device_option(predict_net, predict_net_ssa, predict_net_device_types)
+
+ # update ops in init_net
+ init_net_ssa, versions = core.get_ssa(init_net)
+ init_net_output_device_types = {
+ (name, versions[name]): predict_net_device_types[(name, 0)]
+ for name in init_net.external_output
+ }
+ init_net_device_types = infer_device_type(
+ init_net, known_status=init_net_output_device_types, device_name_style="pytorch"
+ )
+ _assign_op_device_option(init_net, init_net_ssa, init_net_device_types)
+
+
+def export_caffe2_detection_model(model: torch.nn.Module, tensor_inputs: List[torch.Tensor]):
+ """
+ Export a caffe2-compatible Detectron2 model to caffe2 format via ONNX.
+
+ Arg:
+ model: a caffe2-compatible version of detectron2 model, defined in caffe2_modeling.py
+ tensor_inputs: a list of tensors that caffe2 model takes as input.
+ """
+ model = copy.deepcopy(model)
+ assert isinstance(model, torch.nn.Module)
+ assert hasattr(model, "encode_additional_info")
+
+ # Export via ONNX
+ logger.info(
+ "Exporting a {} model via ONNX ...".format(type(model).__name__)
+ + " Some warnings from ONNX are expected and are usually not to worry about."
+ )
+ onnx_model = export_onnx_model(model, (tensor_inputs,))
+ # Convert ONNX model to Caffe2 protobuf
+ init_net, predict_net = Caffe2Backend.onnx_graph_to_caffe2_net(onnx_model)
+ ops_table = [[op.type, op.input, op.output] for op in predict_net.op]
+ table = tabulate(ops_table, headers=["type", "input", "output"], tablefmt="pipe")
+ logger.info(
+ "ONNX export Done. Exported predict_net (before optimizations):\n" + colored(table, "cyan")
+ )
+
+ # Apply protobuf optimization
+ fuse_alias_placeholder(predict_net, init_net)
+ if any(t.device.type != "cpu" for t in tensor_inputs):
+ fuse_copy_between_cpu_and_gpu(predict_net)
+ remove_dead_end_ops(init_net)
+ _assign_device_option(predict_net, init_net, tensor_inputs)
+ params, device_options = get_params_from_init_net(init_net)
+ predict_net, params = remove_reshape_for_fc(predict_net, params)
+ init_net = construct_init_net_from_params(params, device_options)
+ group_norm_replace_aten_with_caffe2(predict_net)
+
+ # Record necessary information for running the pb model in Detectron2 system.
+ model.encode_additional_info(predict_net, init_net)
+
+ logger.info("Operators used in predict_net: \n{}".format(_op_stats(predict_net)))
+ logger.info("Operators used in init_net: \n{}".format(_op_stats(init_net)))
+
+ return predict_net, init_net
+
+
+def run_and_save_graph(predict_net, init_net, tensor_inputs, graph_save_path):
+ """
+ Run the caffe2 model on given inputs, recording the shape and draw the graph.
+
+ predict_net/init_net: caffe2 model.
+ tensor_inputs: a list of tensors that caffe2 model takes as input.
+ graph_save_path: path for saving graph of exported model.
+ """
+
+ logger.info("Saving graph of ONNX exported model to {} ...".format(graph_save_path))
+ save_graph(predict_net, graph_save_path, op_only=False)
+
+ # Run the exported Caffe2 net
+ logger.info("Running ONNX exported model ...")
+ with ScopedWS("__ws_tmp__", True) as ws:
+ ws.RunNetOnce(init_net)
+ initialized_blobs = set(ws.Blobs())
+ uninitialized = [inp for inp in predict_net.external_input if inp not in initialized_blobs]
+ for name, blob in zip(uninitialized, tensor_inputs):
+ ws.FeedBlob(name, blob)
+
+ try:
+ ws.RunNetOnce(predict_net)
+ except RuntimeError as e:
+ logger.warning("Encountered RuntimeError: \n{}".format(str(e)))
+
+ ws_blobs = {b: ws.FetchBlob(b) for b in ws.Blobs()}
+ blob_sizes = {b: ws_blobs[b].shape for b in ws_blobs if isinstance(ws_blobs[b], np.ndarray)}
+
+ logger.info("Saving graph with blob shapes to {} ...".format(graph_save_path))
+ save_graph(predict_net, graph_save_path, op_only=False, blob_sizes=blob_sizes)
+
+ return ws_blobs
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/caffe2_inference.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/caffe2_inference.py
new file mode 100644
index 0000000000000000000000000000000000000000..deb886c0417285ed1d5ad85eb941fa1ac757cdab
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/caffe2_inference.py
@@ -0,0 +1,161 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+
+import logging
+import numpy as np
+from itertools import count
+import torch
+from caffe2.proto import caffe2_pb2
+from caffe2.python import core
+
+from .caffe2_modeling import META_ARCH_CAFFE2_EXPORT_TYPE_MAP, convert_batched_inputs_to_c2_format
+from .shared import ScopedWS, get_pb_arg_vali, get_pb_arg_vals, infer_device_type
+
+logger = logging.getLogger(__name__)
+
+
+# ===== ref: mobile-vision predictor's 'Caffe2Wrapper' class ======
+class ProtobufModel(torch.nn.Module):
+ """
+ Wrapper of a caffe2's protobuf model.
+ It works just like nn.Module, but running caffe2 under the hood.
+ Input/Output are tuple[tensor] that match the caffe2 net's external_input/output.
+ """
+
+ _ids = count(0)
+
+ def __init__(self, predict_net, init_net):
+ logger.info(f"Initializing ProtobufModel for: {predict_net.name} ...")
+ super().__init__()
+ assert isinstance(predict_net, caffe2_pb2.NetDef)
+ assert isinstance(init_net, caffe2_pb2.NetDef)
+ # create unique temporary workspace for each instance
+ self.ws_name = "__tmp_ProtobufModel_{}__".format(next(self._ids))
+ self.net = core.Net(predict_net)
+
+ logger.info("Running init_net once to fill the parameters ...")
+ with ScopedWS(self.ws_name, is_reset=True, is_cleanup=False) as ws:
+ ws.RunNetOnce(init_net)
+ uninitialized_external_input = []
+ for blob in self.net.Proto().external_input:
+ if blob not in ws.Blobs():
+ uninitialized_external_input.append(blob)
+ ws.CreateBlob(blob)
+ ws.CreateNet(self.net)
+
+ self._error_msgs = set()
+ self._input_blobs = uninitialized_external_input
+
+ def _infer_output_devices(self, inputs):
+ """
+ Returns:
+ list[str]: list of device for each external output
+ """
+
+ def _get_device_type(torch_tensor):
+ assert torch_tensor.device.type in ["cpu", "cuda"]
+ assert torch_tensor.device.index == 0
+ return torch_tensor.device.type
+
+ predict_net = self.net.Proto()
+ input_device_types = {
+ (name, 0): _get_device_type(tensor) for name, tensor in zip(self._input_blobs, inputs)
+ }
+ device_type_map = infer_device_type(
+ predict_net, known_status=input_device_types, device_name_style="pytorch"
+ )
+ ssa, versions = core.get_ssa(predict_net)
+ versioned_outputs = [(name, versions[name]) for name in predict_net.external_output]
+ output_devices = [device_type_map[outp] for outp in versioned_outputs]
+ return output_devices
+
+ def forward(self, inputs):
+ """
+ Args:
+ inputs (tuple[torch.Tensor])
+
+ Returns:
+ tuple[torch.Tensor]
+ """
+ assert len(inputs) == len(self._input_blobs), (
+ f"Length of inputs ({len(inputs)}) "
+ f"doesn't match the required input blobs: {self._input_blobs}"
+ )
+
+ with ScopedWS(self.ws_name, is_reset=False, is_cleanup=False) as ws:
+ for b, tensor in zip(self._input_blobs, inputs):
+ ws.FeedBlob(b, tensor)
+
+ try:
+ ws.RunNet(self.net.Proto().name)
+ except RuntimeError as e:
+ if not str(e) in self._error_msgs:
+ self._error_msgs.add(str(e))
+ logger.warning("Encountered new RuntimeError: \n{}".format(str(e)))
+ logger.warning("Catch the error and use partial results.")
+
+ c2_outputs = [ws.FetchBlob(b) for b in self.net.Proto().external_output]
+ # Remove outputs of current run, this is necessary in order to
+ # prevent fetching the result from previous run if the model fails
+ # in the middle.
+ for b in self.net.Proto().external_output:
+ # Needs to create uninitialized blob to make the net runable.
+ # This is "equivalent" to: ws.RemoveBlob(b) then ws.CreateBlob(b),
+ # but there'no such API.
+ ws.FeedBlob(b, f"{b}, a C++ native class of type nullptr (uninitialized).")
+
+ # Cast output to torch.Tensor on the desired device
+ output_devices = (
+ self._infer_output_devices(inputs)
+ if any(t.device.type != "cpu" for t in inputs)
+ else ["cpu" for _ in self.net.Proto().external_output]
+ )
+
+ outputs = []
+ for name, c2_output, device in zip(
+ self.net.Proto().external_output, c2_outputs, output_devices
+ ):
+ if not isinstance(c2_output, np.ndarray):
+ raise RuntimeError(
+ "Invalid output for blob {}, received: {}".format(name, c2_output)
+ )
+ outputs.append(torch.tensor(c2_output).to(device=device))
+ return tuple(outputs)
+
+
+class ProtobufDetectionModel(torch.nn.Module):
+ """
+ A class works just like a pytorch meta arch in terms of inference, but running
+ caffe2 model under the hood.
+ """
+
+ def __init__(self, predict_net, init_net, *, convert_outputs=None):
+ """
+ Args:
+ predict_net, init_net (core.Net): caffe2 nets
+ convert_outptus (callable): a function that converts caffe2
+ outputs to the same format of the original pytorch model.
+ By default, use the one defined in the caffe2 meta_arch.
+ """
+ super().__init__()
+ self.protobuf_model = ProtobufModel(predict_net, init_net)
+ self.size_divisibility = get_pb_arg_vali(predict_net, "size_divisibility", 0)
+ self.device = get_pb_arg_vals(predict_net, "device", b"cpu").decode("ascii")
+
+ if convert_outputs is None:
+ meta_arch = get_pb_arg_vals(predict_net, "meta_architecture", b"GeneralizedRCNN")
+ meta_arch = META_ARCH_CAFFE2_EXPORT_TYPE_MAP[meta_arch.decode("ascii")]
+ self._convert_outputs = meta_arch.get_outputs_converter(predict_net, init_net)
+ else:
+ self._convert_outputs = convert_outputs
+
+ def _convert_inputs(self, batched_inputs):
+ # currently all models convert inputs in the same way
+ return convert_batched_inputs_to_c2_format(
+ batched_inputs, self.size_divisibility, self.device
+ )
+
+ def forward(self, batched_inputs):
+ c2_inputs = self._convert_inputs(batched_inputs)
+ c2_results = self.protobuf_model(c2_inputs)
+ c2_results = dict(zip(self.protobuf_model.net.Proto().external_output, c2_results))
+ return self._convert_outputs(batched_inputs, c2_inputs, c2_results)
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/caffe2_modeling.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/caffe2_modeling.py
new file mode 100644
index 0000000000000000000000000000000000000000..e0128e4672bc08eb2983d3d382614c6381baefd9
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/caffe2_modeling.py
@@ -0,0 +1,419 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+
+import functools
+import io
+import struct
+import types
+import torch
+
+from annotator.oneformer.detectron2.modeling import meta_arch
+from annotator.oneformer.detectron2.modeling.box_regression import Box2BoxTransform
+from annotator.oneformer.detectron2.modeling.roi_heads import keypoint_head
+from annotator.oneformer.detectron2.structures import Boxes, ImageList, Instances, RotatedBoxes
+
+from .c10 import Caffe2Compatible
+from .caffe2_patch import ROIHeadsPatcher, patch_generalized_rcnn
+from .shared import (
+ alias,
+ check_set_pb_arg,
+ get_pb_arg_floats,
+ get_pb_arg_valf,
+ get_pb_arg_vali,
+ get_pb_arg_vals,
+ mock_torch_nn_functional_interpolate,
+)
+
+
+def assemble_rcnn_outputs_by_name(image_sizes, tensor_outputs, force_mask_on=False):
+ """
+ A function to assemble caffe2 model's outputs (i.e. Dict[str, Tensor])
+ to detectron2's format (i.e. list of Instances instance).
+ This only works when the model follows the Caffe2 detectron's naming convention.
+
+ Args:
+ image_sizes (List[List[int, int]]): [H, W] of every image.
+ tensor_outputs (Dict[str, Tensor]): external_output to its tensor.
+
+ force_mask_on (Bool): if true, the it make sure there'll be pred_masks even
+ if the mask is not found from tensor_outputs (usually due to model crash)
+ """
+
+ results = [Instances(image_size) for image_size in image_sizes]
+
+ batch_splits = tensor_outputs.get("batch_splits", None)
+ if batch_splits:
+ raise NotImplementedError()
+ assert len(image_sizes) == 1
+ result = results[0]
+
+ bbox_nms = tensor_outputs["bbox_nms"]
+ score_nms = tensor_outputs["score_nms"]
+ class_nms = tensor_outputs["class_nms"]
+ # Detection will always success because Conv support 0-batch
+ assert bbox_nms is not None
+ assert score_nms is not None
+ assert class_nms is not None
+ if bbox_nms.shape[1] == 5:
+ result.pred_boxes = RotatedBoxes(bbox_nms)
+ else:
+ result.pred_boxes = Boxes(bbox_nms)
+ result.scores = score_nms
+ result.pred_classes = class_nms.to(torch.int64)
+
+ mask_fcn_probs = tensor_outputs.get("mask_fcn_probs", None)
+ if mask_fcn_probs is not None:
+ # finish the mask pred
+ mask_probs_pred = mask_fcn_probs
+ num_masks = mask_probs_pred.shape[0]
+ class_pred = result.pred_classes
+ indices = torch.arange(num_masks, device=class_pred.device)
+ mask_probs_pred = mask_probs_pred[indices, class_pred][:, None]
+ result.pred_masks = mask_probs_pred
+ elif force_mask_on:
+ # NOTE: there's no way to know the height/width of mask here, it won't be
+ # used anyway when batch size is 0, so just set them to 0.
+ result.pred_masks = torch.zeros([0, 1, 0, 0], dtype=torch.uint8)
+
+ keypoints_out = tensor_outputs.get("keypoints_out", None)
+ kps_score = tensor_outputs.get("kps_score", None)
+ if keypoints_out is not None:
+ # keypoints_out: [N, 4, #kypoints], where 4 is in order of (x, y, score, prob)
+ keypoints_tensor = keypoints_out
+ # NOTE: it's possible that prob is not calculated if "should_output_softmax"
+ # is set to False in HeatmapMaxKeypoint, so just using raw score, seems
+ # it doesn't affect mAP. TODO: check more carefully.
+ keypoint_xyp = keypoints_tensor.transpose(1, 2)[:, :, [0, 1, 2]]
+ result.pred_keypoints = keypoint_xyp
+ elif kps_score is not None:
+ # keypoint heatmap to sparse data structure
+ pred_keypoint_logits = kps_score
+ keypoint_head.keypoint_rcnn_inference(pred_keypoint_logits, [result])
+
+ return results
+
+
+def _cast_to_f32(f64):
+ return struct.unpack("f", struct.pack("f", f64))[0]
+
+
+def set_caffe2_compatible_tensor_mode(model, enable=True):
+ def _fn(m):
+ if isinstance(m, Caffe2Compatible):
+ m.tensor_mode = enable
+
+ model.apply(_fn)
+
+
+def convert_batched_inputs_to_c2_format(batched_inputs, size_divisibility, device):
+ """
+ See get_caffe2_inputs() below.
+ """
+ assert all(isinstance(x, dict) for x in batched_inputs)
+ assert all(x["image"].dim() == 3 for x in batched_inputs)
+
+ images = [x["image"] for x in batched_inputs]
+ images = ImageList.from_tensors(images, size_divisibility)
+
+ im_info = []
+ for input_per_image, image_size in zip(batched_inputs, images.image_sizes):
+ target_height = input_per_image.get("height", image_size[0])
+ target_width = input_per_image.get("width", image_size[1]) # noqa
+ # NOTE: The scale inside im_info is kept as convention and for providing
+ # post-processing information if further processing is needed. For
+ # current Caffe2 model definitions that don't include post-processing inside
+ # the model, this number is not used.
+ # NOTE: There can be a slight difference between width and height
+ # scales, using a single number can results in numerical difference
+ # compared with D2's post-processing.
+ scale = target_height / image_size[0]
+ im_info.append([image_size[0], image_size[1], scale])
+ im_info = torch.Tensor(im_info)
+
+ return images.tensor.to(device), im_info.to(device)
+
+
+class Caffe2MetaArch(Caffe2Compatible, torch.nn.Module):
+ """
+ Base class for caffe2-compatible implementation of a meta architecture.
+ The forward is traceable and its traced graph can be converted to caffe2
+ graph through ONNX.
+ """
+
+ def __init__(self, cfg, torch_model):
+ """
+ Args:
+ cfg (CfgNode):
+ torch_model (nn.Module): the detectron2 model (meta_arch) to be
+ converted.
+ """
+ super().__init__()
+ self._wrapped_model = torch_model
+ self.eval()
+ set_caffe2_compatible_tensor_mode(self, True)
+
+ def get_caffe2_inputs(self, batched_inputs):
+ """
+ Convert pytorch-style structured inputs to caffe2-style inputs that
+ are tuples of tensors.
+
+ Args:
+ batched_inputs (list[dict]): inputs to a detectron2 model
+ in its standard format. Each dict has "image" (CHW tensor), and optionally
+ "height" and "width".
+
+ Returns:
+ tuple[Tensor]:
+ tuple of tensors that will be the inputs to the
+ :meth:`forward` method. For existing models, the first
+ is an NCHW tensor (padded and batched); the second is
+ a im_info Nx3 tensor, where the rows are
+ (height, width, unused legacy parameter)
+ """
+ return convert_batched_inputs_to_c2_format(
+ batched_inputs,
+ self._wrapped_model.backbone.size_divisibility,
+ self._wrapped_model.device,
+ )
+
+ def encode_additional_info(self, predict_net, init_net):
+ """
+ Save extra metadata that will be used by inference in the output protobuf.
+ """
+ pass
+
+ def forward(self, inputs):
+ """
+ Run the forward in caffe2-style. It has to use caffe2-compatible ops
+ and the method will be used for tracing.
+
+ Args:
+ inputs (tuple[Tensor]): inputs defined by :meth:`get_caffe2_input`.
+ They will be the inputs of the converted caffe2 graph.
+
+ Returns:
+ tuple[Tensor]: output tensors. They will be the outputs of the
+ converted caffe2 graph.
+ """
+ raise NotImplementedError
+
+ def _caffe2_preprocess_image(self, inputs):
+ """
+ Caffe2 implementation of preprocess_image, which is called inside each MetaArch's forward.
+ It normalizes the input images, and the final caffe2 graph assumes the
+ inputs have been batched already.
+ """
+ data, im_info = inputs
+ data = alias(data, "data")
+ im_info = alias(im_info, "im_info")
+ mean, std = self._wrapped_model.pixel_mean, self._wrapped_model.pixel_std
+ normalized_data = (data - mean) / std
+ normalized_data = alias(normalized_data, "normalized_data")
+
+ # Pack (data, im_info) into ImageList which is recognized by self.inference.
+ images = ImageList(tensor=normalized_data, image_sizes=im_info)
+ return images
+
+ @staticmethod
+ def get_outputs_converter(predict_net, init_net):
+ """
+ Creates a function that converts outputs of the caffe2 model to
+ detectron2's standard format.
+ The function uses information in `predict_net` and `init_net` that are
+ available at inferene time. Therefore the function logic can be used in inference.
+
+ The returned function has the following signature:
+
+ def convert(batched_inputs, c2_inputs, c2_results) -> detectron2_outputs
+
+ Where
+
+ * batched_inputs (list[dict]): the original input format of the meta arch
+ * c2_inputs (tuple[Tensor]): the caffe2 inputs.
+ * c2_results (dict[str, Tensor]): the caffe2 output format,
+ corresponding to the outputs of the :meth:`forward` function.
+ * detectron2_outputs: the original output format of the meta arch.
+
+ This function can be used to compare the outputs of the original meta arch and
+ the converted caffe2 graph.
+
+ Returns:
+ callable: a callable of the above signature.
+ """
+ raise NotImplementedError
+
+
+class Caffe2GeneralizedRCNN(Caffe2MetaArch):
+ def __init__(self, cfg, torch_model):
+ assert isinstance(torch_model, meta_arch.GeneralizedRCNN)
+ torch_model = patch_generalized_rcnn(torch_model)
+ super().__init__(cfg, torch_model)
+
+ try:
+ use_heatmap_max_keypoint = cfg.EXPORT_CAFFE2.USE_HEATMAP_MAX_KEYPOINT
+ except AttributeError:
+ use_heatmap_max_keypoint = False
+ self.roi_heads_patcher = ROIHeadsPatcher(
+ self._wrapped_model.roi_heads, use_heatmap_max_keypoint
+ )
+
+ def encode_additional_info(self, predict_net, init_net):
+ size_divisibility = self._wrapped_model.backbone.size_divisibility
+ check_set_pb_arg(predict_net, "size_divisibility", "i", size_divisibility)
+ check_set_pb_arg(
+ predict_net, "device", "s", str.encode(str(self._wrapped_model.device), "ascii")
+ )
+ check_set_pb_arg(predict_net, "meta_architecture", "s", b"GeneralizedRCNN")
+
+ @mock_torch_nn_functional_interpolate()
+ def forward(self, inputs):
+ if not self.tensor_mode:
+ return self._wrapped_model.inference(inputs)
+ images = self._caffe2_preprocess_image(inputs)
+ features = self._wrapped_model.backbone(images.tensor)
+ proposals, _ = self._wrapped_model.proposal_generator(images, features)
+ with self.roi_heads_patcher.mock_roi_heads():
+ detector_results, _ = self._wrapped_model.roi_heads(images, features, proposals)
+ return tuple(detector_results[0].flatten())
+
+ @staticmethod
+ def get_outputs_converter(predict_net, init_net):
+ def f(batched_inputs, c2_inputs, c2_results):
+ _, im_info = c2_inputs
+ image_sizes = [[int(im[0]), int(im[1])] for im in im_info]
+ results = assemble_rcnn_outputs_by_name(image_sizes, c2_results)
+ return meta_arch.GeneralizedRCNN._postprocess(results, batched_inputs, image_sizes)
+
+ return f
+
+
+class Caffe2RetinaNet(Caffe2MetaArch):
+ def __init__(self, cfg, torch_model):
+ assert isinstance(torch_model, meta_arch.RetinaNet)
+ super().__init__(cfg, torch_model)
+
+ @mock_torch_nn_functional_interpolate()
+ def forward(self, inputs):
+ assert self.tensor_mode
+ images = self._caffe2_preprocess_image(inputs)
+
+ # explicitly return the images sizes to avoid removing "im_info" by ONNX
+ # since it's not used in the forward path
+ return_tensors = [images.image_sizes]
+
+ features = self._wrapped_model.backbone(images.tensor)
+ features = [features[f] for f in self._wrapped_model.head_in_features]
+ for i, feature_i in enumerate(features):
+ features[i] = alias(feature_i, "feature_{}".format(i), is_backward=True)
+ return_tensors.append(features[i])
+
+ pred_logits, pred_anchor_deltas = self._wrapped_model.head(features)
+ for i, (box_cls_i, box_delta_i) in enumerate(zip(pred_logits, pred_anchor_deltas)):
+ return_tensors.append(alias(box_cls_i, "box_cls_{}".format(i)))
+ return_tensors.append(alias(box_delta_i, "box_delta_{}".format(i)))
+
+ return tuple(return_tensors)
+
+ def encode_additional_info(self, predict_net, init_net):
+ size_divisibility = self._wrapped_model.backbone.size_divisibility
+ check_set_pb_arg(predict_net, "size_divisibility", "i", size_divisibility)
+ check_set_pb_arg(
+ predict_net, "device", "s", str.encode(str(self._wrapped_model.device), "ascii")
+ )
+ check_set_pb_arg(predict_net, "meta_architecture", "s", b"RetinaNet")
+
+ # Inference parameters:
+ check_set_pb_arg(
+ predict_net, "score_threshold", "f", _cast_to_f32(self._wrapped_model.test_score_thresh)
+ )
+ check_set_pb_arg(
+ predict_net, "topk_candidates", "i", self._wrapped_model.test_topk_candidates
+ )
+ check_set_pb_arg(
+ predict_net, "nms_threshold", "f", _cast_to_f32(self._wrapped_model.test_nms_thresh)
+ )
+ check_set_pb_arg(
+ predict_net,
+ "max_detections_per_image",
+ "i",
+ self._wrapped_model.max_detections_per_image,
+ )
+
+ check_set_pb_arg(
+ predict_net,
+ "bbox_reg_weights",
+ "floats",
+ [_cast_to_f32(w) for w in self._wrapped_model.box2box_transform.weights],
+ )
+ self._encode_anchor_generator_cfg(predict_net)
+
+ def _encode_anchor_generator_cfg(self, predict_net):
+ # serialize anchor_generator for future use
+ serialized_anchor_generator = io.BytesIO()
+ torch.save(self._wrapped_model.anchor_generator, serialized_anchor_generator)
+ # Ideally we can put anchor generating inside the model, then we don't
+ # need to store this information.
+ bytes = serialized_anchor_generator.getvalue()
+ check_set_pb_arg(predict_net, "serialized_anchor_generator", "s", bytes)
+
+ @staticmethod
+ def get_outputs_converter(predict_net, init_net):
+ self = types.SimpleNamespace()
+ serialized_anchor_generator = io.BytesIO(
+ get_pb_arg_vals(predict_net, "serialized_anchor_generator", None)
+ )
+ self.anchor_generator = torch.load(serialized_anchor_generator)
+ bbox_reg_weights = get_pb_arg_floats(predict_net, "bbox_reg_weights", None)
+ self.box2box_transform = Box2BoxTransform(weights=tuple(bbox_reg_weights))
+ self.test_score_thresh = get_pb_arg_valf(predict_net, "score_threshold", None)
+ self.test_topk_candidates = get_pb_arg_vali(predict_net, "topk_candidates", None)
+ self.test_nms_thresh = get_pb_arg_valf(predict_net, "nms_threshold", None)
+ self.max_detections_per_image = get_pb_arg_vali(
+ predict_net, "max_detections_per_image", None
+ )
+
+ # hack to reuse inference code from RetinaNet
+ for meth in [
+ "forward_inference",
+ "inference_single_image",
+ "_transpose_dense_predictions",
+ "_decode_multi_level_predictions",
+ "_decode_per_level_predictions",
+ ]:
+ setattr(self, meth, functools.partial(getattr(meta_arch.RetinaNet, meth), self))
+
+ def f(batched_inputs, c2_inputs, c2_results):
+ _, im_info = c2_inputs
+ image_sizes = [[int(im[0]), int(im[1])] for im in im_info]
+ dummy_images = ImageList(
+ torch.randn(
+ (
+ len(im_info),
+ 3,
+ )
+ + tuple(image_sizes[0])
+ ),
+ image_sizes,
+ )
+
+ num_features = len([x for x in c2_results.keys() if x.startswith("box_cls_")])
+ pred_logits = [c2_results["box_cls_{}".format(i)] for i in range(num_features)]
+ pred_anchor_deltas = [c2_results["box_delta_{}".format(i)] for i in range(num_features)]
+
+ # For each feature level, feature should have the same batch size and
+ # spatial dimension as the box_cls and box_delta.
+ dummy_features = [x.clone()[:, 0:0, :, :] for x in pred_logits]
+ # self.num_classess can be inferred
+ self.num_classes = pred_logits[0].shape[1] // (pred_anchor_deltas[0].shape[1] // 4)
+
+ results = self.forward_inference(
+ dummy_images, dummy_features, [pred_logits, pred_anchor_deltas]
+ )
+ return meta_arch.GeneralizedRCNN._postprocess(results, batched_inputs, image_sizes)
+
+ return f
+
+
+META_ARCH_CAFFE2_EXPORT_TYPE_MAP = {
+ "GeneralizedRCNN": Caffe2GeneralizedRCNN,
+ "RetinaNet": Caffe2RetinaNet,
+}
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/caffe2_patch.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/caffe2_patch.py
new file mode 100644
index 0000000000000000000000000000000000000000..9c197cac1e7d5f665b6cbda46268716b1222f217
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/caffe2_patch.py
@@ -0,0 +1,152 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+
+import contextlib
+from unittest import mock
+import torch
+
+from annotator.oneformer.detectron2.modeling import poolers
+from annotator.oneformer.detectron2.modeling.proposal_generator import rpn
+from annotator.oneformer.detectron2.modeling.roi_heads import keypoint_head, mask_head
+from annotator.oneformer.detectron2.modeling.roi_heads.fast_rcnn import FastRCNNOutputLayers
+
+from .c10 import (
+ Caffe2Compatible,
+ Caffe2FastRCNNOutputsInference,
+ Caffe2KeypointRCNNInference,
+ Caffe2MaskRCNNInference,
+ Caffe2ROIPooler,
+ Caffe2RPN,
+)
+
+
+class GenericMixin(object):
+ pass
+
+
+class Caffe2CompatibleConverter(object):
+ """
+ A GenericUpdater which implements the `create_from` interface, by modifying
+ module object and assign it with another class replaceCls.
+ """
+
+ def __init__(self, replaceCls):
+ self.replaceCls = replaceCls
+
+ def create_from(self, module):
+ # update module's class to the new class
+ assert isinstance(module, torch.nn.Module)
+ if issubclass(self.replaceCls, GenericMixin):
+ # replaceCls should act as mixin, create a new class on-the-fly
+ new_class = type(
+ "{}MixedWith{}".format(self.replaceCls.__name__, module.__class__.__name__),
+ (self.replaceCls, module.__class__),
+ {}, # {"new_method": lambda self: ...},
+ )
+ module.__class__ = new_class
+ else:
+ # replaceCls is complete class, this allow arbitrary class swap
+ module.__class__ = self.replaceCls
+
+ # initialize Caffe2Compatible
+ if isinstance(module, Caffe2Compatible):
+ module.tensor_mode = False
+
+ return module
+
+
+def patch(model, target, updater, *args, **kwargs):
+ """
+ recursively (post-order) update all modules with the target type and its
+ subclasses, make a initialization/composition/inheritance/... via the
+ updater.create_from.
+ """
+ for name, module in model.named_children():
+ model._modules[name] = patch(module, target, updater, *args, **kwargs)
+ if isinstance(model, target):
+ return updater.create_from(model, *args, **kwargs)
+ return model
+
+
+def patch_generalized_rcnn(model):
+ ccc = Caffe2CompatibleConverter
+ model = patch(model, rpn.RPN, ccc(Caffe2RPN))
+ model = patch(model, poolers.ROIPooler, ccc(Caffe2ROIPooler))
+
+ return model
+
+
+@contextlib.contextmanager
+def mock_fastrcnn_outputs_inference(
+ tensor_mode, check=True, box_predictor_type=FastRCNNOutputLayers
+):
+ with mock.patch.object(
+ box_predictor_type,
+ "inference",
+ autospec=True,
+ side_effect=Caffe2FastRCNNOutputsInference(tensor_mode),
+ ) as mocked_func:
+ yield
+ if check:
+ assert mocked_func.call_count > 0
+
+
+@contextlib.contextmanager
+def mock_mask_rcnn_inference(tensor_mode, patched_module, check=True):
+ with mock.patch(
+ "{}.mask_rcnn_inference".format(patched_module), side_effect=Caffe2MaskRCNNInference()
+ ) as mocked_func:
+ yield
+ if check:
+ assert mocked_func.call_count > 0
+
+
+@contextlib.contextmanager
+def mock_keypoint_rcnn_inference(tensor_mode, patched_module, use_heatmap_max_keypoint, check=True):
+ with mock.patch(
+ "{}.keypoint_rcnn_inference".format(patched_module),
+ side_effect=Caffe2KeypointRCNNInference(use_heatmap_max_keypoint),
+ ) as mocked_func:
+ yield
+ if check:
+ assert mocked_func.call_count > 0
+
+
+class ROIHeadsPatcher:
+ def __init__(self, heads, use_heatmap_max_keypoint):
+ self.heads = heads
+ self.use_heatmap_max_keypoint = use_heatmap_max_keypoint
+
+ @contextlib.contextmanager
+ def mock_roi_heads(self, tensor_mode=True):
+ """
+ Patching several inference functions inside ROIHeads and its subclasses
+
+ Args:
+ tensor_mode (bool): whether the inputs/outputs are caffe2's tensor
+ format or not. Default to True.
+ """
+ # NOTE: this requries the `keypoint_rcnn_inference` and `mask_rcnn_inference`
+ # are called inside the same file as BaseXxxHead due to using mock.patch.
+ kpt_heads_mod = keypoint_head.BaseKeypointRCNNHead.__module__
+ mask_head_mod = mask_head.BaseMaskRCNNHead.__module__
+
+ mock_ctx_managers = [
+ mock_fastrcnn_outputs_inference(
+ tensor_mode=tensor_mode,
+ check=True,
+ box_predictor_type=type(self.heads.box_predictor),
+ )
+ ]
+ if getattr(self.heads, "keypoint_on", False):
+ mock_ctx_managers += [
+ mock_keypoint_rcnn_inference(
+ tensor_mode, kpt_heads_mod, self.use_heatmap_max_keypoint
+ )
+ ]
+ if getattr(self.heads, "mask_on", False):
+ mock_ctx_managers += [mock_mask_rcnn_inference(tensor_mode, mask_head_mod)]
+
+ with contextlib.ExitStack() as stack: # python 3.3+
+ for mgr in mock_ctx_managers:
+ stack.enter_context(mgr)
+ yield
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/flatten.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/flatten.py
new file mode 100644
index 0000000000000000000000000000000000000000..3fcb2bf49a0adad2798a10781a42accd9571218f
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/flatten.py
@@ -0,0 +1,330 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import collections
+from dataclasses import dataclass
+from typing import Callable, List, Optional, Tuple
+import torch
+from torch import nn
+
+from annotator.oneformer.detectron2.structures import Boxes, Instances, ROIMasks
+from annotator.oneformer.detectron2.utils.registry import _convert_target_to_string, locate
+
+from .torchscript_patch import patch_builtin_len
+
+
+@dataclass
+class Schema:
+ """
+ A Schema defines how to flatten a possibly hierarchical object into tuple of
+ primitive objects, so it can be used as inputs/outputs of PyTorch's tracing.
+
+ PyTorch does not support tracing a function that produces rich output
+ structures (e.g. dict, Instances, Boxes). To trace such a function, we
+ flatten the rich object into tuple of tensors, and return this tuple of tensors
+ instead. Meanwhile, we also need to know how to "rebuild" the original object
+ from the flattened results, so we can evaluate the flattened results.
+ A Schema defines how to flatten an object, and while flattening it, it records
+ necessary schemas so that the object can be rebuilt using the flattened outputs.
+
+ The flattened object and the schema object is returned by ``.flatten`` classmethod.
+ Then the original object can be rebuilt with the ``__call__`` method of schema.
+
+ A Schema is a dataclass that can be serialized easily.
+ """
+
+ # inspired by FetchMapper in tensorflow/python/client/session.py
+
+ @classmethod
+ def flatten(cls, obj):
+ raise NotImplementedError
+
+ def __call__(self, values):
+ raise NotImplementedError
+
+ @staticmethod
+ def _concat(values):
+ ret = ()
+ sizes = []
+ for v in values:
+ assert isinstance(v, tuple), "Flattened results must be a tuple"
+ ret = ret + v
+ sizes.append(len(v))
+ return ret, sizes
+
+ @staticmethod
+ def _split(values, sizes):
+ if len(sizes):
+ expected_len = sum(sizes)
+ assert (
+ len(values) == expected_len
+ ), f"Values has length {len(values)} but expect length {expected_len}."
+ ret = []
+ for k in range(len(sizes)):
+ begin, end = sum(sizes[:k]), sum(sizes[: k + 1])
+ ret.append(values[begin:end])
+ return ret
+
+
+@dataclass
+class ListSchema(Schema):
+ schemas: List[Schema] # the schemas that define how to flatten each element in the list
+ sizes: List[int] # the flattened length of each element
+
+ def __call__(self, values):
+ values = self._split(values, self.sizes)
+ if len(values) != len(self.schemas):
+ raise ValueError(
+ f"Values has length {len(values)} but schemas " f"has length {len(self.schemas)}!"
+ )
+ values = [m(v) for m, v in zip(self.schemas, values)]
+ return list(values)
+
+ @classmethod
+ def flatten(cls, obj):
+ res = [flatten_to_tuple(k) for k in obj]
+ values, sizes = cls._concat([k[0] for k in res])
+ return values, cls([k[1] for k in res], sizes)
+
+
+@dataclass
+class TupleSchema(ListSchema):
+ def __call__(self, values):
+ return tuple(super().__call__(values))
+
+
+@dataclass
+class IdentitySchema(Schema):
+ def __call__(self, values):
+ return values[0]
+
+ @classmethod
+ def flatten(cls, obj):
+ return (obj,), cls()
+
+
+@dataclass
+class DictSchema(ListSchema):
+ keys: List[str]
+
+ def __call__(self, values):
+ values = super().__call__(values)
+ return dict(zip(self.keys, values))
+
+ @classmethod
+ def flatten(cls, obj):
+ for k in obj.keys():
+ if not isinstance(k, str):
+ raise KeyError("Only support flattening dictionaries if keys are str.")
+ keys = sorted(obj.keys())
+ values = [obj[k] for k in keys]
+ ret, schema = ListSchema.flatten(values)
+ return ret, cls(schema.schemas, schema.sizes, keys)
+
+
+@dataclass
+class InstancesSchema(DictSchema):
+ def __call__(self, values):
+ image_size, fields = values[-1], values[:-1]
+ fields = super().__call__(fields)
+ return Instances(image_size, **fields)
+
+ @classmethod
+ def flatten(cls, obj):
+ ret, schema = super().flatten(obj.get_fields())
+ size = obj.image_size
+ if not isinstance(size, torch.Tensor):
+ size = torch.tensor(size)
+ return ret + (size,), schema
+
+
+@dataclass
+class TensorWrapSchema(Schema):
+ """
+ For classes that are simple wrapper of tensors, e.g.
+ Boxes, RotatedBoxes, BitMasks
+ """
+
+ class_name: str
+
+ def __call__(self, values):
+ return locate(self.class_name)(values[0])
+
+ @classmethod
+ def flatten(cls, obj):
+ return (obj.tensor,), cls(_convert_target_to_string(type(obj)))
+
+
+# if more custom structures needed in the future, can allow
+# passing in extra schemas for custom types
+def flatten_to_tuple(obj):
+ """
+ Flatten an object so it can be used for PyTorch tracing.
+ Also returns how to rebuild the original object from the flattened outputs.
+
+ Returns:
+ res (tuple): the flattened results that can be used as tracing outputs
+ schema: an object with a ``__call__`` method such that ``schema(res) == obj``.
+ It is a pure dataclass that can be serialized.
+ """
+ schemas = [
+ ((str, bytes), IdentitySchema),
+ (list, ListSchema),
+ (tuple, TupleSchema),
+ (collections.abc.Mapping, DictSchema),
+ (Instances, InstancesSchema),
+ ((Boxes, ROIMasks), TensorWrapSchema),
+ ]
+ for klass, schema in schemas:
+ if isinstance(obj, klass):
+ F = schema
+ break
+ else:
+ F = IdentitySchema
+
+ return F.flatten(obj)
+
+
+class TracingAdapter(nn.Module):
+ """
+ A model may take rich input/output format (e.g. dict or custom classes),
+ but `torch.jit.trace` requires tuple of tensors as input/output.
+ This adapter flattens input/output format of a model so it becomes traceable.
+
+ It also records the necessary schema to rebuild model's inputs/outputs from flattened
+ inputs/outputs.
+
+ Example:
+ ::
+ outputs = model(inputs) # inputs/outputs may be rich structure
+ adapter = TracingAdapter(model, inputs)
+
+ # can now trace the model, with adapter.flattened_inputs, or another
+ # tuple of tensors with the same length and meaning
+ traced = torch.jit.trace(adapter, adapter.flattened_inputs)
+
+ # traced model can only produce flattened outputs (tuple of tensors)
+ flattened_outputs = traced(*adapter.flattened_inputs)
+ # adapter knows the schema to convert it back (new_outputs == outputs)
+ new_outputs = adapter.outputs_schema(flattened_outputs)
+ """
+
+ flattened_inputs: Tuple[torch.Tensor] = None
+ """
+ Flattened version of inputs given to this class's constructor.
+ """
+
+ inputs_schema: Schema = None
+ """
+ Schema of the inputs given to this class's constructor.
+ """
+
+ outputs_schema: Schema = None
+ """
+ Schema of the output produced by calling the given model with inputs.
+ """
+
+ def __init__(
+ self,
+ model: nn.Module,
+ inputs,
+ inference_func: Optional[Callable] = None,
+ allow_non_tensor: bool = False,
+ ):
+ """
+ Args:
+ model: an nn.Module
+ inputs: An input argument or a tuple of input arguments used to call model.
+ After flattening, it has to only consist of tensors.
+ inference_func: a callable that takes (model, *inputs), calls the
+ model with inputs, and return outputs. By default it
+ is ``lambda model, *inputs: model(*inputs)``. Can be override
+ if you need to call the model differently.
+ allow_non_tensor: allow inputs/outputs to contain non-tensor objects.
+ This option will filter out non-tensor objects to make the
+ model traceable, but ``inputs_schema``/``outputs_schema`` cannot be
+ used anymore because inputs/outputs cannot be rebuilt from pure tensors.
+ This is useful when you're only interested in the single trace of
+ execution (e.g. for flop count), but not interested in
+ generalizing the traced graph to new inputs.
+ """
+ super().__init__()
+ if isinstance(model, (nn.parallel.distributed.DistributedDataParallel, nn.DataParallel)):
+ model = model.module
+ self.model = model
+ if not isinstance(inputs, tuple):
+ inputs = (inputs,)
+ self.inputs = inputs
+ self.allow_non_tensor = allow_non_tensor
+
+ if inference_func is None:
+ inference_func = lambda model, *inputs: model(*inputs) # noqa
+ self.inference_func = inference_func
+
+ self.flattened_inputs, self.inputs_schema = flatten_to_tuple(inputs)
+
+ if all(isinstance(x, torch.Tensor) for x in self.flattened_inputs):
+ return
+ if self.allow_non_tensor:
+ self.flattened_inputs = tuple(
+ [x for x in self.flattened_inputs if isinstance(x, torch.Tensor)]
+ )
+ self.inputs_schema = None
+ else:
+ for input in self.flattened_inputs:
+ if not isinstance(input, torch.Tensor):
+ raise ValueError(
+ "Inputs for tracing must only contain tensors. "
+ f"Got a {type(input)} instead."
+ )
+
+ def forward(self, *args: torch.Tensor):
+ with torch.no_grad(), patch_builtin_len():
+ if self.inputs_schema is not None:
+ inputs_orig_format = self.inputs_schema(args)
+ else:
+ if len(args) != len(self.flattened_inputs) or any(
+ x is not y for x, y in zip(args, self.flattened_inputs)
+ ):
+ raise ValueError(
+ "TracingAdapter does not contain valid inputs_schema."
+ " So it cannot generalize to other inputs and must be"
+ " traced with `.flattened_inputs`."
+ )
+ inputs_orig_format = self.inputs
+
+ outputs = self.inference_func(self.model, *inputs_orig_format)
+ flattened_outputs, schema = flatten_to_tuple(outputs)
+
+ flattened_output_tensors = tuple(
+ [x for x in flattened_outputs if isinstance(x, torch.Tensor)]
+ )
+ if len(flattened_output_tensors) < len(flattened_outputs):
+ if self.allow_non_tensor:
+ flattened_outputs = flattened_output_tensors
+ self.outputs_schema = None
+ else:
+ raise ValueError(
+ "Model cannot be traced because some model outputs "
+ "cannot flatten to tensors."
+ )
+ else: # schema is valid
+ if self.outputs_schema is None:
+ self.outputs_schema = schema
+ else:
+ assert self.outputs_schema == schema, (
+ "Model should always return outputs with the same "
+ "structure so it can be traced!"
+ )
+ return flattened_outputs
+
+ def _create_wrapper(self, traced_model):
+ """
+ Return a function that has an input/output interface the same as the
+ original model, but it calls the given traced model under the hood.
+ """
+
+ def forward(*args):
+ flattened_inputs, _ = flatten_to_tuple(args)
+ flattened_outputs = traced_model(*flattened_inputs)
+ return self.outputs_schema(flattened_outputs)
+
+ return forward
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/shared.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/shared.py
new file mode 100644
index 0000000000000000000000000000000000000000..53ba9335e26819f9381115eba17bbbe3816b469c
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/shared.py
@@ -0,0 +1,1039 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+
+import collections
+import copy
+import functools
+import logging
+import numpy as np
+import os
+from typing import Any, Callable, Dict, List, Optional, Tuple, Union
+from unittest import mock
+import caffe2.python.utils as putils
+import torch
+import torch.nn.functional as F
+from caffe2.proto import caffe2_pb2
+from caffe2.python import core, net_drawer, workspace
+from torch.nn.functional import interpolate as interp
+
+logger = logging.getLogger(__name__)
+
+
+# ==== torch/utils_toffee/cast.py =======================================
+
+
+def to_device(t, device_str):
+ """
+ This function is a replacement of .to(another_device) such that it allows the
+ casting to be traced properly by explicitly calling the underlying copy ops.
+ It also avoids introducing unncessary op when casting to the same device.
+ """
+ src = t.device
+ dst = torch.device(device_str)
+
+ if src == dst:
+ return t
+ elif src.type == "cuda" and dst.type == "cpu":
+ return torch.ops._caffe2.CopyGPUToCPU(t)
+ elif src.type == "cpu" and dst.type == "cuda":
+ return torch.ops._caffe2.CopyCPUToGPU(t)
+ else:
+ raise RuntimeError("Can't cast tensor from device {} to device {}".format(src, dst))
+
+
+# ==== torch/utils_toffee/interpolate.py =======================================
+
+
+# Note: borrowed from vision/detection/fair/detectron/detectron/modeling/detector.py
+def BilinearInterpolation(tensor_in, up_scale):
+ assert up_scale % 2 == 0, "Scale should be even"
+
+ def upsample_filt(size):
+ factor = (size + 1) // 2
+ if size % 2 == 1:
+ center = factor - 1
+ else:
+ center = factor - 0.5
+
+ og = np.ogrid[:size, :size]
+ return (1 - abs(og[0] - center) / factor) * (1 - abs(og[1] - center) / factor)
+
+ kernel_size = int(up_scale) * 2
+ bil_filt = upsample_filt(kernel_size)
+
+ dim = int(tensor_in.shape[1])
+ kernel = np.zeros((dim, dim, kernel_size, kernel_size), dtype=np.float32)
+ kernel[range(dim), range(dim), :, :] = bil_filt
+
+ tensor_out = F.conv_transpose2d(
+ tensor_in,
+ weight=to_device(torch.Tensor(kernel), tensor_in.device),
+ bias=None,
+ stride=int(up_scale),
+ padding=int(up_scale / 2),
+ )
+
+ return tensor_out
+
+
+# NOTE: ONNX is incompatible with traced torch.nn.functional.interpolate if
+# using dynamic `scale_factor` rather than static `size`. (T43166860)
+# NOTE: Caffe2 Int8 conversion might not be able to quantize `size` properly.
+def onnx_compatibale_interpolate(
+ input, size=None, scale_factor=None, mode="nearest", align_corners=None
+):
+ # NOTE: The input dimensions are interpreted in the form:
+ # `mini-batch x channels x [optional depth] x [optional height] x width`.
+ if size is None and scale_factor is not None:
+ if input.dim() == 4:
+ if isinstance(scale_factor, (int, float)):
+ height_scale, width_scale = (scale_factor, scale_factor)
+ else:
+ assert isinstance(scale_factor, (tuple, list))
+ assert len(scale_factor) == 2
+ height_scale, width_scale = scale_factor
+
+ assert not align_corners, "No matching C2 op for align_corners == True"
+ if mode == "nearest":
+ return torch.ops._caffe2.ResizeNearest(
+ input, order="NCHW", width_scale=width_scale, height_scale=height_scale
+ )
+ elif mode == "bilinear":
+ logger.warning(
+ "Use F.conv_transpose2d for bilinear interpolate"
+ " because there's no such C2 op, this may cause significant"
+ " slowdown and the boundary pixels won't be as same as"
+ " using F.interpolate due to padding."
+ )
+ assert height_scale == width_scale
+ return BilinearInterpolation(input, up_scale=height_scale)
+ logger.warning("Output size is not static, it might cause ONNX conversion issue")
+
+ return interp(input, size, scale_factor, mode, align_corners)
+
+
+def mock_torch_nn_functional_interpolate():
+ def decorator(func):
+ @functools.wraps(func)
+ def _mock_torch_nn_functional_interpolate(*args, **kwargs):
+ if torch.onnx.is_in_onnx_export():
+ with mock.patch(
+ "torch.nn.functional.interpolate", side_effect=onnx_compatibale_interpolate
+ ):
+ return func(*args, **kwargs)
+ else:
+ return func(*args, **kwargs)
+
+ return _mock_torch_nn_functional_interpolate
+
+ return decorator
+
+
+# ==== torch/utils_caffe2/ws_utils.py ==========================================
+
+
+class ScopedWS(object):
+ def __init__(self, ws_name, is_reset, is_cleanup=False):
+ self.ws_name = ws_name
+ self.is_reset = is_reset
+ self.is_cleanup = is_cleanup
+ self.org_ws = ""
+
+ def __enter__(self):
+ self.org_ws = workspace.CurrentWorkspace()
+ if self.ws_name is not None:
+ workspace.SwitchWorkspace(self.ws_name, True)
+ if self.is_reset:
+ workspace.ResetWorkspace()
+
+ return workspace
+
+ def __exit__(self, *args):
+ if self.is_cleanup:
+ workspace.ResetWorkspace()
+ if self.ws_name is not None:
+ workspace.SwitchWorkspace(self.org_ws)
+
+
+def fetch_any_blob(name):
+ bb = None
+ try:
+ bb = workspace.FetchBlob(name)
+ except TypeError:
+ bb = workspace.FetchInt8Blob(name)
+ except Exception as e:
+ logger.error("Get blob {} error: {}".format(name, e))
+
+ return bb
+
+
+# ==== torch/utils_caffe2/protobuf.py ==========================================
+
+
+def get_pb_arg(pb, arg_name):
+ for x in pb.arg:
+ if x.name == arg_name:
+ return x
+ return None
+
+
+def get_pb_arg_valf(pb, arg_name, default_val):
+ arg = get_pb_arg(pb, arg_name)
+ return arg.f if arg is not None else default_val
+
+
+def get_pb_arg_floats(pb, arg_name, default_val):
+ arg = get_pb_arg(pb, arg_name)
+ return list(map(float, arg.floats)) if arg is not None else default_val
+
+
+def get_pb_arg_ints(pb, arg_name, default_val):
+ arg = get_pb_arg(pb, arg_name)
+ return list(map(int, arg.ints)) if arg is not None else default_val
+
+
+def get_pb_arg_vali(pb, arg_name, default_val):
+ arg = get_pb_arg(pb, arg_name)
+ return arg.i if arg is not None else default_val
+
+
+def get_pb_arg_vals(pb, arg_name, default_val):
+ arg = get_pb_arg(pb, arg_name)
+ return arg.s if arg is not None else default_val
+
+
+def get_pb_arg_valstrings(pb, arg_name, default_val):
+ arg = get_pb_arg(pb, arg_name)
+ return list(arg.strings) if arg is not None else default_val
+
+
+def check_set_pb_arg(pb, arg_name, arg_attr, arg_value, allow_override=False):
+ arg = get_pb_arg(pb, arg_name)
+ if arg is None:
+ arg = putils.MakeArgument(arg_name, arg_value)
+ assert hasattr(arg, arg_attr)
+ pb.arg.extend([arg])
+ if allow_override and getattr(arg, arg_attr) != arg_value:
+ logger.warning(
+ "Override argument {}: {} -> {}".format(arg_name, getattr(arg, arg_attr), arg_value)
+ )
+ setattr(arg, arg_attr, arg_value)
+ else:
+ assert arg is not None
+ assert getattr(arg, arg_attr) == arg_value, "Existing value {}, new value {}".format(
+ getattr(arg, arg_attr), arg_value
+ )
+
+
+def _create_const_fill_op_from_numpy(name, tensor, device_option=None):
+ assert type(tensor) == np.ndarray
+ kTypeNameMapper = {
+ np.dtype("float32"): "GivenTensorFill",
+ np.dtype("int32"): "GivenTensorIntFill",
+ np.dtype("int64"): "GivenTensorInt64Fill",
+ np.dtype("uint8"): "GivenTensorStringFill",
+ }
+
+ args_dict = {}
+ if tensor.dtype == np.dtype("uint8"):
+ args_dict.update({"values": [str(tensor.data)], "shape": [1]})
+ else:
+ args_dict.update({"values": tensor, "shape": tensor.shape})
+
+ if device_option is not None:
+ args_dict["device_option"] = device_option
+
+ return core.CreateOperator(kTypeNameMapper[tensor.dtype], [], [name], **args_dict)
+
+
+def _create_const_fill_op_from_c2_int8_tensor(name, int8_tensor):
+ assert type(int8_tensor) == workspace.Int8Tensor
+ kTypeNameMapper = {
+ np.dtype("int32"): "Int8GivenIntTensorFill",
+ np.dtype("uint8"): "Int8GivenTensorFill",
+ }
+
+ tensor = int8_tensor.data
+ assert tensor.dtype in [np.dtype("uint8"), np.dtype("int32")]
+ values = tensor.tobytes() if tensor.dtype == np.dtype("uint8") else tensor
+
+ return core.CreateOperator(
+ kTypeNameMapper[tensor.dtype],
+ [],
+ [name],
+ values=values,
+ shape=tensor.shape,
+ Y_scale=int8_tensor.scale,
+ Y_zero_point=int8_tensor.zero_point,
+ )
+
+
+def create_const_fill_op(
+ name: str,
+ blob: Union[np.ndarray, workspace.Int8Tensor],
+ device_option: Optional[caffe2_pb2.DeviceOption] = None,
+) -> caffe2_pb2.OperatorDef:
+ """
+ Given a blob object, return the Caffe2 operator that creates this blob
+ as constant. Currently support NumPy tensor and Caffe2 Int8Tensor.
+ """
+
+ tensor_type = type(blob)
+ assert tensor_type in [
+ np.ndarray,
+ workspace.Int8Tensor,
+ ], 'Error when creating const fill op for "{}", unsupported blob type: {}'.format(
+ name, type(blob)
+ )
+
+ if tensor_type == np.ndarray:
+ return _create_const_fill_op_from_numpy(name, blob, device_option)
+ elif tensor_type == workspace.Int8Tensor:
+ assert device_option is None
+ return _create_const_fill_op_from_c2_int8_tensor(name, blob)
+
+
+def construct_init_net_from_params(
+ params: Dict[str, Any], device_options: Optional[Dict[str, caffe2_pb2.DeviceOption]] = None
+) -> caffe2_pb2.NetDef:
+ """
+ Construct the init_net from params dictionary
+ """
+ init_net = caffe2_pb2.NetDef()
+ device_options = device_options or {}
+ for name, blob in params.items():
+ if isinstance(blob, str):
+ logger.warning(
+ (
+ "Blob {} with type {} is not supported in generating init net,"
+ " skipped.".format(name, type(blob))
+ )
+ )
+ continue
+ init_net.op.extend(
+ [create_const_fill_op(name, blob, device_option=device_options.get(name, None))]
+ )
+ init_net.external_output.append(name)
+ return init_net
+
+
+def get_producer_map(ssa):
+ """
+ Return dict from versioned blob to (i, j),
+ where i is index of producer op, j is the index of output of that op.
+ """
+ producer_map = {}
+ for i in range(len(ssa)):
+ outputs = ssa[i][1]
+ for j, outp in enumerate(outputs):
+ producer_map[outp] = (i, j)
+ return producer_map
+
+
+def get_consumer_map(ssa):
+ """
+ Return dict from versioned blob to list of (i, j),
+ where i is index of consumer op, j is the index of input of that op.
+ """
+ consumer_map = collections.defaultdict(list)
+ for i in range(len(ssa)):
+ inputs = ssa[i][0]
+ for j, inp in enumerate(inputs):
+ consumer_map[inp].append((i, j))
+ return consumer_map
+
+
+def get_params_from_init_net(
+ init_net: caffe2_pb2.NetDef,
+) -> [Dict[str, Any], Dict[str, caffe2_pb2.DeviceOption]]:
+ """
+ Take the output blobs from init_net by running it.
+ Outputs:
+ params: dict from blob name to numpy array
+ device_options: dict from blob name to the device option of its creating op
+ """
+ # NOTE: this assumes that the params is determined by producer op with the
+ # only exception be CopyGPUToCPU which is CUDA op but returns CPU tensor.
+ def _get_device_option(producer_op):
+ if producer_op.type == "CopyGPUToCPU":
+ return caffe2_pb2.DeviceOption()
+ else:
+ return producer_op.device_option
+
+ with ScopedWS("__get_params_from_init_net__", is_reset=True, is_cleanup=True) as ws:
+ ws.RunNetOnce(init_net)
+ params = {b: fetch_any_blob(b) for b in init_net.external_output}
+ ssa, versions = core.get_ssa(init_net)
+ producer_map = get_producer_map(ssa)
+ device_options = {
+ b: _get_device_option(init_net.op[producer_map[(b, versions[b])][0]])
+ for b in init_net.external_output
+ }
+ return params, device_options
+
+
+def _updater_raise(op, input_types, output_types):
+ raise RuntimeError(
+ "Failed to apply updater for op {} given input_types {} and"
+ " output_types {}".format(op, input_types, output_types)
+ )
+
+
+def _generic_status_identifier(
+ predict_net: caffe2_pb2.NetDef,
+ status_updater: Callable,
+ known_status: Dict[Tuple[str, int], Any],
+) -> Dict[Tuple[str, int], Any]:
+ """
+ Statically infer the status of each blob, the status can be such as device type
+ (CPU/GPU), layout (NCHW/NHWC), data type (float32/int8), etc. "Blob" here
+ is versioned blob (Tuple[str, int]) in the format compatible with ssa.
+ Inputs:
+ predict_net: the caffe2 network
+ status_updater: a callable, given an op and the status of its input/output,
+ it returns the updated status of input/output. `None` is used for
+ representing unknown status.
+ known_status: a dict containing known status, used as initialization.
+ Outputs:
+ A dict mapping from versioned blob to its status
+ """
+ ssa, versions = core.get_ssa(predict_net)
+ versioned_ext_input = [(b, 0) for b in predict_net.external_input]
+ versioned_ext_output = [(b, versions[b]) for b in predict_net.external_output]
+ all_versioned_blobs = set().union(*[set(x[0] + x[1]) for x in ssa])
+
+ allowed_vbs = all_versioned_blobs.union(versioned_ext_input).union(versioned_ext_output)
+ assert all(k in allowed_vbs for k in known_status)
+ assert all(v is not None for v in known_status.values())
+ _known_status = copy.deepcopy(known_status)
+
+ def _check_and_update(key, value):
+ assert value is not None
+ if key in _known_status:
+ if not _known_status[key] == value:
+ raise RuntimeError(
+ "Confilict status for {}, existing status {}, new status {}".format(
+ key, _known_status[key], value
+ )
+ )
+ _known_status[key] = value
+
+ def _update_i(op, ssa_i):
+ versioned_inputs = ssa_i[0]
+ versioned_outputs = ssa_i[1]
+
+ inputs_status = [_known_status.get(b, None) for b in versioned_inputs]
+ outputs_status = [_known_status.get(b, None) for b in versioned_outputs]
+
+ new_inputs_status, new_outputs_status = status_updater(op, inputs_status, outputs_status)
+
+ for versioned_blob, status in zip(
+ versioned_inputs + versioned_outputs, new_inputs_status + new_outputs_status
+ ):
+ if status is not None:
+ _check_and_update(versioned_blob, status)
+
+ for op, ssa_i in zip(predict_net.op, ssa):
+ _update_i(op, ssa_i)
+ for op, ssa_i in zip(reversed(predict_net.op), reversed(ssa)):
+ _update_i(op, ssa_i)
+
+ # NOTE: This strictly checks all the blob from predict_net must be assgined
+ # a known status. However sometimes it's impossible (eg. having deadend op),
+ # we may relax this constraint if
+ for k in all_versioned_blobs:
+ if k not in _known_status:
+ raise NotImplementedError(
+ "Can not infer the status for {}. Currently only support the case where"
+ " a single forward and backward pass can identify status for all blobs.".format(k)
+ )
+
+ return _known_status
+
+
+def infer_device_type(
+ predict_net: caffe2_pb2.NetDef,
+ known_status: Dict[Tuple[str, int], Any],
+ device_name_style: str = "caffe2",
+) -> Dict[Tuple[str, int], str]:
+ """Return the device type ("cpu" or "gpu"/"cuda") of each (versioned) blob"""
+
+ assert device_name_style in ["caffe2", "pytorch"]
+ _CPU_STR = "cpu"
+ _GPU_STR = "gpu" if device_name_style == "caffe2" else "cuda"
+
+ def _copy_cpu_to_gpu_updater(op, input_types, output_types):
+ if input_types[0] == _GPU_STR or output_types[0] == _CPU_STR:
+ _updater_raise(op, input_types, output_types)
+ return ([_CPU_STR], [_GPU_STR])
+
+ def _copy_gpu_to_cpu_updater(op, input_types, output_types):
+ if input_types[0] == _CPU_STR or output_types[0] == _GPU_STR:
+ _updater_raise(op, input_types, output_types)
+ return ([_GPU_STR], [_CPU_STR])
+
+ def _other_ops_updater(op, input_types, output_types):
+ non_none_types = [x for x in input_types + output_types if x is not None]
+ if len(non_none_types) > 0:
+ the_type = non_none_types[0]
+ if not all(x == the_type for x in non_none_types):
+ _updater_raise(op, input_types, output_types)
+ else:
+ the_type = None
+ return ([the_type for _ in op.input], [the_type for _ in op.output])
+
+ def _device_updater(op, *args, **kwargs):
+ return {
+ "CopyCPUToGPU": _copy_cpu_to_gpu_updater,
+ "CopyGPUToCPU": _copy_gpu_to_cpu_updater,
+ }.get(op.type, _other_ops_updater)(op, *args, **kwargs)
+
+ return _generic_status_identifier(predict_net, _device_updater, known_status)
+
+
+# ==== torch/utils_caffe2/vis.py ===============================================
+
+
+def _modify_blob_names(ops, blob_rename_f):
+ ret = []
+
+ def _replace_list(blob_list, replaced_list):
+ del blob_list[:]
+ blob_list.extend(replaced_list)
+
+ for x in ops:
+ cur = copy.deepcopy(x)
+ _replace_list(cur.input, list(map(blob_rename_f, cur.input)))
+ _replace_list(cur.output, list(map(blob_rename_f, cur.output)))
+ ret.append(cur)
+
+ return ret
+
+
+def _rename_blob(name, blob_sizes, blob_ranges):
+ def _list_to_str(bsize):
+ ret = ", ".join([str(x) for x in bsize])
+ ret = "[" + ret + "]"
+ return ret
+
+ ret = name
+ if blob_sizes is not None and name in blob_sizes:
+ ret += "\n" + _list_to_str(blob_sizes[name])
+ if blob_ranges is not None and name in blob_ranges:
+ ret += "\n" + _list_to_str(blob_ranges[name])
+
+ return ret
+
+
+# graph_name could not contain word 'graph'
+def save_graph(net, file_name, graph_name="net", op_only=True, blob_sizes=None, blob_ranges=None):
+ blob_rename_f = functools.partial(_rename_blob, blob_sizes=blob_sizes, blob_ranges=blob_ranges)
+ return save_graph_base(net, file_name, graph_name, op_only, blob_rename_f)
+
+
+def save_graph_base(net, file_name, graph_name="net", op_only=True, blob_rename_func=None):
+ graph = None
+ ops = net.op
+ if blob_rename_func is not None:
+ ops = _modify_blob_names(ops, blob_rename_func)
+ if not op_only:
+ graph = net_drawer.GetPydotGraph(ops, graph_name, rankdir="TB")
+ else:
+ graph = net_drawer.GetPydotGraphMinimal(
+ ops, graph_name, rankdir="TB", minimal_dependency=True
+ )
+
+ try:
+ par_dir = os.path.dirname(file_name)
+ if not os.path.exists(par_dir):
+ os.makedirs(par_dir)
+
+ format = os.path.splitext(os.path.basename(file_name))[-1]
+ if format == ".png":
+ graph.write_png(file_name)
+ elif format == ".pdf":
+ graph.write_pdf(file_name)
+ elif format == ".svg":
+ graph.write_svg(file_name)
+ else:
+ print("Incorrect format {}".format(format))
+ except Exception as e:
+ print("Error when writing graph to image {}".format(e))
+
+ return graph
+
+
+# ==== torch/utils_toffee/aten_to_caffe2.py ====================================
+
+
+def group_norm_replace_aten_with_caffe2(predict_net: caffe2_pb2.NetDef):
+ """
+ For ONNX exported model, GroupNorm will be represented as ATen op,
+ this can be a drop in replacement from ATen to GroupNorm
+ """
+ count = 0
+ for op in predict_net.op:
+ if op.type == "ATen":
+ op_name = get_pb_arg_vals(op, "operator", None) # return byte in py3
+ if op_name and op_name.decode() == "group_norm":
+ op.arg.remove(get_pb_arg(op, "operator"))
+
+ if get_pb_arg_vali(op, "cudnn_enabled", None):
+ op.arg.remove(get_pb_arg(op, "cudnn_enabled"))
+
+ num_groups = get_pb_arg_vali(op, "num_groups", None)
+ if num_groups is not None:
+ op.arg.remove(get_pb_arg(op, "num_groups"))
+ check_set_pb_arg(op, "group", "i", num_groups)
+
+ op.type = "GroupNorm"
+ count += 1
+ if count > 1:
+ logger.info("Replaced {} ATen operator to GroupNormOp".format(count))
+
+
+# ==== torch/utils_toffee/alias.py =============================================
+
+
+def alias(x, name, is_backward=False):
+ if not torch.onnx.is_in_onnx_export():
+ return x
+ assert isinstance(x, torch.Tensor)
+ return torch.ops._caffe2.AliasWithName(x, name, is_backward=is_backward)
+
+
+def fuse_alias_placeholder(predict_net, init_net):
+ """Remove AliasWithName placeholder and rename the input/output of it"""
+ # First we finish all the re-naming
+ for i, op in enumerate(predict_net.op):
+ if op.type == "AliasWithName":
+ assert len(op.input) == 1
+ assert len(op.output) == 1
+ name = get_pb_arg_vals(op, "name", None).decode()
+ is_backward = bool(get_pb_arg_vali(op, "is_backward", 0))
+ rename_op_input(predict_net, init_net, i, 0, name, from_producer=is_backward)
+ rename_op_output(predict_net, i, 0, name)
+
+ # Remove AliasWithName, should be very safe since it's a non-op
+ new_ops = []
+ for op in predict_net.op:
+ if op.type != "AliasWithName":
+ new_ops.append(op)
+ else:
+ # safety check
+ assert op.input == op.output
+ assert op.input[0] == op.arg[0].s.decode()
+ del predict_net.op[:]
+ predict_net.op.extend(new_ops)
+
+
+# ==== torch/utils_caffe2/graph_transform.py ===================================
+
+
+class IllegalGraphTransformError(ValueError):
+ """When a graph transform function call can't be executed."""
+
+
+def _rename_versioned_blob_in_proto(
+ proto: caffe2_pb2.NetDef,
+ old_name: str,
+ new_name: str,
+ version: int,
+ ssa: List[Tuple[List[Tuple[str, int]], List[Tuple[str, int]]]],
+ start_versions: Dict[str, int],
+ end_versions: Dict[str, int],
+):
+ """In given proto, rename all blobs with matched version"""
+ # Operater list
+ for op, i_th_ssa in zip(proto.op, ssa):
+ versioned_inputs, versioned_outputs = i_th_ssa
+ for i in range(len(op.input)):
+ if versioned_inputs[i] == (old_name, version):
+ op.input[i] = new_name
+ for i in range(len(op.output)):
+ if versioned_outputs[i] == (old_name, version):
+ op.output[i] = new_name
+ # external_input
+ if start_versions.get(old_name, 0) == version:
+ for i in range(len(proto.external_input)):
+ if proto.external_input[i] == old_name:
+ proto.external_input[i] = new_name
+ # external_output
+ if end_versions.get(old_name, 0) == version:
+ for i in range(len(proto.external_output)):
+ if proto.external_output[i] == old_name:
+ proto.external_output[i] = new_name
+
+
+def rename_op_input(
+ predict_net: caffe2_pb2.NetDef,
+ init_net: caffe2_pb2.NetDef,
+ op_id: int,
+ input_id: int,
+ new_name: str,
+ from_producer: bool = False,
+):
+ """
+ Rename the op_id-th operator in predict_net, change it's input_id-th input's
+ name to the new_name. It also does automatic re-route and change
+ external_input and init_net if necessary.
+ - It requires the input is only consumed by this op.
+ - This function modifies predict_net and init_net in-place.
+ - When from_producer is enable, this also updates other operators that consumes
+ the same input. Be cautious because may trigger unintended behavior.
+ """
+ assert isinstance(predict_net, caffe2_pb2.NetDef)
+ assert isinstance(init_net, caffe2_pb2.NetDef)
+
+ init_net_ssa, init_net_versions = core.get_ssa(init_net)
+ predict_net_ssa, predict_net_versions = core.get_ssa(
+ predict_net, copy.deepcopy(init_net_versions)
+ )
+
+ versioned_inputs, versioned_outputs = predict_net_ssa[op_id]
+ old_name, version = versioned_inputs[input_id]
+
+ if from_producer:
+ producer_map = get_producer_map(predict_net_ssa)
+ if not (old_name, version) in producer_map:
+ raise NotImplementedError(
+ "Can't find producer, the input {} is probably from"
+ " init_net, this is not supported yet.".format(old_name)
+ )
+ producer = producer_map[(old_name, version)]
+ rename_op_output(predict_net, producer[0], producer[1], new_name)
+ return
+
+ def contain_targets(op_ssa):
+ return (old_name, version) in op_ssa[0]
+
+ is_consumer = [contain_targets(op_ssa) for op_ssa in predict_net_ssa]
+ if sum(is_consumer) > 1:
+ raise IllegalGraphTransformError(
+ (
+ "Input '{}' of operator(#{}) are consumed by other ops, please use"
+ + " rename_op_output on the producer instead. Offending op: \n{}"
+ ).format(old_name, op_id, predict_net.op[op_id])
+ )
+
+ # update init_net
+ _rename_versioned_blob_in_proto(
+ init_net, old_name, new_name, version, init_net_ssa, {}, init_net_versions
+ )
+ # update predict_net
+ _rename_versioned_blob_in_proto(
+ predict_net,
+ old_name,
+ new_name,
+ version,
+ predict_net_ssa,
+ init_net_versions,
+ predict_net_versions,
+ )
+
+
+def rename_op_output(predict_net: caffe2_pb2.NetDef, op_id: int, output_id: int, new_name: str):
+ """
+ Rename the op_id-th operator in predict_net, change it's output_id-th input's
+ name to the new_name. It also does automatic re-route and change
+ external_output and if necessary.
+ - It allows multiple consumers of its output.
+ - This function modifies predict_net in-place, doesn't need init_net.
+ """
+ assert isinstance(predict_net, caffe2_pb2.NetDef)
+
+ ssa, blob_versions = core.get_ssa(predict_net)
+
+ versioned_inputs, versioned_outputs = ssa[op_id]
+ old_name, version = versioned_outputs[output_id]
+
+ # update predict_net
+ _rename_versioned_blob_in_proto(
+ predict_net, old_name, new_name, version, ssa, {}, blob_versions
+ )
+
+
+def get_sub_graph_external_input_output(
+ predict_net: caffe2_pb2.NetDef, sub_graph_op_indices: List[int]
+) -> Tuple[List[Tuple[str, int]], List[Tuple[str, int]]]:
+ """
+ Return the list of external input/output of sub-graph,
+ each element is tuple of the name and corresponding version in predict_net.
+
+ external input/output is defined the same way as caffe2 NetDef.
+ """
+ ssa, versions = core.get_ssa(predict_net)
+
+ all_inputs = []
+ all_outputs = []
+ for op_id in sub_graph_op_indices:
+ all_inputs += [inp for inp in ssa[op_id][0] if inp not in all_inputs]
+ all_outputs += list(ssa[op_id][1]) # ssa output won't repeat
+
+ # for versioned blobs, external inputs are just those blob in all_inputs
+ # but not in all_outputs
+ ext_inputs = [inp for inp in all_inputs if inp not in all_outputs]
+
+ # external outputs are essentially outputs of this subgraph that are used
+ # outside of this sub-graph (including predict_net.external_output)
+ all_other_inputs = sum(
+ (ssa[i][0] for i in range(len(ssa)) if i not in sub_graph_op_indices),
+ [(outp, versions[outp]) for outp in predict_net.external_output],
+ )
+ ext_outputs = [outp for outp in all_outputs if outp in set(all_other_inputs)]
+
+ return ext_inputs, ext_outputs
+
+
+class DiGraph:
+ """A DAG representation of caffe2 graph, each vertice is a versioned blob."""
+
+ def __init__(self):
+ self.vertices = set()
+ self.graph = collections.defaultdict(list)
+
+ def add_edge(self, u, v):
+ self.graph[u].append(v)
+ self.vertices.add(u)
+ self.vertices.add(v)
+
+ # grab from https://www.geeksforgeeks.org/find-paths-given-source-destination/
+ def get_all_paths(self, s, d):
+ visited = {k: False for k in self.vertices}
+ path = []
+ all_paths = []
+
+ def _get_all_paths_util(graph, u, d, visited, path):
+ visited[u] = True
+ path.append(u)
+ if u == d:
+ all_paths.append(copy.deepcopy(path))
+ else:
+ for i in graph[u]:
+ if not visited[i]:
+ _get_all_paths_util(graph, i, d, visited, path)
+ path.pop()
+ visited[u] = False
+
+ _get_all_paths_util(self.graph, s, d, visited, path)
+ return all_paths
+
+ @staticmethod
+ def from_ssa(ssa):
+ graph = DiGraph()
+ for op_id in range(len(ssa)):
+ for inp in ssa[op_id][0]:
+ for outp in ssa[op_id][1]:
+ graph.add_edge(inp, outp)
+ return graph
+
+
+def _get_dependency_chain(ssa, versioned_target, versioned_source):
+ """
+ Return the index list of relevant operator to produce target blob from source blob,
+ if there's no dependency, return empty list.
+ """
+
+ # finding all paths between nodes can be O(N!), thus we can only search
+ # in the subgraph using the op starting from the first consumer of source blob
+ # to the producer of the target blob.
+ consumer_map = get_consumer_map(ssa)
+ producer_map = get_producer_map(ssa)
+ start_op = min(x[0] for x in consumer_map[versioned_source]) - 15
+ end_op = (
+ producer_map[versioned_target][0] + 15 if versioned_target in producer_map else start_op
+ )
+ sub_graph_ssa = ssa[start_op : end_op + 1]
+ if len(sub_graph_ssa) > 30:
+ logger.warning(
+ "Subgraph bebetween {} and {} is large (from op#{} to op#{}), it"
+ " might take non-trival time to find all paths between them.".format(
+ versioned_source, versioned_target, start_op, end_op
+ )
+ )
+
+ dag = DiGraph.from_ssa(sub_graph_ssa)
+ paths = dag.get_all_paths(versioned_source, versioned_target) # include two ends
+ ops_in_paths = [[producer_map[blob][0] for blob in path[1:]] for path in paths]
+ return sorted(set().union(*[set(ops) for ops in ops_in_paths]))
+
+
+def identify_reshape_sub_graph(predict_net: caffe2_pb2.NetDef) -> List[List[int]]:
+ """
+ Idenfity the reshape sub-graph in a protobuf.
+ The reshape sub-graph is defined as matching the following pattern:
+
+ (input_blob) -> Op_1 -> ... -> Op_N -> (new_shape) -─┐
+ └-------------------------------------------> Reshape -> (output_blob)
+
+ Return:
+ List of sub-graphs, each sub-graph is represented as a list of indices
+ of the relavent ops, [Op_1, Op_2, ..., Op_N, Reshape]
+ """
+
+ ssa, _ = core.get_ssa(predict_net)
+
+ ret = []
+ for i, op in enumerate(predict_net.op):
+ if op.type == "Reshape":
+ assert len(op.input) == 2
+ input_ssa = ssa[i][0]
+ data_source = input_ssa[0]
+ shape_source = input_ssa[1]
+ op_indices = _get_dependency_chain(ssa, shape_source, data_source)
+ ret.append(op_indices + [i])
+ return ret
+
+
+def remove_reshape_for_fc(predict_net, params):
+ """
+ In PyTorch nn.Linear has to take 2D tensor, this often leads to reshape
+ a 4D tensor to 2D by calling .view(). However this (dynamic) reshaping
+ doesn't work well with ONNX and Int8 tools, and cause using extra
+ ops (eg. ExpandDims) that might not be available on mobile.
+ Luckily Caffe2 supports 4D tensor for FC, so we can remove those reshape
+ after exporting ONNX model.
+ """
+ from caffe2.python import core
+
+ # find all reshape sub-graph that can be removed, which is now all Reshape
+ # sub-graph whose output is only consumed by FC.
+ # TODO: to make it safer, we may need the actually value to better determine
+ # if a Reshape before FC is removable.
+ reshape_sub_graphs = identify_reshape_sub_graph(predict_net)
+ sub_graphs_to_remove = []
+ for reshape_sub_graph in reshape_sub_graphs:
+ reshape_op_id = reshape_sub_graph[-1]
+ assert predict_net.op[reshape_op_id].type == "Reshape"
+ ssa, _ = core.get_ssa(predict_net)
+ reshape_output = ssa[reshape_op_id][1][0]
+ consumers = [i for i in range(len(ssa)) if reshape_output in ssa[i][0]]
+ if all(predict_net.op[consumer].type == "FC" for consumer in consumers):
+ # safety check if the sub-graph is isolated, for this reshape sub-graph,
+ # it means it has one non-param external input and one external output.
+ ext_inputs, ext_outputs = get_sub_graph_external_input_output(
+ predict_net, reshape_sub_graph
+ )
+ non_params_ext_inputs = [inp for inp in ext_inputs if inp[1] != 0]
+ if len(non_params_ext_inputs) == 1 and len(ext_outputs) == 1:
+ sub_graphs_to_remove.append(reshape_sub_graph)
+
+ # perform removing subgraph by:
+ # 1: rename the Reshape's output to its input, then the graph can be
+ # seen as in-place itentify, meaning whose external input/output are the same.
+ # 2: simply remove those ops.
+ remove_op_ids = []
+ params_to_remove = []
+ for sub_graph in sub_graphs_to_remove:
+ logger.info(
+ "Remove Reshape sub-graph:\n{}".format(
+ "".join(["(#{:>4})\n{}".format(i, predict_net.op[i]) for i in sub_graph])
+ )
+ )
+ reshape_op_id = sub_graph[-1]
+ new_reshap_output = predict_net.op[reshape_op_id].input[0]
+ rename_op_output(predict_net, reshape_op_id, 0, new_reshap_output)
+ ext_inputs, ext_outputs = get_sub_graph_external_input_output(predict_net, sub_graph)
+ non_params_ext_inputs = [inp for inp in ext_inputs if inp[1] != 0]
+ params_ext_inputs = [inp for inp in ext_inputs if inp[1] == 0]
+ assert len(non_params_ext_inputs) == 1 and len(ext_outputs) == 1
+ assert ext_outputs[0][0] == non_params_ext_inputs[0][0]
+ assert ext_outputs[0][1] == non_params_ext_inputs[0][1] + 1
+ remove_op_ids.extend(sub_graph)
+ params_to_remove.extend(params_ext_inputs)
+
+ predict_net = copy.deepcopy(predict_net)
+ new_ops = [op for i, op in enumerate(predict_net.op) if i not in remove_op_ids]
+ del predict_net.op[:]
+ predict_net.op.extend(new_ops)
+ for versioned_params in params_to_remove:
+ name = versioned_params[0]
+ logger.info("Remove params: {} from init_net and predict_net.external_input".format(name))
+ del params[name]
+ predict_net.external_input.remove(name)
+
+ return predict_net, params
+
+
+def fuse_copy_between_cpu_and_gpu(predict_net: caffe2_pb2.NetDef):
+ """
+ In-place fuse extra copy ops between cpu/gpu for the following case:
+ a -CopyAToB-> b -CopyBToA> c1 -NextOp1-> d1
+ -CopyBToA> c2 -NextOp2-> d2
+ The fused network will look like:
+ a -NextOp1-> d1
+ -NextOp2-> d2
+ """
+
+ _COPY_OPS = ["CopyCPUToGPU", "CopyGPUToCPU"]
+
+ def _fuse_once(predict_net):
+ ssa, blob_versions = core.get_ssa(predict_net)
+ consumer_map = get_consumer_map(ssa)
+ versioned_external_output = [
+ (name, blob_versions[name]) for name in predict_net.external_output
+ ]
+
+ for op_id, op in enumerate(predict_net.op):
+ if op.type in _COPY_OPS:
+ fw_copy_versioned_output = ssa[op_id][1][0]
+ consumer_ids = [x[0] for x in consumer_map[fw_copy_versioned_output]]
+ reverse_op_type = _COPY_OPS[1 - _COPY_OPS.index(op.type)]
+
+ is_fusable = (
+ len(consumer_ids) > 0
+ and fw_copy_versioned_output not in versioned_external_output
+ and all(
+ predict_net.op[_op_id].type == reverse_op_type
+ and ssa[_op_id][1][0] not in versioned_external_output
+ for _op_id in consumer_ids
+ )
+ )
+
+ if is_fusable:
+ for rv_copy_op_id in consumer_ids:
+ # making each NextOp uses "a" directly and removing Copy ops
+ rs_copy_versioned_output = ssa[rv_copy_op_id][1][0]
+ next_op_id, inp_id = consumer_map[rs_copy_versioned_output][0]
+ predict_net.op[next_op_id].input[inp_id] = op.input[0]
+ # remove CopyOps
+ new_ops = [
+ op
+ for i, op in enumerate(predict_net.op)
+ if i != op_id and i not in consumer_ids
+ ]
+ del predict_net.op[:]
+ predict_net.op.extend(new_ops)
+ return True
+
+ return False
+
+ # _fuse_once returns False is nothing can be fused
+ while _fuse_once(predict_net):
+ pass
+
+
+def remove_dead_end_ops(net_def: caffe2_pb2.NetDef):
+ """remove ops if its output is not used or not in external_output"""
+ ssa, versions = core.get_ssa(net_def)
+ versioned_external_output = [(name, versions[name]) for name in net_def.external_output]
+ consumer_map = get_consumer_map(ssa)
+ removed_op_ids = set()
+
+ def _is_dead_end(versioned_blob):
+ return not (
+ versioned_blob in versioned_external_output
+ or (
+ len(consumer_map[versioned_blob]) > 0
+ and all(x[0] not in removed_op_ids for x in consumer_map[versioned_blob])
+ )
+ )
+
+ for i, ssa_i in reversed(list(enumerate(ssa))):
+ versioned_outputs = ssa_i[1]
+ if all(_is_dead_end(outp) for outp in versioned_outputs):
+ removed_op_ids.add(i)
+
+ # simply removing those deadend ops should have no effect to external_output
+ new_ops = [op for i, op in enumerate(net_def.op) if i not in removed_op_ids]
+ del net_def.op[:]
+ net_def.op.extend(new_ops)
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/torchscript.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/torchscript.py
new file mode 100644
index 0000000000000000000000000000000000000000..8ce1c81e1b7abb65415055ae0d1d4b83e1ae111d
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/torchscript.py
@@ -0,0 +1,132 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+
+import os
+import torch
+
+from annotator.oneformer.detectron2.utils.file_io import PathManager
+
+from .torchscript_patch import freeze_training_mode, patch_instances
+
+__all__ = ["scripting_with_instances", "dump_torchscript_IR"]
+
+
+def scripting_with_instances(model, fields):
+ """
+ Run :func:`torch.jit.script` on a model that uses the :class:`Instances` class. Since
+ attributes of :class:`Instances` are "dynamically" added in eager mode，it is difficult
+ for scripting to support it out of the box. This function is made to support scripting
+ a model that uses :class:`Instances`. It does the following:
+
+ 1. Create a scriptable ``new_Instances`` class which behaves similarly to ``Instances``,
+ but with all attributes been "static".
+ The attributes need to be statically declared in the ``fields`` argument.
+ 2. Register ``new_Instances``, and force scripting compiler to
+ use it when trying to compile ``Instances``.
+
+ After this function, the process will be reverted. User should be able to script another model
+ using different fields.
+
+ Example:
+ Assume that ``Instances`` in the model consist of two attributes named
+ ``proposal_boxes`` and ``objectness_logits`` with type :class:`Boxes` and
+ :class:`Tensor` respectively during inference. You can call this function like:
+ ::
+ fields = {"proposal_boxes": Boxes, "objectness_logits": torch.Tensor}
+ torchscipt_model = scripting_with_instances(model, fields)
+
+ Note:
+ It only support models in evaluation mode.
+
+ Args:
+ model (nn.Module): The input model to be exported by scripting.
+ fields (Dict[str, type]): Attribute names and corresponding type that
+ ``Instances`` will use in the model. Note that all attributes used in ``Instances``
+ need to be added, regardless of whether they are inputs/outputs of the model.
+ Data type not defined in detectron2 is not supported for now.
+
+ Returns:
+ torch.jit.ScriptModule: the model in torchscript format
+ """
+ assert (
+ not model.training
+ ), "Currently we only support exporting models in evaluation mode to torchscript"
+
+ with freeze_training_mode(model), patch_instances(fields):
+ scripted_model = torch.jit.script(model)
+ return scripted_model
+
+
+# alias for old name
+export_torchscript_with_instances = scripting_with_instances
+
+
+def dump_torchscript_IR(model, dir):
+ """
+ Dump IR of a TracedModule/ScriptModule/Function in various format (code, graph,
+ inlined graph). Useful for debugging.
+
+ Args:
+ model (TracedModule/ScriptModule/ScriptFUnction): traced or scripted module
+ dir (str): output directory to dump files.
+ """
+ dir = os.path.expanduser(dir)
+ PathManager.mkdirs(dir)
+
+ def _get_script_mod(mod):
+ if isinstance(mod, torch.jit.TracedModule):
+ return mod._actual_script_module
+ return mod
+
+ # Dump pretty-printed code: https://pytorch.org/docs/stable/jit.html#inspecting-code
+ with PathManager.open(os.path.join(dir, "model_ts_code.txt"), "w") as f:
+
+ def get_code(mod):
+ # Try a few ways to get code using private attributes.
+ try:
+ # This contains more information than just `mod.code`
+ return _get_script_mod(mod)._c.code
+ except AttributeError:
+ pass
+ try:
+ return mod.code
+ except AttributeError:
+ return None
+
+ def dump_code(prefix, mod):
+ code = get_code(mod)
+ name = prefix or "root model"
+ if code is None:
+ f.write(f"Could not found code for {name} (type={mod.original_name})\n")
+ f.write("\n")
+ else:
+ f.write(f"\nCode for {name}, type={mod.original_name}:\n")
+ f.write(code)
+ f.write("\n")
+ f.write("-" * 80)
+
+ for name, m in mod.named_children():
+ dump_code(prefix + "." + name, m)
+
+ if isinstance(model, torch.jit.ScriptFunction):
+ f.write(get_code(model))
+ else:
+ dump_code("", model)
+
+ def _get_graph(model):
+ try:
+ # Recursively dump IR of all modules
+ return _get_script_mod(model)._c.dump_to_str(True, False, False)
+ except AttributeError:
+ return model.graph.str()
+
+ with PathManager.open(os.path.join(dir, "model_ts_IR.txt"), "w") as f:
+ f.write(_get_graph(model))
+
+ # Dump IR of the entire graph (all submodules inlined)
+ with PathManager.open(os.path.join(dir, "model_ts_IR_inlined.txt"), "w") as f:
+ f.write(str(model.inlined_graph))
+
+ if not isinstance(model, torch.jit.ScriptFunction):
+ # Dump the model structure in pytorch style
+ with PathManager.open(os.path.join(dir, "model.txt"), "w") as f:
+ f.write(str(model))
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/torchscript_patch.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/torchscript_patch.py
new file mode 100644
index 0000000000000000000000000000000000000000..24c69b25dbec19221bcd8fc2e928a8393dd3aaf6
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/export/torchscript_patch.py
@@ -0,0 +1,406 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+
+import os
+import sys
+import tempfile
+from contextlib import ExitStack, contextmanager
+from copy import deepcopy
+from unittest import mock
+import torch
+from torch import nn
+
+# need some explicit imports due to https://github.com/pytorch/pytorch/issues/38964
+import annotator.oneformer.detectron2 # noqa F401
+from annotator.oneformer.detectron2.structures import Boxes, Instances
+from annotator.oneformer.detectron2.utils.env import _import_file
+
+_counter = 0
+
+
+def _clear_jit_cache():
+ from torch.jit._recursive import concrete_type_store
+ from torch.jit._state import _jit_caching_layer
+
+ concrete_type_store.type_store.clear() # for modules
+ _jit_caching_layer.clear() # for free functions
+
+
+def _add_instances_conversion_methods(newInstances):
+ """
+ Add from_instances methods to the scripted Instances class.
+ """
+ cls_name = newInstances.__name__
+
+ @torch.jit.unused
+ def from_instances(instances: Instances):
+ """
+ Create scripted Instances from original Instances
+ """
+ fields = instances.get_fields()
+ image_size = instances.image_size
+ ret = newInstances(image_size)
+ for name, val in fields.items():
+ assert hasattr(ret, f"_{name}"), f"No attribute named {name} in {cls_name}"
+ setattr(ret, name, deepcopy(val))
+ return ret
+
+ newInstances.from_instances = from_instances
+
+
+@contextmanager
+def patch_instances(fields):
+ """
+ A contextmanager, under which the Instances class in detectron2 is replaced
+ by a statically-typed scriptable class, defined by `fields`.
+ See more in `scripting_with_instances`.
+ """
+
+ with tempfile.TemporaryDirectory(prefix="detectron2") as dir, tempfile.NamedTemporaryFile(
+ mode="w", encoding="utf-8", suffix=".py", dir=dir, delete=False
+ ) as f:
+ try:
+ # Objects that use Instances should not reuse previously-compiled
+ # results in cache, because `Instances` could be a new class each time.
+ _clear_jit_cache()
+
+ cls_name, s = _gen_instance_module(fields)
+ f.write(s)
+ f.flush()
+ f.close()
+
+ module = _import(f.name)
+ new_instances = getattr(module, cls_name)
+ _ = torch.jit.script(new_instances)
+ # let torchscript think Instances was scripted already
+ Instances.__torch_script_class__ = True
+ # let torchscript find new_instances when looking for the jit type of Instances
+ Instances._jit_override_qualname = torch._jit_internal._qualified_name(new_instances)
+
+ _add_instances_conversion_methods(new_instances)
+ yield new_instances
+ finally:
+ try:
+ del Instances.__torch_script_class__
+ del Instances._jit_override_qualname
+ except AttributeError:
+ pass
+ sys.modules.pop(module.__name__)
+
+
+def _gen_instance_class(fields):
+ """
+ Args:
+ fields (dict[name: type])
+ """
+
+ class _FieldType:
+ def __init__(self, name, type_):
+ assert isinstance(name, str), f"Field name must be str, got {name}"
+ self.name = name
+ self.type_ = type_
+ self.annotation = f"{type_.__module__}.{type_.__name__}"
+
+ fields = [_FieldType(k, v) for k, v in fields.items()]
+
+ def indent(level, s):
+ return " " * 4 * level + s
+
+ lines = []
+
+ global _counter
+ _counter += 1
+
+ cls_name = "ScriptedInstances{}".format(_counter)
+
+ field_names = tuple(x.name for x in fields)
+ extra_args = ", ".join([f"{f.name}: Optional[{f.annotation}] = None" for f in fields])
+ lines.append(
+ f"""
+class {cls_name}:
+ def __init__(self, image_size: Tuple[int, int], {extra_args}):
+ self.image_size = image_size
+ self._field_names = {field_names}
+"""
+ )
+
+ for f in fields:
+ lines.append(
+ indent(2, f"self._{f.name} = torch.jit.annotate(Optional[{f.annotation}], {f.name})")
+ )
+
+ for f in fields:
+ lines.append(
+ f"""
+ @property
+ def {f.name}(self) -> {f.annotation}:
+ # has to use a local for type refinement
+ # https://pytorch.org/docs/stable/jit_language_reference.html#optional-type-refinement
+ t = self._{f.name}
+ assert t is not None, "{f.name} is None and cannot be accessed!"
+ return t
+
+ @{f.name}.setter
+ def {f.name}(self, value: {f.annotation}) -> None:
+ self._{f.name} = value
+"""
+ )
+
+ # support method `__len__`
+ lines.append(
+ """
+ def __len__(self) -> int:
+"""
+ )
+ for f in fields:
+ lines.append(
+ f"""
+ t = self._{f.name}
+ if t is not None:
+ return len(t)
+"""
+ )
+ lines.append(
+ """
+ raise NotImplementedError("Empty Instances does not support __len__!")
+"""
+ )
+
+ # support method `has`
+ lines.append(
+ """
+ def has(self, name: str) -> bool:
+"""
+ )
+ for f in fields:
+ lines.append(
+ f"""
+ if name == "{f.name}":
+ return self._{f.name} is not None
+"""
+ )
+ lines.append(
+ """
+ return False
+"""
+ )
+
+ # support method `to`
+ none_args = ", None" * len(fields)
+ lines.append(
+ f"""
+ def to(self, device: torch.device) -> "{cls_name}":
+ ret = {cls_name}(self.image_size{none_args})
+"""
+ )
+ for f in fields:
+ if hasattr(f.type_, "to"):
+ lines.append(
+ f"""
+ t = self._{f.name}
+ if t is not None:
+ ret._{f.name} = t.to(device)
+"""
+ )
+ else:
+ # For now, ignore fields that cannot be moved to devices.
+ # Maybe can support other tensor-like classes (e.g. __torch_function__)
+ pass
+ lines.append(
+ """
+ return ret
+"""
+ )
+
+ # support method `getitem`
+ none_args = ", None" * len(fields)
+ lines.append(
+ f"""
+ def __getitem__(self, item) -> "{cls_name}":
+ ret = {cls_name}(self.image_size{none_args})
+"""
+ )
+ for f in fields:
+ lines.append(
+ f"""
+ t = self._{f.name}
+ if t is not None:
+ ret._{f.name} = t[item]
+"""
+ )
+ lines.append(
+ """
+ return ret
+"""
+ )
+
+ # support method `cat`
+ # this version does not contain checks that all instances have same size and fields
+ none_args = ", None" * len(fields)
+ lines.append(
+ f"""
+ def cat(self, instances: List["{cls_name}"]) -> "{cls_name}":
+ ret = {cls_name}(self.image_size{none_args})
+"""
+ )
+ for f in fields:
+ lines.append(
+ f"""
+ t = self._{f.name}
+ if t is not None:
+ values: List[{f.annotation}] = [x.{f.name} for x in instances]
+ if torch.jit.isinstance(t, torch.Tensor):
+ ret._{f.name} = torch.cat(values, dim=0)
+ else:
+ ret._{f.name} = t.cat(values)
+"""
+ )
+ lines.append(
+ """
+ return ret"""
+ )
+
+ # support method `get_fields()`
+ lines.append(
+ """
+ def get_fields(self) -> Dict[str, Tensor]:
+ ret = {}
+ """
+ )
+ for f in fields:
+ if f.type_ == Boxes:
+ stmt = "t.tensor"
+ elif f.type_ == torch.Tensor:
+ stmt = "t"
+ else:
+ stmt = f'assert False, "unsupported type {str(f.type_)}"'
+ lines.append(
+ f"""
+ t = self._{f.name}
+ if t is not None:
+ ret["{f.name}"] = {stmt}
+ """
+ )
+ lines.append(
+ """
+ return ret"""
+ )
+ return cls_name, os.linesep.join(lines)
+
+
+def _gen_instance_module(fields):
+ # TODO: find a more automatic way to enable import of other classes
+ s = """
+from copy import deepcopy
+import torch
+from torch import Tensor
+import typing
+from typing import *
+
+import annotator.oneformer.detectron2
+from annotator.oneformer.detectron2.structures import Boxes, Instances
+
+"""
+
+ cls_name, cls_def = _gen_instance_class(fields)
+ s += cls_def
+ return cls_name, s
+
+
+def _import(path):
+ return _import_file(
+ "{}{}".format(sys.modules[__name__].__name__, _counter), path, make_importable=True
+ )
+
+
+@contextmanager
+def patch_builtin_len(modules=()):
+ """
+ Patch the builtin len() function of a few detectron2 modules
+ to use __len__ instead, because __len__ does not convert values to
+ integers and therefore is friendly to tracing.
+
+ Args:
+ modules (list[stsr]): names of extra modules to patch len(), in
+ addition to those in detectron2.
+ """
+
+ def _new_len(obj):
+ return obj.__len__()
+
+ with ExitStack() as stack:
+ MODULES = [
+ "detectron2.modeling.roi_heads.fast_rcnn",
+ "detectron2.modeling.roi_heads.mask_head",
+ "detectron2.modeling.roi_heads.keypoint_head",
+ ] + list(modules)
+ ctxs = [stack.enter_context(mock.patch(mod + ".len")) for mod in MODULES]
+ for m in ctxs:
+ m.side_effect = _new_len
+ yield
+
+
+def patch_nonscriptable_classes():
+ """
+ Apply patches on a few nonscriptable detectron2 classes.
+ Should not have side-effects on eager usage.
+ """
+ # __prepare_scriptable__ can also be added to models for easier maintenance.
+ # But it complicates the clean model code.
+
+ from annotator.oneformer.detectron2.modeling.backbone import ResNet, FPN
+
+ # Due to https://github.com/pytorch/pytorch/issues/36061,
+ # we change backbone to use ModuleList for scripting.
+ # (note: this changes param names in state_dict)
+
+ def prepare_resnet(self):
+ ret = deepcopy(self)
+ ret.stages = nn.ModuleList(ret.stages)
+ for k in self.stage_names:
+ delattr(ret, k)
+ return ret
+
+ ResNet.__prepare_scriptable__ = prepare_resnet
+
+ def prepare_fpn(self):
+ ret = deepcopy(self)
+ ret.lateral_convs = nn.ModuleList(ret.lateral_convs)
+ ret.output_convs = nn.ModuleList(ret.output_convs)
+ for name, _ in self.named_children():
+ if name.startswith("fpn_"):
+ delattr(ret, name)
+ return ret
+
+ FPN.__prepare_scriptable__ = prepare_fpn
+
+ # Annotate some attributes to be constants for the purpose of scripting,
+ # even though they are not constants in eager mode.
+ from annotator.oneformer.detectron2.modeling.roi_heads import StandardROIHeads
+
+ if hasattr(StandardROIHeads, "__annotations__"):
+ # copy first to avoid editing annotations of base class
+ StandardROIHeads.__annotations__ = deepcopy(StandardROIHeads.__annotations__)
+ StandardROIHeads.__annotations__["mask_on"] = torch.jit.Final[bool]
+ StandardROIHeads.__annotations__["keypoint_on"] = torch.jit.Final[bool]
+
+
+# These patches are not supposed to have side-effects.
+patch_nonscriptable_classes()
+
+
+@contextmanager
+def freeze_training_mode(model):
+ """
+ A context manager that annotates the "training" attribute of every submodule
+ to constant, so that the training codepath in these modules can be
+ meta-compiled away. Upon exiting, the annotations are reverted.
+ """
+ classes = {type(x) for x in model.modules()}
+ # __constants__ is the old way to annotate constants and not compatible
+ # with __annotations__ .
+ classes = {x for x in classes if not hasattr(x, "__constants__")}
+ for cls in classes:
+ cls.__annotations__["training"] = torch.jit.Final[bool]
+ yield
+ for cls in classes:
+ cls.__annotations__["training"] = bool
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/__init__.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..761a3d1c7afa049e9779ee9fc4d299e9aae38cad
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/__init__.py
@@ -0,0 +1,26 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+from .batch_norm import FrozenBatchNorm2d, get_norm, NaiveSyncBatchNorm, CycleBatchNormList
+from .deform_conv import DeformConv, ModulatedDeformConv
+from .mask_ops import paste_masks_in_image
+from .nms import batched_nms, batched_nms_rotated, nms, nms_rotated
+from .roi_align import ROIAlign, roi_align
+from .roi_align_rotated import ROIAlignRotated, roi_align_rotated
+from .shape_spec import ShapeSpec
+from .wrappers import (
+ BatchNorm2d,
+ Conv2d,
+ ConvTranspose2d,
+ cat,
+ interpolate,
+ Linear,
+ nonzero_tuple,
+ cross_entropy,
+ empty_input_loss_func_wrapper,
+ shapes_to_tensor,
+ move_device_like,
+)
+from .blocks import CNNBlockBase, DepthwiseSeparableConv2d
+from .aspp import ASPP
+from .losses import ciou_loss, diou_loss
+
+__all__ = [k for k in globals().keys() if not k.startswith("_")]
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/aspp.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/aspp.py
new file mode 100644
index 0000000000000000000000000000000000000000..14861aa9ede4fea6a69a49f189bcab997b558148
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/aspp.py
@@ -0,0 +1,144 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+
+from copy import deepcopy
+import fvcore.nn.weight_init as weight_init
+import torch
+from torch import nn
+from torch.nn import functional as F
+
+from .batch_norm import get_norm
+from .blocks import DepthwiseSeparableConv2d
+from .wrappers import Conv2d
+
+
+class ASPP(nn.Module):
+ """
+ Atrous Spatial Pyramid Pooling (ASPP).
+ """
+
+ def __init__(
+ self,
+ in_channels,
+ out_channels,
+ dilations,
+ *,
+ norm,
+ activation,
+ pool_kernel_size=None,
+ dropout: float = 0.0,
+ use_depthwise_separable_conv=False,
+ ):
+ """
+ Args:
+ in_channels (int): number of input channels for ASPP.
+ out_channels (int): number of output channels.
+ dilations (list): a list of 3 dilations in ASPP.
+ norm (str or callable): normalization for all conv layers.
+ See :func:`layers.get_norm` for supported format. norm is
+ applied to all conv layers except the conv following
+ global average pooling.
+ activation (callable): activation function.
+ pool_kernel_size (tuple, list): the average pooling size (kh, kw)
+ for image pooling layer in ASPP. If set to None, it always
+ performs global average pooling. If not None, it must be
+ divisible by the shape of inputs in forward(). It is recommended
+ to use a fixed input feature size in training, and set this
+ option to match this size, so that it performs global average
+ pooling in training, and the size of the pooling window stays
+ consistent in inference.
+ dropout (float): apply dropout on the output of ASPP. It is used in
+ the official DeepLab implementation with a rate of 0.1:
+ https://github.com/tensorflow/models/blob/21b73d22f3ed05b650e85ac50849408dd36de32e/research/deeplab/model.py#L532 # noqa
+ use_depthwise_separable_conv (bool): use DepthwiseSeparableConv2d
+ for 3x3 convs in ASPP, proposed in :paper:`DeepLabV3+`.
+ """
+ super(ASPP, self).__init__()
+ assert len(dilations) == 3, "ASPP expects 3 dilations, got {}".format(len(dilations))
+ self.pool_kernel_size = pool_kernel_size
+ self.dropout = dropout
+ use_bias = norm == ""
+ self.convs = nn.ModuleList()
+ # conv 1x1
+ self.convs.append(
+ Conv2d(
+ in_channels,
+ out_channels,
+ kernel_size=1,
+ bias=use_bias,
+ norm=get_norm(norm, out_channels),
+ activation=deepcopy(activation),
+ )
+ )
+ weight_init.c2_xavier_fill(self.convs[-1])
+ # atrous convs
+ for dilation in dilations:
+ if use_depthwise_separable_conv:
+ self.convs.append(
+ DepthwiseSeparableConv2d(
+ in_channels,
+ out_channels,
+ kernel_size=3,
+ padding=dilation,
+ dilation=dilation,
+ norm1=norm,
+ activation1=deepcopy(activation),
+ norm2=norm,
+ activation2=deepcopy(activation),
+ )
+ )
+ else:
+ self.convs.append(
+ Conv2d(
+ in_channels,
+ out_channels,
+ kernel_size=3,
+ padding=dilation,
+ dilation=dilation,
+ bias=use_bias,
+ norm=get_norm(norm, out_channels),
+ activation=deepcopy(activation),
+ )
+ )
+ weight_init.c2_xavier_fill(self.convs[-1])
+ # image pooling
+ # We do not add BatchNorm because the spatial resolution is 1x1,
+ # the original TF implementation has BatchNorm.
+ if pool_kernel_size is None:
+ image_pooling = nn.Sequential(
+ nn.AdaptiveAvgPool2d(1),
+ Conv2d(in_channels, out_channels, 1, bias=True, activation=deepcopy(activation)),
+ )
+ else:
+ image_pooling = nn.Sequential(
+ nn.AvgPool2d(kernel_size=pool_kernel_size, stride=1),
+ Conv2d(in_channels, out_channels, 1, bias=True, activation=deepcopy(activation)),
+ )
+ weight_init.c2_xavier_fill(image_pooling[1])
+ self.convs.append(image_pooling)
+
+ self.project = Conv2d(
+ 5 * out_channels,
+ out_channels,
+ kernel_size=1,
+ bias=use_bias,
+ norm=get_norm(norm, out_channels),
+ activation=deepcopy(activation),
+ )
+ weight_init.c2_xavier_fill(self.project)
+
+ def forward(self, x):
+ size = x.shape[-2:]
+ if self.pool_kernel_size is not None:
+ if size[0] % self.pool_kernel_size[0] or size[1] % self.pool_kernel_size[1]:
+ raise ValueError(
+ "`pool_kernel_size` must be divisible by the shape of inputs. "
+ "Input size: {} `pool_kernel_size`: {}".format(size, self.pool_kernel_size)
+ )
+ res = []
+ for conv in self.convs:
+ res.append(conv(x))
+ res[-1] = F.interpolate(res[-1], size=size, mode="bilinear", align_corners=False)
+ res = torch.cat(res, dim=1)
+ res = self.project(res)
+ res = F.dropout(res, self.dropout, training=self.training) if self.dropout > 0 else res
+ return res
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/batch_norm.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/batch_norm.py
new file mode 100644
index 0000000000000000000000000000000000000000..32a1e05470065e75b6caad18d36211d27af8eec0
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/batch_norm.py
@@ -0,0 +1,300 @@
+# Copyright (c) Facebook, Inc. and its affiliates.
+import torch
+import torch.distributed as dist
+from fvcore.nn.distributed import differentiable_all_reduce
+from torch import nn
+from torch.nn import functional as F
+
+from annotator.oneformer.detectron2.utils import comm, env
+
+from .wrappers import BatchNorm2d
+
+
+class FrozenBatchNorm2d(nn.Module):
+ """
+ BatchNorm2d where the batch statistics and the affine parameters are fixed.
+
+ It contains non-trainable buffers called
+ "weight" and "bias", "running_mean", "running_var",
+ initialized to perform identity transformation.
+
+ The pre-trained backbone models from Caffe2 only contain "weight" and "bias",
+ which are computed from the original four parameters of BN.
+ The affine transform `x * weight + bias` will perform the equivalent
+ computation of `(x - running_mean) / sqrt(running_var) * weight + bias`.
+ When loading a backbone model from Caffe2, "running_mean" and "running_var"
+ will be left unchanged as identity transformation.
+
+ Other pre-trained backbone models may contain all 4 parameters.
+
+ The forward is implemented by `F.batch_norm(..., training=False)`.
+ """
+
+ _version = 3
+
+ def __init__(self, num_features, eps=1e-5):
+ super().__init__()
+ self.num_features = num_features
+ self.eps = eps
+ self.register_buffer("weight", torch.ones(num_features))
+ self.register_buffer("bias", torch.zeros(num_features))
+ self.register_buffer("running_mean", torch.zeros(num_features))
+ self.register_buffer("running_var", torch.ones(num_features) - eps)
+
+ def forward(self, x):
+ if x.requires_grad:
+ # When gradients are needed, F.batch_norm will use extra memory
+ # because its backward op computes gradients for weight/bias as well.
+ scale = self.weight * (self.running_var + self.eps).rsqrt()
+ bias = self.bias - self.running_mean * scale
+ scale = scale.reshape(1, -1, 1, 1)
+ bias = bias.reshape(1, -1, 1, 1)
+ out_dtype = x.dtype # may be half
+ return x * scale.to(out_dtype) + bias.to(out_dtype)
+ else:
+ # When gradients are not needed, F.batch_norm is a single fused op
+ # and provide more optimization opportunities.
+ return F.batch_norm(
+ x,
+ self.running_mean,
+ self.running_var,
+ self.weight,
+ self.bias,
+ training=False,
+ eps=self.eps,
+ )
+
+ def _load_from_state_dict(
+ self, state_dict, prefix, local_metadata, strict, missing_keys, unexpected_keys, error_msgs
+ ):
+ version = local_metadata.get("version", None)
+
+ if version is None or version < 2:
+ # No running_mean/var in early versions
+ # This will silent the warnings
+ if prefix + "running_mean" not in state_dict:
+ state_dict[prefix + "running_mean"] = torch.zeros_like(self.running_mean)
+ if prefix + "running_var" not in state_dict:
+ state_dict[prefix + "running_var"] = torch.ones_like(self.running_var)
+
+ super()._load_from_state_dict(
+ state_dict, prefix, local_metadata, strict, missing_keys, unexpected_keys, error_msgs
+ )
+
+ def __repr__(self):
+ return "FrozenBatchNorm2d(num_features={}, eps={})".format(self.num_features, self.eps)
+
+ @classmethod
+ def convert_frozen_batchnorm(cls, module):
+ """
+ Convert all BatchNorm/SyncBatchNorm in module into FrozenBatchNorm.
+
+ Args:
+ module (torch.nn.Module):
+
+ Returns:
+ If module is BatchNorm/SyncBatchNorm, returns a new module.
+ Otherwise, in-place convert module and return it.
+
+ Similar to convert_sync_batchnorm in
+ https://github.com/pytorch/pytorch/blob/master/torch/nn/modules/batchnorm.py
+ """
+ bn_module = nn.modules.batchnorm
+ bn_module = (bn_module.BatchNorm2d, bn_module.SyncBatchNorm)
+ res = module
+ if isinstance(module, bn_module):
+ res = cls(module.num_features)
+ if module.affine:
+ res.weight.data = module.weight.data.clone().detach()
+ res.bias.data = module.bias.data.clone().detach()
+ res.running_mean.data = module.running_mean.data
+ res.running_var.data = module.running_var.data
+ res.eps = module.eps
+ else:
+ for name, child in module.named_children():
+ new_child = cls.convert_frozen_batchnorm(child)
+ if new_child is not child:
+ res.add_module(name, new_child)
+ return res
+
+
+def get_norm(norm, out_channels):
+ """
+ Args:
+ norm (str or callable): either one of BN, SyncBN, FrozenBN, GN;
+ or a callable that takes a channel number and returns
+ the normalization layer as a nn.Module.
+
+ Returns:
+ nn.Module or None: the normalization layer
+ """
+ if norm is None:
+ return None
+ if isinstance(norm, str):
+ if len(norm) == 0:
+ return None
+ norm = {
+ "BN": BatchNorm2d,
+ # Fixed in https://github.com/pytorch/pytorch/pull/36382
+ "SyncBN": NaiveSyncBatchNorm if env.TORCH_VERSION <= (1, 5) else nn.SyncBatchNorm,
+ "FrozenBN": FrozenBatchNorm2d,
+ "GN": lambda channels: nn.GroupNorm(32, channels),
+ # for debugging:
+ "nnSyncBN": nn.SyncBatchNorm,
+ "naiveSyncBN": NaiveSyncBatchNorm,
+ # expose stats_mode N as an option to caller, required for zero-len inputs
+ "naiveSyncBN_N": lambda channels: NaiveSyncBatchNorm(channels, stats_mode="N"),
+ "LN": lambda channels: LayerNorm(channels),
+ }[norm]
+ return norm(out_channels)
+
+
+class NaiveSyncBatchNorm(BatchNorm2d):
+ """
+ In PyTorch<=1.5, ``nn.SyncBatchNorm`` has incorrect gradient
+ when the batch size on each worker is different.
+ (e.g., when scale augmentation is used, or when it is applied to mask head).
+
+ This is a slower but correct alternative to `nn.SyncBatchNorm`.
+
+ Note:
+ There isn't a single definition of Sync BatchNorm.
+
+ When ``stats_mode==""``, this module computes overall statistics by using
+ statistics of each worker with equal weight. The result is true statistics
+ of all samples (as if they are all on one worker) only when all workers
+ have the same (N, H, W). This mode does not support inputs with zero batch size.
+
+ When ``stats_mode=="N"``, this module computes overall statistics by weighting
+ the statistics of each worker by their ``N``. The result is true statistics
+ of all samples (as if they are all on one worker) only when all workers
+ have the same (H, W). It is slower than ``stats_mode==""``.
+
+ Even though the result of this module may not be the true statistics of all samples,
+ it may still be reasonable because it might be preferrable to assign equal weights
+ to all workers, regardless of their (H, W) dimension, instead of putting larger weight
+ on larger images. From preliminary experiments, little difference is found between such
+ a simplified implementation and an accurate computation of overall mean & variance.
+ """
+
+ def __init__(self, *args, stats_mode="", **kwargs):
+ super().__init__(*args, **kwargs)
+ assert stats_mode in ["", "N"]
+ self._stats_mode = stats_mode
+
+ def forward(self, input):
+ if comm.get_world_size() == 1 or not self.training:
+ return super().forward(input)
+
+ B, C = input.shape[0], input.shape[1]
+
+ half_input = input.dtype == torch.float16
+ if half_input:
+ # fp16 does not have good enough numerics for the reduction here
+ input = input.float()
+ mean = torch.mean(input, dim=[0, 2, 3])
+ meansqr = torch.mean(input * input, dim=[0, 2, 3])
+
+ if self._stats_mode == "":
+ assert B > 0, 'SyncBatchNorm(stats_mode="") does not support zero batch size.'
+ vec = torch.cat([mean, meansqr], dim=0)
+ vec = differentiable_all_reduce(vec) * (1.0 / dist.get_world_size())
+ mean, meansqr = torch.split(vec, C)
+ momentum = self.momentum
+ else:
+ if B == 0:
+ vec = torch.zeros([2 * C + 1], device=mean.device, dtype=mean.dtype)
+ vec = vec + input.sum() # make sure there is gradient w.r.t input
+ else:
+ vec = torch.cat(
+ [mean, meansqr, torch.ones([1], device=mean.device, dtype=mean.dtype)], dim=0
+ )
+ vec = differentiable_all_reduce(vec * B)
+
+ total_batch = vec[-1].detach()
+ momentum = total_batch.clamp(max=1) * self.momentum # no update if total_batch is 0
+ mean, meansqr, _ = torch.split(vec / total_batch.clamp(min=1), C) # avoid div-by-zero
+
+ var = meansqr - mean * mean
+ invstd = torch.rsqrt(var + self.eps)
+ scale = self.weight * invstd
+ bias = self.bias - mean * scale
+ scale = scale.reshape(1, -1, 1, 1)
+ bias = bias.reshape(1, -1, 1, 1)
+
+ self.running_mean += momentum * (mean.detach() - self.running_mean)
+ self.running_var += momentum * (var.detach() - self.running_var)
+ ret = input * scale + bias
+ if half_input:
+ ret = ret.half()
+ return ret
+
+
+class CycleBatchNormList(nn.ModuleList):
+ """
+ Implement domain-specific BatchNorm by cycling.
+
+ When a BatchNorm layer is used for multiple input domains or input
+ features, it might need to maintain a separate test-time statistics
+ for each domain. See Sec 5.2 in :paper:`rethinking-batchnorm`.
+
+ This module implements it by using N separate BN layers
+ and it cycles through them every time a forward() is called.
+
+ NOTE: The caller of this module MUST guarantee to always call
+ this module by multiple of N times. Otherwise its test-time statistics
+ will be incorrect.
+ """
+
+ def __init__(self, length: int, bn_class=nn.BatchNorm2d, **kwargs):
+ """
+ Args:
+ length: number of BatchNorm layers to cycle.
+ bn_class: the BatchNorm class to use
+ kwargs: arguments of the BatchNorm class, such as num_features.
+ """
+ self._affine = kwargs.pop("affine", True)
+ super().__init__([bn_class(**kwargs, affine=False) for k in range(length)])
+ if self._affine:
+ # shared affine, domain-specific BN
+ channels = self[0].num_features
+ self.weight = nn.Parameter(torch.ones(channels))
+ self.bias = nn.Parameter(torch.zeros(channels))
+ self._pos = 0
+
+ def forward(self, x):
+ ret = self[self._pos](x)
+ self._pos = (self._pos + 1) % len(self)
+
+ if self._affine:
+ w = self.weight.reshape(1, -1, 1, 1)
+ b = self.bias.reshape(1, -1, 1, 1)
+ return ret * w + b
+ else:
+ return ret
+
+ def extra_repr(self):
+ return f"affine={self._affine}"
+
+
+class LayerNorm(nn.Module):
+ """
+ A LayerNorm variant, popularized by Transformers, that performs point-wise mean and
+ variance normalization over the channel dimension for inputs that have shape
+ (batch_size, channels, height, width).
+ https://github.com/facebookresearch/ConvNeXt/blob/d1fa8f6fef0a165b27399986cc2bdacc92777e40/models/convnext.py#L119 # noqa B950
+ """
+
+ def __init__(self, normalized_shape, eps=1e-6):
+ super().__init__()
+ self.weight = nn.Parameter(torch.ones(normalized_shape))
+ self.bias = nn.Parameter(torch.zeros(normalized_shape))
+ self.eps = eps
+ self.normalized_shape = (normalized_shape,)
+
+ def forward(self, x):
+ u = x.mean(1, keepdim=True)
+ s = (x - u).pow(2).mean(1, keepdim=True)
+ x = (x - u) / torch.sqrt(s + self.eps)
+ x = self.weight[:, None, None] * x + self.bias[:, None, None]
+ return x
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/blocks.py b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/blocks.py
new file mode 100644
index 0000000000000000000000000000000000000000..1995a4bf7339e8deb7eaaffda4f819dda55e7ac7
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/blocks.py
@@ -0,0 +1,111 @@
+# -*- coding: utf-8 -*-
+# Copyright (c) Facebook, Inc. and its affiliates.
+
+import fvcore.nn.weight_init as weight_init
+from torch import nn
+
+from .batch_norm import FrozenBatchNorm2d, get_norm
+from .wrappers import Conv2d
+
+
+"""
+CNN building blocks.
+"""
+
+
+class CNNBlockBase(nn.Module):
+ """
+ A CNN block is assumed to have input channels, output channels and a stride.
+ The input and output of `forward()` method must be NCHW tensors.
+ The method can perform arbitrary computation but must match the given
+ channels and stride specification.
+
+ Attribute:
+ in_channels (int):
+ out_channels (int):
+ stride (int):
+ """
+
+ def __init__(self, in_channels, out_channels, stride):
+ """
+ The `__init__` method of any subclass should also contain these arguments.
+
+ Args:
+ in_channels (int):
+ out_channels (int):
+ stride (int):
+ """
+ super().__init__()
+ self.in_channels = in_channels
+ self.out_channels = out_channels
+ self.stride = stride
+
+ def freeze(self):
+ """
+ Make this block not trainable.
+ This method sets all parameters to `requires_grad=False`,
+ and convert all BatchNorm layers to FrozenBatchNorm
+
+ Returns:
+ the block itself
+ """
+ for p in self.parameters():
+ p.requires_grad = False
+ FrozenBatchNorm2d.convert_frozen_batchnorm(self)
+ return self
+
+
+class DepthwiseSeparableConv2d(nn.Module):
+ """
+ A kxk depthwise convolution + a 1x1 convolution.
+
+ In :paper:`xception`, norm & activation are applied on the second conv.
+ :paper:`mobilenet` uses norm & activation on both convs.
+ """
+
+ def __init__(
+ self,
+ in_channels,
+ out_channels,
+ kernel_size=3,
+ padding=1,
+ dilation=1,
+ *,
+ norm1=None,
+ activation1=None,
+ norm2=None,
+ activation2=None,
+ ):
+ """
+ Args:
+ norm1, norm2 (str or callable): normalization for the two conv layers.
+ activation1, activation2 (callable(Tensor) -> Tensor): activation
+ function for the two conv layers.
+ """
+ super().__init__()
+ self.depthwise = Conv2d(
+ in_channels,
+ in_channels,
+ kernel_size=kernel_size,
+ padding=padding,
+ dilation=dilation,
+ groups=in_channels,
+ bias=not norm1,
+ norm=get_norm(norm1, in_channels),
+ activation=activation1,
+ )
+ self.pointwise = Conv2d(
+ in_channels,
+ out_channels,
+ kernel_size=1,
+ bias=not norm2,
+ norm=get_norm(norm2, out_channels),
+ activation=activation2,
+ )
+
+ # default initialization
+ weight_init.c2_msra_fill(self.depthwise)
+ weight_init.c2_msra_fill(self.pointwise)
+
+ def forward(self, x):
+ return self.pointwise(self.depthwise(x))
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/README.md b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..778ed3da0bae89820831bcd8a72ff7b9cad8d4dd
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/README.md
@@ -0,0 +1,7 @@
+
+
+To add a new Op:
+
+1. Create a new directory
+2. Implement new ops there
+3. Delcare its Python interface in `vision.cpp`.
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/ROIAlignRotated/ROIAlignRotated.h b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/ROIAlignRotated/ROIAlignRotated.h
new file mode 100644
index 0000000000000000000000000000000000000000..03f4211003f42f601f0cfcf4a690f5da4a0a1f67
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/ROIAlignRotated/ROIAlignRotated.h
@@ -0,0 +1,115 @@
+// Copyright (c) Facebook, Inc. and its affiliates.
+#pragma once
+#include
+
+namespace detectron2 {
+
+at::Tensor ROIAlignRotated_forward_cpu(
+ const at::Tensor& input,
+ const at::Tensor& rois,
+ const float spatial_scale,
+ const int pooled_height,
+ const int pooled_width,
+ const int sampling_ratio);
+
+at::Tensor ROIAlignRotated_backward_cpu(
+ const at::Tensor& grad,
+ const at::Tensor& rois,
+ const float spatial_scale,
+ const int pooled_height,
+ const int pooled_width,
+ const int batch_size,
+ const int channels,
+ const int height,
+ const int width,
+ const int sampling_ratio);
+
+#if defined(WITH_CUDA) || defined(WITH_HIP)
+at::Tensor ROIAlignRotated_forward_cuda(
+ const at::Tensor& input,
+ const at::Tensor& rois,
+ const float spatial_scale,
+ const int pooled_height,
+ const int pooled_width,
+ const int sampling_ratio);
+
+at::Tensor ROIAlignRotated_backward_cuda(
+ const at::Tensor& grad,
+ const at::Tensor& rois,
+ const float spatial_scale,
+ const int pooled_height,
+ const int pooled_width,
+ const int batch_size,
+ const int channels,
+ const int height,
+ const int width,
+ const int sampling_ratio);
+#endif
+
+// Interface for Python
+inline at::Tensor ROIAlignRotated_forward(
+ const at::Tensor& input,
+ const at::Tensor& rois,
+ const double spatial_scale,
+ const int64_t pooled_height,
+ const int64_t pooled_width,
+ const int64_t sampling_ratio) {
+ if (input.is_cuda()) {
+#if defined(WITH_CUDA) || defined(WITH_HIP)
+ return ROIAlignRotated_forward_cuda(
+ input,
+ rois,
+ spatial_scale,
+ pooled_height,
+ pooled_width,
+ sampling_ratio);
+#else
+ AT_ERROR("Detectron2 is not compiled with GPU support!");
+#endif
+ }
+ return ROIAlignRotated_forward_cpu(
+ input, rois, spatial_scale, pooled_height, pooled_width, sampling_ratio);
+}
+
+inline at::Tensor ROIAlignRotated_backward(
+ const at::Tensor& grad,
+ const at::Tensor& rois,
+ const double spatial_scale,
+ const int64_t pooled_height,
+ const int64_t pooled_width,
+ const int64_t batch_size,
+ const int64_t channels,
+ const int64_t height,
+ const int64_t width,
+ const int64_t sampling_ratio) {
+ if (grad.is_cuda()) {
+#if defined(WITH_CUDA) || defined(WITH_HIP)
+ return ROIAlignRotated_backward_cuda(
+ grad,
+ rois,
+ spatial_scale,
+ pooled_height,
+ pooled_width,
+ batch_size,
+ channels,
+ height,
+ width,
+ sampling_ratio);
+#else
+ AT_ERROR("Detectron2 is not compiled with GPU support!");
+#endif
+ }
+ return ROIAlignRotated_backward_cpu(
+ grad,
+ rois,
+ spatial_scale,
+ pooled_height,
+ pooled_width,
+ batch_size,
+ channels,
+ height,
+ width,
+ sampling_ratio);
+}
+
+} // namespace detectron2
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/ROIAlignRotated/ROIAlignRotated_cpu.cpp b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/ROIAlignRotated/ROIAlignRotated_cpu.cpp
new file mode 100644
index 0000000000000000000000000000000000000000..2a3d3056cc71a4acaafb570739a9dd247a7eb1ed
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/ROIAlignRotated/ROIAlignRotated_cpu.cpp
@@ -0,0 +1,522 @@
+// Copyright (c) Facebook, Inc. and its affiliates.
+#include
+#include "ROIAlignRotated.h"
+
+// Note: this implementation originates from the Caffe2 ROIAlignRotated Op
+// and PyTorch ROIAlign (non-rotated) Op implementations.
+// The key difference between this implementation and those ones is
+// we don't do "legacy offset" in this version, as there aren't many previous
+// works, if any, using the "legacy" ROIAlignRotated Op.
+// This would make the interface a bit cleaner.
+
+namespace detectron2 {
+
+namespace {
+template
+struct PreCalc {
+ int pos1;
+ int pos2;
+ int pos3;
+ int pos4;
+ T w1;
+ T w2;
+ T w3;
+ T w4;
+};
+
+template
+void pre_calc_for_bilinear_interpolate(
+ const int height,
+ const int width,
+ const int pooled_height,
+ const int pooled_width,
+ const int iy_upper,
+ const int ix_upper,
+ T roi_start_h,
+ T roi_start_w,
+ T bin_size_h,
+ T bin_size_w,
+ int roi_bin_grid_h,
+ int roi_bin_grid_w,
+ T roi_center_h,
+ T roi_center_w,
+ T cos_theta,
+ T sin_theta,
+ std::vector>& pre_calc) {
+ int pre_calc_index = 0;
+ for (int ph = 0; ph < pooled_height; ph++) {
+ for (int pw = 0; pw < pooled_width; pw++) {
+ for (int iy = 0; iy < iy_upper; iy++) {
+ const T yy = roi_start_h + ph * bin_size_h +
+ static_cast(iy + .5f) * bin_size_h /
+ static_cast(roi_bin_grid_h); // e.g., 0.5, 1.5
+ for (int ix = 0; ix < ix_upper; ix++) {
+ const T xx = roi_start_w + pw * bin_size_w +
+ static_cast(ix + .5f) * bin_size_w /
+ static_cast(roi_bin_grid_w);
+
+ // Rotate by theta around the center and translate
+ // In image space, (y, x) is the order for Right Handed System,
+ // and this is essentially multiplying the point by a rotation matrix
+ // to rotate it counterclockwise through angle theta.
+ T y = yy * cos_theta - xx * sin_theta + roi_center_h;
+ T x = yy * sin_theta + xx * cos_theta + roi_center_w;
+ // deal with: inverse elements are out of feature map boundary
+ if (y < -1.0 || y > height || x < -1.0 || x > width) {
+ // empty
+ PreCalc pc;
+ pc.pos1 = 0;
+ pc.pos2 = 0;
+ pc.pos3 = 0;
+ pc.pos4 = 0;
+ pc.w1 = 0;
+ pc.w2 = 0;
+ pc.w3 = 0;
+ pc.w4 = 0;
+ pre_calc[pre_calc_index] = pc;
+ pre_calc_index += 1;
+ continue;
+ }
+
+ if (y < 0) {
+ y = 0;
+ }
+ if (x < 0) {
+ x = 0;
+ }
+
+ int y_low = (int)y;
+ int x_low = (int)x;
+ int y_high;
+ int x_high;
+
+ if (y_low >= height - 1) {
+ y_high = y_low = height - 1;
+ y = (T)y_low;
+ } else {
+ y_high = y_low + 1;
+ }
+
+ if (x_low >= width - 1) {
+ x_high = x_low = width - 1;
+ x = (T)x_low;
+ } else {
+ x_high = x_low + 1;
+ }
+
+ T ly = y - y_low;
+ T lx = x - x_low;
+ T hy = 1. - ly, hx = 1. - lx;
+ T w1 = hy * hx, w2 = hy * lx, w3 = ly * hx, w4 = ly * lx;
+
+ // save weights and indices
+ PreCalc pc;
+ pc.pos1 = y_low * width + x_low;
+ pc.pos2 = y_low * width + x_high;
+ pc.pos3 = y_high * width + x_low;
+ pc.pos4 = y_high * width + x_high;
+ pc.w1 = w1;
+ pc.w2 = w2;
+ pc.w3 = w3;
+ pc.w4 = w4;
+ pre_calc[pre_calc_index] = pc;
+
+ pre_calc_index += 1;
+ }
+ }
+ }
+ }
+}
+
+template
+void bilinear_interpolate_gradient(
+ const int height,
+ const int width,
+ T y,
+ T x,
+ T& w1,
+ T& w2,
+ T& w3,
+ T& w4,
+ int& x_low,
+ int& x_high,
+ int& y_low,
+ int& y_high) {
+ // deal with cases that inverse elements are out of feature map boundary
+ if (y < -1.0 || y > height || x < -1.0 || x > width) {
+ // empty
+ w1 = w2 = w3 = w4 = 0.;
+ x_low = x_high = y_low = y_high = -1;
+ return;
+ }
+
+ if (y < 0) {
+ y = 0;
+ }
+
+ if (x < 0) {
+ x = 0;
+ }
+
+ y_low = (int)y;
+ x_low = (int)x;
+
+ if (y_low >= height - 1) {
+ y_high = y_low = height - 1;
+ y = (T)y_low;
+ } else {
+ y_high = y_low + 1;
+ }
+
+ if (x_low >= width - 1) {
+ x_high = x_low = width - 1;
+ x = (T)x_low;
+ } else {
+ x_high = x_low + 1;
+ }
+
+ T ly = y - y_low;
+ T lx = x - x_low;
+ T hy = 1. - ly, hx = 1. - lx;
+
+ // reference in forward
+ // T v1 = input[y_low * width + x_low];
+ // T v2 = input[y_low * width + x_high];
+ // T v3 = input[y_high * width + x_low];
+ // T v4 = input[y_high * width + x_high];
+ // T val = (w1 * v1 + w2 * v2 + w3 * v3 + w4 * v4);
+
+ w1 = hy * hx, w2 = hy * lx, w3 = ly * hx, w4 = ly * lx;
+
+ return;
+}
+
+template
+inline void add(T* address, const T& val) {
+ *address += val;
+}
+
+} // namespace
+
+template
+void ROIAlignRotatedForward(
+ const int nthreads,
+ const T* input,
+ const T& spatial_scale,
+ const int channels,
+ const int height,
+ const int width,
+ const int pooled_height,
+ const int pooled_width,
+ const int sampling_ratio,
+ const T* rois,
+ T* output) {
+ int n_rois = nthreads / channels / pooled_width / pooled_height;
+ // (n, c, ph, pw) is an element in the pooled output
+ // can be parallelized using omp
+ // #pragma omp parallel for num_threads(32)
+ for (int n = 0; n < n_rois; n++) {
+ int index_n = n * channels * pooled_width * pooled_height;
+
+ const T* current_roi = rois + n * 6;
+ int roi_batch_ind = current_roi[0];
+
+ // Do not use rounding; this implementation detail is critical
+ // ROIAlignRotated supports align == true, i.e., continuous coordinate
+ // by default, thus the 0.5 offset
+ T offset = (T)0.5;
+ T roi_center_w = current_roi[1] * spatial_scale - offset;
+ T roi_center_h = current_roi[2] * spatial_scale - offset;
+ T roi_width = current_roi[3] * spatial_scale;
+ T roi_height = current_roi[4] * spatial_scale;
+ T theta = current_roi[5] * M_PI / 180.0;
+ T cos_theta = cos(theta);
+ T sin_theta = sin(theta);
+
+ AT_ASSERTM(
+ roi_width >= 0 && roi_height >= 0,
+ "ROIs in ROIAlignRotated do not have non-negative size!");
+
+ T bin_size_h = static_cast(roi_height) / static_cast(pooled_height);
+ T bin_size_w = static_cast(roi_width) / static_cast(pooled_width);
+
+ // We use roi_bin_grid to sample the grid and mimic integral
+ int roi_bin_grid_h = (sampling_ratio > 0)
+ ? sampling_ratio
+ : ceil(roi_height / pooled_height); // e.g., = 2
+ int roi_bin_grid_w =
+ (sampling_ratio > 0) ? sampling_ratio : ceil(roi_width / pooled_width);
+
+ // We do average (integral) pooling inside a bin
+ const T count = std::max(roi_bin_grid_h * roi_bin_grid_w, 1); // e.g. = 4
+
+ // we want to precalculate indices and weights shared by all channels,
+ // this is the key point of optimization
+ std::vector> pre_calc(
+ roi_bin_grid_h * roi_bin_grid_w * pooled_width * pooled_height);
+
+ // roi_start_h and roi_start_w are computed wrt the center of RoI (x, y).
+ // Appropriate translation needs to be applied after.
+ T roi_start_h = -roi_height / 2.0;
+ T roi_start_w = -roi_width / 2.0;
+
+ pre_calc_for_bilinear_interpolate(
+ height,
+ width,
+ pooled_height,
+ pooled_width,
+ roi_bin_grid_h,
+ roi_bin_grid_w,
+ roi_start_h,
+ roi_start_w,
+ bin_size_h,
+ bin_size_w,
+ roi_bin_grid_h,
+ roi_bin_grid_w,
+ roi_center_h,
+ roi_center_w,
+ cos_theta,
+ sin_theta,
+ pre_calc);
+
+ for (int c = 0; c < channels; c++) {
+ int index_n_c = index_n + c * pooled_width * pooled_height;
+ const T* offset_input =
+ input + (roi_batch_ind * channels + c) * height * width;
+ int pre_calc_index = 0;
+
+ for (int ph = 0; ph < pooled_height; ph++) {
+ for (int pw = 0; pw < pooled_width; pw++) {
+ int index = index_n_c + ph * pooled_width + pw;
+
+ T output_val = 0.;
+ for (int iy = 0; iy < roi_bin_grid_h; iy++) {
+ for (int ix = 0; ix < roi_bin_grid_w; ix++) {
+ PreCalc pc = pre_calc[pre_calc_index];
+ output_val += pc.w1 * offset_input[pc.pos1] +
+ pc.w2 * offset_input[pc.pos2] +
+ pc.w3 * offset_input[pc.pos3] + pc.w4 * offset_input[pc.pos4];
+
+ pre_calc_index += 1;
+ }
+ }
+ output_val /= count;
+
+ output[index] = output_val;
+ } // for pw
+ } // for ph
+ } // for c
+ } // for n
+}
+
+template
+void ROIAlignRotatedBackward(
+ const int nthreads,
+ // may not be contiguous. should index using n_stride, etc
+ const T* grad_output,
+ const T& spatial_scale,
+ const int channels,
+ const int height,
+ const int width,
+ const int pooled_height,
+ const int pooled_width,
+ const int sampling_ratio,
+ T* grad_input,
+ const T* rois,
+ const int n_stride,
+ const int c_stride,
+ const int h_stride,
+ const int w_stride) {
+ for (int index = 0; index < nthreads; index++) {
+ // (n, c, ph, pw) is an element in the pooled output
+ int pw = index % pooled_width;
+ int ph = (index / pooled_width) % pooled_height;
+ int c = (index / pooled_width / pooled_height) % channels;
+ int n = index / pooled_width / pooled_height / channels;
+
+ const T* current_roi = rois + n * 6;
+ int roi_batch_ind = current_roi[0];
+
+ // Do not use rounding; this implementation detail is critical
+ // ROIAlignRotated supports align == true, i.e., continuous coordinate
+ // by default, thus the 0.5 offset
+ T offset = (T)0.5;
+ T roi_center_w = current_roi[1] * spatial_scale - offset;
+ T roi_center_h = current_roi[2] * spatial_scale - offset;
+ T roi_width = current_roi[3] * spatial_scale;
+ T roi_height = current_roi[4] * spatial_scale;
+ T theta = current_roi[5] * M_PI / 180.0;
+ T cos_theta = cos(theta);
+ T sin_theta = sin(theta);
+
+ AT_ASSERTM(
+ roi_width >= 0 && roi_height >= 0,
+ "ROIs in ROIAlignRotated do not have non-negative size!");
+
+ T bin_size_h = static_cast(roi_height) / static_cast(pooled_height);
+ T bin_size_w = static_cast(roi_width) / static_cast(pooled_width);
+
+ T* offset_grad_input =
+ grad_input + ((roi_batch_ind * channels + c) * height * width);
+
+ int output_offset = n * n_stride + c * c_stride;
+ const T* offset_grad_output = grad_output + output_offset;
+ const T grad_output_this_bin =
+ offset_grad_output[ph * h_stride + pw * w_stride];
+
+ // We use roi_bin_grid to sample the grid and mimic integral
+ int roi_bin_grid_h = (sampling_ratio > 0)
+ ? sampling_ratio
+ : ceil(roi_height / pooled_height); // e.g., = 2
+ int roi_bin_grid_w =
+ (sampling_ratio > 0) ? sampling_ratio : ceil(roi_width / pooled_width);
+
+ // roi_start_h and roi_start_w are computed wrt the center of RoI (x, y).
+ // Appropriate translation needs to be applied after.
+ T roi_start_h = -roi_height / 2.0;
+ T roi_start_w = -roi_width / 2.0;
+
+ // We do average (integral) pooling inside a bin
+ const T count = roi_bin_grid_h * roi_bin_grid_w; // e.g. = 4
+
+ for (int iy = 0; iy < roi_bin_grid_h; iy++) {
+ const T yy = roi_start_h + ph * bin_size_h +
+ static_cast(iy + .5f) * bin_size_h /
+ static_cast(roi_bin_grid_h); // e.g., 0.5, 1.5
+ for (int ix = 0; ix < roi_bin_grid_w; ix++) {
+ const T xx = roi_start_w + pw * bin_size_w +
+ static_cast(ix + .5f) * bin_size_w /
+ static_cast(roi_bin_grid_w);
+
+ // Rotate by theta around the center and translate
+ T y = yy * cos_theta - xx * sin_theta + roi_center_h;
+ T x = yy * sin_theta + xx * cos_theta + roi_center_w;
+
+ T w1, w2, w3, w4;
+ int x_low, x_high, y_low, y_high;
+
+ bilinear_interpolate_gradient(
+ height, width, y, x, w1, w2, w3, w4, x_low, x_high, y_low, y_high);
+
+ T g1 = grad_output_this_bin * w1 / count;
+ T g2 = grad_output_this_bin * w2 / count;
+ T g3 = grad_output_this_bin * w3 / count;
+ T g4 = grad_output_this_bin * w4 / count;
+
+ if (x_low >= 0 && x_high >= 0 && y_low >= 0 && y_high >= 0) {
+ // atomic add is not needed for now since it is single threaded
+ add(offset_grad_input + y_low * width + x_low, static_cast(g1));
+ add(offset_grad_input + y_low * width + x_high, static_cast(g2));
+ add(offset_grad_input + y_high * width + x_low, static_cast(g3));
+ add(offset_grad_input + y_high * width + x_high, static_cast(g4));
+ } // if
+ } // ix
+ } // iy
+ } // for
+} // ROIAlignRotatedBackward
+
+at::Tensor ROIAlignRotated_forward_cpu(
+ const at::Tensor& input,
+ const at::Tensor& rois,
+ const float spatial_scale,
+ const int pooled_height,
+ const int pooled_width,
+ const int sampling_ratio) {
+ AT_ASSERTM(input.device().is_cpu(), "input must be a CPU tensor");
+ AT_ASSERTM(rois.device().is_cpu(), "rois must be a CPU tensor");
+
+ at::TensorArg input_t{input, "input", 1}, rois_t{rois, "rois", 2};
+
+ at::CheckedFrom c = "ROIAlign_forward_cpu";
+ at::checkAllSameType(c, {input_t, rois_t});
+
+ auto num_rois = rois.size(0);
+ auto channels = input.size(1);
+ auto height = input.size(2);
+ auto width = input.size(3);
+
+ at::Tensor output = at::zeros(
+ {num_rois, channels, pooled_height, pooled_width}, input.options());
+
+ auto output_size = num_rois * pooled_height * pooled_width * channels;
+
+ if (output.numel() == 0) {
+ return output;
+ }
+
+ auto input_ = input.contiguous(), rois_ = rois.contiguous();
+ AT_DISPATCH_FLOATING_TYPES_AND_HALF(
+ input.scalar_type(), "ROIAlignRotated_forward", [&] {
+ ROIAlignRotatedForward(
+ output_size,
+ input_.data_ptr(),
+ spatial_scale,
+ channels,
+ height,
+ width,
+ pooled_height,
+ pooled_width,
+ sampling_ratio,
+ rois_.data_ptr(),
+ output.data_ptr());
+ });
+ return output;
+}
+
+at::Tensor ROIAlignRotated_backward_cpu(
+ const at::Tensor& grad,
+ const at::Tensor& rois,
+ const float spatial_scale,
+ const int pooled_height,
+ const int pooled_width,
+ const int batch_size,
+ const int channels,
+ const int height,
+ const int width,
+ const int sampling_ratio) {
+ AT_ASSERTM(grad.device().is_cpu(), "grad must be a CPU tensor");
+ AT_ASSERTM(rois.device().is_cpu(), "rois must be a CPU tensor");
+
+ at::TensorArg grad_t{grad, "grad", 1}, rois_t{rois, "rois", 2};
+
+ at::CheckedFrom c = "ROIAlignRotated_backward_cpu";
+ at::checkAllSameType(c, {grad_t, rois_t});
+
+ at::Tensor grad_input =
+ at::zeros({batch_size, channels, height, width}, grad.options());
+
+ // handle possibly empty gradients
+ if (grad.numel() == 0) {
+ return grad_input;
+ }
+
+ // get stride values to ensure indexing into gradients is correct.
+ int n_stride = grad.stride(0);
+ int c_stride = grad.stride(1);
+ int h_stride = grad.stride(2);
+ int w_stride = grad.stride(3);
+
+ auto rois_ = rois.contiguous();
+ AT_DISPATCH_FLOATING_TYPES_AND_HALF(
+ grad.scalar_type(), "ROIAlignRotated_forward", [&] {
+ ROIAlignRotatedBackward(
+ grad.numel(),
+ grad.data_ptr(),
+ spatial_scale,
+ channels,
+ height,
+ width,
+ pooled_height,
+ pooled_width,
+ sampling_ratio,
+ grad_input.data_ptr(),
+ rois_.data_ptr(),
+ n_stride,
+ c_stride,
+ h_stride,
+ w_stride);
+ });
+ return grad_input;
+}
+
+} // namespace detectron2
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/ROIAlignRotated/ROIAlignRotated_cuda.cu b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/ROIAlignRotated/ROIAlignRotated_cuda.cu
new file mode 100644
index 0000000000000000000000000000000000000000..fca186519143b168a912c880a4cf495a0a5a9322
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/ROIAlignRotated/ROIAlignRotated_cuda.cu
@@ -0,0 +1,443 @@
+// Copyright (c) Facebook, Inc. and its affiliates.
+#include
+#include
+#include
+#include
+
+// TODO make it in a common file
+#define CUDA_1D_KERNEL_LOOP(i, n) \
+ for (int i = blockIdx.x * blockDim.x + threadIdx.x; i < n; \
+ i += blockDim.x * gridDim.x)
+
+// Note: this implementation originates from the Caffe2 ROIAlignRotated Op
+// and PyTorch ROIAlign (non-rotated) Op implementations.
+// The key difference between this implementation and those ones is
+// we don't do "legacy offset" in this version, as there aren't many previous
+// works, if any, using the "legacy" ROIAlignRotated Op.
+// This would make the interface a bit cleaner.
+
+namespace detectron2 {
+
+namespace {
+
+template
+__device__ T bilinear_interpolate(
+ const T* input,
+ const int height,
+ const int width,
+ T y,
+ T x) {
+ // deal with cases that inverse elements are out of feature map boundary
+ if (y < -1.0 || y > height || x < -1.0 || x > width) {
+ // empty
+ return 0;
+ }
+
+ if (y < 0) {
+ y = 0;
+ }
+
+ if (x < 0) {
+ x = 0;
+ }
+
+ int y_low = (int)y;
+ int x_low = (int)x;
+ int y_high;
+ int x_high;
+
+ if (y_low >= height - 1) {
+ y_high = y_low = height - 1;
+ y = (T)y_low;
+ } else {
+ y_high = y_low + 1;
+ }
+
+ if (x_low >= width - 1) {
+ x_high = x_low = width - 1;
+ x = (T)x_low;
+ } else {
+ x_high = x_low + 1;
+ }
+
+ T ly = y - y_low;
+ T lx = x - x_low;
+ T hy = 1. - ly, hx = 1. - lx;
+ // do bilinear interpolation
+ T v1 = input[y_low * width + x_low];
+ T v2 = input[y_low * width + x_high];
+ T v3 = input[y_high * width + x_low];
+ T v4 = input[y_high * width + x_high];
+ T w1 = hy * hx, w2 = hy * lx, w3 = ly * hx, w4 = ly * lx;
+
+ T val = (w1 * v1 + w2 * v2 + w3 * v3 + w4 * v4);
+
+ return val;
+}
+
+template
+__device__ void bilinear_interpolate_gradient(
+ const int height,
+ const int width,
+ T y,
+ T x,
+ T& w1,
+ T& w2,
+ T& w3,
+ T& w4,
+ int& x_low,
+ int& x_high,
+ int& y_low,
+ int& y_high) {
+ // deal with cases that inverse elements are out of feature map boundary
+ if (y < -1.0 || y > height || x < -1.0 || x > width) {
+ // empty
+ w1 = w2 = w3 = w4 = 0.;
+ x_low = x_high = y_low = y_high = -1;
+ return;
+ }
+
+ if (y < 0) {
+ y = 0;
+ }
+
+ if (x < 0) {
+ x = 0;
+ }
+
+ y_low = (int)y;
+ x_low = (int)x;
+
+ if (y_low >= height - 1) {
+ y_high = y_low = height - 1;
+ y = (T)y_low;
+ } else {
+ y_high = y_low + 1;
+ }
+
+ if (x_low >= width - 1) {
+ x_high = x_low = width - 1;
+ x = (T)x_low;
+ } else {
+ x_high = x_low + 1;
+ }
+
+ T ly = y - y_low;
+ T lx = x - x_low;
+ T hy = 1. - ly, hx = 1. - lx;
+
+ // reference in forward
+ // T v1 = input[y_low * width + x_low];
+ // T v2 = input[y_low * width + x_high];
+ // T v3 = input[y_high * width + x_low];
+ // T v4 = input[y_high * width + x_high];
+ // T val = (w1 * v1 + w2 * v2 + w3 * v3 + w4 * v4);
+
+ w1 = hy * hx, w2 = hy * lx, w3 = ly * hx, w4 = ly * lx;
+
+ return;
+}
+
+} // namespace
+
+template
+__global__ void RoIAlignRotatedForward(
+ const int nthreads,
+ const T* input,
+ const T spatial_scale,
+ const int channels,
+ const int height,
+ const int width,
+ const int pooled_height,
+ const int pooled_width,
+ const int sampling_ratio,
+ const T* rois,
+ T* top_data) {
+ CUDA_1D_KERNEL_LOOP(index, nthreads) {
+ // (n, c, ph, pw) is an element in the pooled output
+ int pw = index % pooled_width;
+ int ph = (index / pooled_width) % pooled_height;
+ int c = (index / pooled_width / pooled_height) % channels;
+ int n = index / pooled_width / pooled_height / channels;
+
+ const T* current_roi = rois + n * 6;
+ int roi_batch_ind = current_roi[0];
+
+ // Do not use rounding; this implementation detail is critical
+ // ROIAlignRotated supports align == true, i.e., continuous coordinate
+ // by default, thus the 0.5 offset
+ T offset = (T)0.5;
+ T roi_center_w = current_roi[1] * spatial_scale - offset;
+ T roi_center_h = current_roi[2] * spatial_scale - offset;
+ T roi_width = current_roi[3] * spatial_scale;
+ T roi_height = current_roi[4] * spatial_scale;
+ T theta = current_roi[5] * M_PI / 180.0;
+ T cos_theta = cos(theta);
+ T sin_theta = sin(theta);
+
+ T bin_size_h = static_cast(roi_height) / static_cast(pooled_height);
+ T bin_size_w = static_cast(roi_width) / static_cast(pooled_width);
+
+ const T* offset_input =
+ input + (roi_batch_ind * channels + c) * height * width;
+
+ // We use roi_bin_grid to sample the grid and mimic integral
+ int roi_bin_grid_h = (sampling_ratio > 0)
+ ? sampling_ratio
+ : ceil(roi_height / pooled_height); // e.g., = 2
+ int roi_bin_grid_w =
+ (sampling_ratio > 0) ? sampling_ratio : ceil(roi_width / pooled_width);
+
+ // roi_start_h and roi_start_w are computed wrt the center of RoI (x, y).
+ // Appropriate translation needs to be applied after.
+ T roi_start_h = -roi_height / 2.0;
+ T roi_start_w = -roi_width / 2.0;
+
+ // We do average (inte gral) pooling inside a bin
+ const T count = max(roi_bin_grid_h * roi_bin_grid_w, 1); // e.g. = 4
+
+ T output_val = 0.;
+ for (int iy = 0; iy < roi_bin_grid_h; iy++) // e.g., iy = 0, 1
+ {
+ const T yy = roi_start_h + ph * bin_size_h +
+ static_cast(iy + .5f) * bin_size_h /
+ static_cast(roi_bin_grid_h); // e.g., 0.5, 1.5
+ for (int ix = 0; ix < roi_bin_grid_w; ix++) {
+ const T xx = roi_start_w + pw * bin_size_w +
+ static_cast(ix + .5f) * bin_size_w /
+ static_cast(roi_bin_grid_w);
+
+ // Rotate by theta around the center and translate
+ T y = yy * cos_theta - xx * sin_theta + roi_center_h;
+ T x = yy * sin_theta + xx * cos_theta + roi_center_w;
+
+ T val = bilinear_interpolate(offset_input, height, width, y, x);
+ output_val += val;
+ }
+ }
+ output_val /= count;
+
+ top_data[index] = output_val;
+ }
+}
+
+template
+__global__ void RoIAlignRotatedBackwardFeature(
+ const int nthreads,
+ const T* top_diff,
+ const int num_rois,
+ const T spatial_scale,
+ const int channels,
+ const int height,
+ const int width,
+ const int pooled_height,
+ const int pooled_width,
+ const int sampling_ratio,
+ T* bottom_diff,
+ const T* rois) {
+ CUDA_1D_KERNEL_LOOP(index, nthreads) {
+ // (n, c, ph, pw) is an element in the pooled output
+ int pw = index % pooled_width;
+ int ph = (index / pooled_width) % pooled_height;
+ int c = (index / pooled_width / pooled_height) % channels;
+ int n = index / pooled_width / pooled_height / channels;
+
+ const T* current_roi = rois + n * 6;
+ int roi_batch_ind = current_roi[0];
+
+ // Do not use rounding; this implementation detail is critical
+ // ROIAlignRotated supports align == true, i.e., continuous coordinate
+ // by default, thus the 0.5 offset
+ T offset = (T)0.5;
+ T roi_center_w = current_roi[1] * spatial_scale - offset;
+ T roi_center_h = current_roi[2] * spatial_scale - offset;
+ T roi_width = current_roi[3] * spatial_scale;
+ T roi_height = current_roi[4] * spatial_scale;
+ T theta = current_roi[5] * M_PI / 180.0;
+ T cos_theta = cos(theta);
+ T sin_theta = sin(theta);
+
+ T bin_size_h = static_cast(roi_height) / static_cast(pooled_height);
+ T bin_size_w = static_cast(roi_width) / static_cast(pooled_width);
+
+ T* offset_bottom_diff =
+ bottom_diff + (roi_batch_ind * channels + c) * height * width;
+
+ int top_offset = (n * channels + c) * pooled_height * pooled_width;
+ const T* offset_top_diff = top_diff + top_offset;
+ const T top_diff_this_bin = offset_top_diff[ph * pooled_width + pw];
+
+ // We use roi_bin_grid to sample the grid and mimic integral
+ int roi_bin_grid_h = (sampling_ratio > 0)
+ ? sampling_ratio
+ : ceil(roi_height / pooled_height); // e.g., = 2
+ int roi_bin_grid_w =
+ (sampling_ratio > 0) ? sampling_ratio : ceil(roi_width / pooled_width);
+
+ // roi_start_h and roi_start_w are computed wrt the center of RoI (x, y).
+ // Appropriate translation needs to be applied after.
+ T roi_start_h = -roi_height / 2.0;
+ T roi_start_w = -roi_width / 2.0;
+
+ // We do average (integral) pooling inside a bin
+ const T count = roi_bin_grid_h * roi_bin_grid_w; // e.g. = 4
+
+ for (int iy = 0; iy < roi_bin_grid_h; iy++) // e.g., iy = 0, 1
+ {
+ const T yy = roi_start_h + ph * bin_size_h +
+ static_cast(iy + .5f) * bin_size_h /
+ static_cast(roi_bin_grid_h); // e.g., 0.5, 1.5
+ for (int ix = 0; ix < roi_bin_grid_w; ix++) {
+ const T xx = roi_start_w + pw * bin_size_w +
+ static_cast(ix + .5f) * bin_size_w /
+ static_cast(roi_bin_grid_w);
+
+ // Rotate by theta around the center and translate
+ T y = yy * cos_theta - xx * sin_theta + roi_center_h;
+ T x = yy * sin_theta + xx * cos_theta + roi_center_w;
+
+ T w1, w2, w3, w4;
+ int x_low, x_high, y_low, y_high;
+
+ bilinear_interpolate_gradient(
+ height, width, y, x, w1, w2, w3, w4, x_low, x_high, y_low, y_high);
+
+ T g1 = top_diff_this_bin * w1 / count;
+ T g2 = top_diff_this_bin * w2 / count;
+ T g3 = top_diff_this_bin * w3 / count;
+ T g4 = top_diff_this_bin * w4 / count;
+
+ if (x_low >= 0 && x_high >= 0 && y_low >= 0 && y_high >= 0) {
+ atomicAdd(
+ offset_bottom_diff + y_low * width + x_low, static_cast(g1));
+ atomicAdd(
+ offset_bottom_diff + y_low * width + x_high, static_cast(g2));
+ atomicAdd(
+ offset_bottom_diff + y_high * width + x_low, static_cast(g3));
+ atomicAdd(
+ offset_bottom_diff + y_high * width + x_high, static_cast(g4));
+ } // if
+ } // ix
+ } // iy
+ } // CUDA_1D_KERNEL_LOOP
+} // RoIAlignRotatedBackward
+
+at::Tensor ROIAlignRotated_forward_cuda(
+ const at::Tensor& input,
+ const at::Tensor& rois,
+ const float spatial_scale,
+ const int pooled_height,
+ const int pooled_width,
+ const int sampling_ratio) {
+ AT_ASSERTM(input.device().is_cuda(), "input must be a CUDA tensor");
+ AT_ASSERTM(rois.device().is_cuda(), "rois must be a CUDA tensor");
+ at::TensorArg input_t{input, "input", 1}, rois_t{rois, "rois", 2};
+
+ at::CheckedFrom c = "ROIAlignRotated_forward_cuda";
+ at::checkAllSameGPU(c, {input_t, rois_t});
+ at::checkAllSameType(c, {input_t, rois_t});
+ at::cuda::CUDAGuard device_guard(input.device());
+
+ auto num_rois = rois.size(0);
+ auto channels = input.size(1);
+ auto height = input.size(2);
+ auto width = input.size(3);
+
+ auto output = at::empty(
+ {num_rois, channels, pooled_height, pooled_width}, input.options());
+ auto output_size = num_rois * pooled_height * pooled_width * channels;
+ cudaStream_t stream = at::cuda::getCurrentCUDAStream();
+
+ dim3 grid(std::min(
+ at::cuda::ATenCeilDiv(
+ static_cast(output_size), static_cast(512)),
+ static_cast(4096)));
+ dim3 block(512);
+
+ if (output.numel() == 0) {
+ AT_CUDA_CHECK(cudaGetLastError());
+ return output;
+ }
+
+ auto input_ = input.contiguous(), rois_ = rois.contiguous();
+ AT_DISPATCH_FLOATING_TYPES(
+ input.scalar_type(), "ROIAlignRotated_forward", [&] {
+ RoIAlignRotatedForward<<>>(
+ output_size,
+ input_.data_ptr(),
+ spatial_scale,
+ channels,
+ height,
+ width,
+ pooled_height,
+ pooled_width,
+ sampling_ratio,
+ rois_.data_ptr(),
+ output.data_ptr());
+ });
+ cudaDeviceSynchronize();
+ AT_CUDA_CHECK(cudaGetLastError());
+ return output;
+}
+
+// TODO remove the dependency on input and use instead its sizes -> save memory
+at::Tensor ROIAlignRotated_backward_cuda(
+ const at::Tensor& grad,
+ const at::Tensor& rois,
+ const float spatial_scale,
+ const int pooled_height,
+ const int pooled_width,
+ const int batch_size,
+ const int channels,
+ const int height,
+ const int width,
+ const int sampling_ratio) {
+ AT_ASSERTM(grad.device().is_cuda(), "grad must be a CUDA tensor");
+ AT_ASSERTM(rois.device().is_cuda(), "rois must be a CUDA tensor");
+
+ at::TensorArg grad_t{grad, "grad", 1}, rois_t{rois, "rois", 2};
+ at::CheckedFrom c = "ROIAlign_backward_cuda";
+ at::checkAllSameGPU(c, {grad_t, rois_t});
+ at::checkAllSameType(c, {grad_t, rois_t});
+ at::cuda::CUDAGuard device_guard(grad.device());
+
+ auto num_rois = rois.size(0);
+ auto grad_input =
+ at::zeros({batch_size, channels, height, width}, grad.options());
+
+ cudaStream_t stream = at::cuda::getCurrentCUDAStream();
+
+ dim3 grid(std::min(
+ at::cuda::ATenCeilDiv(
+ static_cast(grad.numel()), static_cast(512)),
+ static_cast(4096)));
+ dim3 block(512);
+
+ // handle possibly empty gradients
+ if (grad.numel() == 0) {
+ AT_CUDA_CHECK(cudaGetLastError());
+ return grad_input;
+ }
+
+ auto grad_ = grad.contiguous(), rois_ = rois.contiguous();
+ AT_DISPATCH_FLOATING_TYPES(
+ grad.scalar_type(), "ROIAlignRotated_backward", [&] {
+ RoIAlignRotatedBackwardFeature<<>>(
+ grad.numel(),
+ grad_.data_ptr(),
+ num_rois,
+ spatial_scale,
+ channels,
+ height,
+ width,
+ pooled_height,
+ pooled_width,
+ sampling_ratio,
+ grad_input.data_ptr(),
+ rois_.data_ptr());
+ });
+ AT_CUDA_CHECK(cudaGetLastError());
+ return grad_input;
+}
+
+} // namespace detectron2
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/box_iou_rotated/box_iou_rotated.h b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/box_iou_rotated/box_iou_rotated.h
new file mode 100644
index 0000000000000000000000000000000000000000..3bf383b8ed9b358b5313d433a9682c294dfb77e4
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/box_iou_rotated/box_iou_rotated.h
@@ -0,0 +1,35 @@
+// Copyright (c) Facebook, Inc. and its affiliates.
+#pragma once
+#include
+
+namespace detectron2 {
+
+at::Tensor box_iou_rotated_cpu(
+ const at::Tensor& boxes1,
+ const at::Tensor& boxes2);
+
+#if defined(WITH_CUDA) || defined(WITH_HIP)
+at::Tensor box_iou_rotated_cuda(
+ const at::Tensor& boxes1,
+ const at::Tensor& boxes2);
+#endif
+
+// Interface for Python
+// inline is needed to prevent multiple function definitions when this header is
+// included by different cpps
+inline at::Tensor box_iou_rotated(
+ const at::Tensor& boxes1,
+ const at::Tensor& boxes2) {
+ assert(boxes1.device().is_cuda() == boxes2.device().is_cuda());
+ if (boxes1.device().is_cuda()) {
+#if defined(WITH_CUDA) || defined(WITH_HIP)
+ return box_iou_rotated_cuda(boxes1.contiguous(), boxes2.contiguous());
+#else
+ AT_ERROR("Detectron2 is not compiled with GPU support!");
+#endif
+ }
+
+ return box_iou_rotated_cpu(boxes1.contiguous(), boxes2.contiguous());
+}
+
+} // namespace detectron2
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/box_iou_rotated/box_iou_rotated_cpu.cpp b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/box_iou_rotated/box_iou_rotated_cpu.cpp
new file mode 100644
index 0000000000000000000000000000000000000000..c843487b5fa4e8077dd27402ec99009266ddda8d
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/box_iou_rotated/box_iou_rotated_cpu.cpp
@@ -0,0 +1,39 @@
+// Copyright (c) Facebook, Inc. and its affiliates.
+#include "box_iou_rotated.h"
+#include "box_iou_rotated_utils.h"
+
+namespace detectron2 {
+
+template
+void box_iou_rotated_cpu_kernel(
+ const at::Tensor& boxes1,
+ const at::Tensor& boxes2,
+ at::Tensor& ious) {
+ auto num_boxes1 = boxes1.size(0);
+ auto num_boxes2 = boxes2.size(0);
+
+ for (int i = 0; i < num_boxes1; i++) {
+ for (int j = 0; j < num_boxes2; j++) {
+ ious[i * num_boxes2 + j] = single_box_iou_rotated(
+ boxes1[i].data_ptr(), boxes2[j].data_ptr());
+ }
+ }
+}
+
+at::Tensor box_iou_rotated_cpu(
+ // input must be contiguous:
+ const at::Tensor& boxes1,
+ const at::Tensor& boxes2) {
+ auto num_boxes1 = boxes1.size(0);
+ auto num_boxes2 = boxes2.size(0);
+ at::Tensor ious =
+ at::empty({num_boxes1 * num_boxes2}, boxes1.options().dtype(at::kFloat));
+
+ box_iou_rotated_cpu_kernel(boxes1, boxes2, ious);
+
+ // reshape from 1d array to 2d array
+ auto shape = std::vector{num_boxes1, num_boxes2};
+ return ious.reshape(shape);
+}
+
+} // namespace detectron2
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/box_iou_rotated/box_iou_rotated_cuda.cu b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/box_iou_rotated/box_iou_rotated_cuda.cu
new file mode 100644
index 0000000000000000000000000000000000000000..952710e53041187907fbd113f8d0d0fa24134a86
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/box_iou_rotated/box_iou_rotated_cuda.cu
@@ -0,0 +1,130 @@
+// Copyright (c) Facebook, Inc. and its affiliates.
+#include
+#include
+#include
+#include
+#include "box_iou_rotated_utils.h"
+
+namespace detectron2 {
+
+// 2D block with 32 * 16 = 512 threads per block
+const int BLOCK_DIM_X = 32;
+const int BLOCK_DIM_Y = 16;
+
+template
+__global__ void box_iou_rotated_cuda_kernel(
+ const int n_boxes1,
+ const int n_boxes2,
+ const T* dev_boxes1,
+ const T* dev_boxes2,
+ T* dev_ious) {
+ const int row_start = blockIdx.x * blockDim.x;
+ const int col_start = blockIdx.y * blockDim.y;
+
+ const int row_size = min(n_boxes1 - row_start, blockDim.x);
+ const int col_size = min(n_boxes2 - col_start, blockDim.y);
+
+ __shared__ float block_boxes1[BLOCK_DIM_X * 5];
+ __shared__ float block_boxes2[BLOCK_DIM_Y * 5];
+
+ // It's safe to copy using threadIdx.x since BLOCK_DIM_X >= BLOCK_DIM_Y
+ if (threadIdx.x < row_size && threadIdx.y == 0) {
+ block_boxes1[threadIdx.x * 5 + 0] =
+ dev_boxes1[(row_start + threadIdx.x) * 5 + 0];
+ block_boxes1[threadIdx.x * 5 + 1] =
+ dev_boxes1[(row_start + threadIdx.x) * 5 + 1];
+ block_boxes1[threadIdx.x * 5 + 2] =
+ dev_boxes1[(row_start + threadIdx.x) * 5 + 2];
+ block_boxes1[threadIdx.x * 5 + 3] =
+ dev_boxes1[(row_start + threadIdx.x) * 5 + 3];
+ block_boxes1[threadIdx.x * 5 + 4] =
+ dev_boxes1[(row_start + threadIdx.x) * 5 + 4];
+ }
+
+ if (threadIdx.x < col_size && threadIdx.y == 0) {
+ block_boxes2[threadIdx.x * 5 + 0] =
+ dev_boxes2[(col_start + threadIdx.x) * 5 + 0];
+ block_boxes2[threadIdx.x * 5 + 1] =
+ dev_boxes2[(col_start + threadIdx.x) * 5 + 1];
+ block_boxes2[threadIdx.x * 5 + 2] =
+ dev_boxes2[(col_start + threadIdx.x) * 5 + 2];
+ block_boxes2[threadIdx.x * 5 + 3] =
+ dev_boxes2[(col_start + threadIdx.x) * 5 + 3];
+ block_boxes2[threadIdx.x * 5 + 4] =
+ dev_boxes2[(col_start + threadIdx.x) * 5 + 4];
+ }
+ __syncthreads();
+
+ if (threadIdx.x < row_size && threadIdx.y < col_size) {
+ int offset = (row_start + threadIdx.x) * n_boxes2 + col_start + threadIdx.y;
+ dev_ious[offset] = single_box_iou_rotated(
+ block_boxes1 + threadIdx.x * 5, block_boxes2 + threadIdx.y * 5);
+ }
+}
+
+at::Tensor box_iou_rotated_cuda(
+ // input must be contiguous
+ const at::Tensor& boxes1,
+ const at::Tensor& boxes2) {
+ using scalar_t = float;
+ AT_ASSERTM(
+ boxes1.scalar_type() == at::kFloat, "boxes1 must be a float tensor");
+ AT_ASSERTM(
+ boxes2.scalar_type() == at::kFloat, "boxes2 must be a float tensor");
+ AT_ASSERTM(boxes1.is_cuda(), "boxes1 must be a CUDA tensor");
+ AT_ASSERTM(boxes2.is_cuda(), "boxes2 must be a CUDA tensor");
+ at::cuda::CUDAGuard device_guard(boxes1.device());
+
+ auto num_boxes1 = boxes1.size(0);
+ auto num_boxes2 = boxes2.size(0);
+
+ at::Tensor ious =
+ at::empty({num_boxes1 * num_boxes2}, boxes1.options().dtype(at::kFloat));
+
+ bool transpose = false;
+ if (num_boxes1 > 0 && num_boxes2 > 0) {
+ scalar_t *data1 = boxes1.data_ptr(),
+ *data2 = boxes2.data_ptr();
+
+ if (num_boxes2 > 65535 * BLOCK_DIM_Y) {
+ AT_ASSERTM(
+ num_boxes1 <= 65535 * BLOCK_DIM_Y,
+ "Too many boxes for box_iou_rotated_cuda!");
+ // x dim is allowed to be large, but y dim cannot,
+ // so we transpose the two to avoid "invalid configuration argument"
+ // error. We assume one of them is small. Otherwise the result is hard to
+ // fit in memory anyway.
+ std::swap(num_boxes1, num_boxes2);
+ std::swap(data1, data2);
+ transpose = true;
+ }
+
+ const int blocks_x =
+ at::cuda::ATenCeilDiv(static_cast(num_boxes1), BLOCK_DIM_X);
+ const int blocks_y =
+ at::cuda::ATenCeilDiv(static_cast(num_boxes2), BLOCK_DIM_Y);
+
+ dim3 blocks(blocks_x, blocks_y);
+ dim3 threads(BLOCK_DIM_X, BLOCK_DIM_Y);
+ cudaStream_t stream = at::cuda::getCurrentCUDAStream();
+
+ box_iou_rotated_cuda_kernel<<>>(
+ num_boxes1,
+ num_boxes2,
+ data1,
+ data2,
+ (scalar_t*)ious.data_ptr());
+
+ AT_CUDA_CHECK(cudaGetLastError());
+ }
+
+ // reshape from 1d array to 2d array
+ auto shape = std::vector{num_boxes1, num_boxes2};
+ if (transpose) {
+ return ious.view(shape).t();
+ } else {
+ return ious.view(shape);
+ }
+}
+
+} // namespace detectron2
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/box_iou_rotated/box_iou_rotated_utils.h b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/box_iou_rotated/box_iou_rotated_utils.h
new file mode 100644
index 0000000000000000000000000000000000000000..b54a5dde2ca11a74d29c4d8adb7fe1634f5baf9c
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/box_iou_rotated/box_iou_rotated_utils.h
@@ -0,0 +1,370 @@
+// Copyright (c) Facebook, Inc. and its affiliates.
+#pragma once
+
+#include
+#include
+
+#if defined(__CUDACC__) || __HCC__ == 1 || __HIP__ == 1
+// Designates functions callable from the host (CPU) and the device (GPU)
+#define HOST_DEVICE __host__ __device__
+#define HOST_DEVICE_INLINE HOST_DEVICE __forceinline__
+#else
+#include
+#define HOST_DEVICE
+#define HOST_DEVICE_INLINE HOST_DEVICE inline
+#endif
+
+namespace detectron2 {
+
+namespace {
+
+template
+struct RotatedBox {
+ T x_ctr, y_ctr, w, h, a;
+};
+
+template
+struct Point {
+ T x, y;
+ HOST_DEVICE_INLINE Point(const T& px = 0, const T& py = 0) : x(px), y(py) {}
+ HOST_DEVICE_INLINE Point operator+(const Point& p) const {
+ return Point(x + p.x, y + p.y);
+ }
+ HOST_DEVICE_INLINE Point& operator+=(const Point& p) {
+ x += p.x;
+ y += p.y;
+ return *this;
+ }
+ HOST_DEVICE_INLINE Point operator-(const Point& p) const {
+ return Point(x - p.x, y - p.y);
+ }
+ HOST_DEVICE_INLINE Point operator*(const T coeff) const {
+ return Point(x * coeff, y * coeff);
+ }
+};
+
+template
+HOST_DEVICE_INLINE T dot_2d(const Point& A, const Point& B) {
+ return A.x * B.x + A.y * B.y;
+}
+
+// R: result type. can be different from input type
+template
+HOST_DEVICE_INLINE R cross_2d(const Point& A, const Point& B) {
+ return static_cast(A.x) * static_cast(B.y) -
+ static_cast(B.x) * static_cast(A.y);
+}
+
+template
+HOST_DEVICE_INLINE void get_rotated_vertices(
+ const RotatedBox& box,
+ Point (&pts)[4]) {
+ // M_PI / 180. == 0.01745329251
+ double theta = box.a * 0.01745329251;
+ T cosTheta2 = (T)cos(theta) * 0.5f;
+ T sinTheta2 = (T)sin(theta) * 0.5f;
+
+ // y: top --> down; x: left --> right
+ pts[0].x = box.x_ctr + sinTheta2 * box.h + cosTheta2 * box.w;
+ pts[0].y = box.y_ctr + cosTheta2 * box.h - sinTheta2 * box.w;
+ pts[1].x = box.x_ctr - sinTheta2 * box.h + cosTheta2 * box.w;
+ pts[1].y = box.y_ctr - cosTheta2 * box.h - sinTheta2 * box.w;
+ pts[2].x = 2 * box.x_ctr - pts[0].x;
+ pts[2].y = 2 * box.y_ctr - pts[0].y;
+ pts[3].x = 2 * box.x_ctr - pts[1].x;
+ pts[3].y = 2 * box.y_ctr - pts[1].y;
+}
+
+template
+HOST_DEVICE_INLINE int get_intersection_points(
+ const Point (&pts1)[4],
+ const Point (&pts2)[4],
+ Point (&intersections)[24]) {
+ // Line vector
+ // A line from p1 to p2 is: p1 + (p2-p1)*t, t=[0,1]
+ Point vec1[4], vec2[4];
+ for (int i = 0; i < 4; i++) {
+ vec1[i] = pts1[(i + 1) % 4] - pts1[i];
+ vec2[i] = pts2[(i + 1) % 4] - pts2[i];
+ }
+
+ // When computing the intersection area, it doesn't hurt if we have
+ // more (duplicated/approximate) intersections/vertices than needed,
+ // while it can cause drastic difference if we miss an intersection/vertex.
+ // Therefore, we add an epsilon to relax the comparisons between
+ // the float point numbers that decide the intersection points.
+ double EPS = 1e-5;
+
+ // Line test - test all line combos for intersection
+ int num = 0; // number of intersections
+ for (int i = 0; i < 4; i++) {
+ for (int j = 0; j < 4; j++) {
+ // Solve for 2x2 Ax=b
+ T det = cross_2d(vec2[j], vec1[i]);
+
+ // This takes care of parallel lines
+ if (fabs(det) <= 1e-14) {
+ continue;
+ }
+
+ auto vec12 = pts2[j] - pts1[i];
+
+ T t1 = cross_2d(vec2[j], vec12) / det;
+ T t2 = cross_2d(vec1[i], vec12) / det;
+
+ if (t1 > -EPS && t1 < 1.0f + EPS && t2 > -EPS && t2 < 1.0f + EPS) {
+ intersections[num++] = pts1[i] + vec1[i] * t1;
+ }
+ }
+ }
+
+ // Check for vertices of rect1 inside rect2
+ {
+ const auto& AB = vec2[0];
+ const auto& DA = vec2[3];
+ auto ABdotAB = dot_2d(AB, AB);
+ auto ADdotAD = dot_2d(DA, DA);
+ for (int i = 0; i < 4; i++) {
+ // assume ABCD is the rectangle, and P is the point to be judged
+ // P is inside ABCD iff. P's projection on AB lies within AB
+ // and P's projection on AD lies within AD
+
+ auto AP = pts1[i] - pts2[0];
+
+ auto APdotAB = dot_2d(AP, AB);
+ auto APdotAD = -dot_2d(AP, DA);
+
+ if ((APdotAB > -EPS) && (APdotAD > -EPS) && (APdotAB < ABdotAB + EPS) &&
+ (APdotAD < ADdotAD + EPS)) {
+ intersections[num++] = pts1[i];
+ }
+ }
+ }
+
+ // Reverse the check - check for vertices of rect2 inside rect1
+ {
+ const auto& AB = vec1[0];
+ const auto& DA = vec1[3];
+ auto ABdotAB = dot_2d(AB, AB);
+ auto ADdotAD = dot_2d(DA, DA);
+ for (int i = 0; i < 4; i++) {
+ auto AP = pts2[i] - pts1[0];
+
+ auto APdotAB = dot_2d(AP, AB);
+ auto APdotAD = -dot_2d(AP, DA);
+
+ if ((APdotAB > -EPS) && (APdotAD > -EPS) && (APdotAB < ABdotAB + EPS) &&
+ (APdotAD < ADdotAD + EPS)) {
+ intersections[num++] = pts2[i];
+ }
+ }
+ }
+
+ return num;
+}
+
+template
+HOST_DEVICE_INLINE int convex_hull_graham(
+ const Point (&p)[24],
+ const int& num_in,
+ Point (&q)[24],
+ bool shift_to_zero = false) {
+ assert(num_in >= 2);
+
+ // Step 1:
+ // Find point with minimum y
+ // if more than 1 points have the same minimum y,
+ // pick the one with the minimum x.
+ int t = 0;
+ for (int i = 1; i < num_in; i++) {
+ if (p[i].y < p[t].y || (p[i].y == p[t].y && p[i].x < p[t].x)) {
+ t = i;
+ }
+ }
+ auto& start = p[t]; // starting point
+
+ // Step 2:
+ // Subtract starting point from every points (for sorting in the next step)
+ for (int i = 0; i < num_in; i++) {
+ q[i] = p[i] - start;
+ }
+
+ // Swap the starting point to position 0
+ auto tmp = q[0];
+ q[0] = q[t];
+ q[t] = tmp;
+
+ // Step 3:
+ // Sort point 1 ~ num_in according to their relative cross-product values
+ // (essentially sorting according to angles)
+ // If the angles are the same, sort according to their distance to origin
+ T dist[24];
+#if defined(__CUDACC__) || __HCC__ == 1 || __HIP__ == 1
+ // compute distance to origin before sort, and sort them together with the
+ // points
+ for (int i = 0; i < num_in; i++) {
+ dist[i] = dot_2d(q[i], q[i]);
+ }
+
+ // CUDA version
+ // In the future, we can potentially use thrust
+ // for sorting here to improve speed (though not guaranteed)
+ for (int i = 1; i < num_in - 1; i++) {
+ for (int j = i + 1; j < num_in; j++) {
+ T crossProduct = cross_2d(q[i], q[j]);
+ if ((crossProduct < -1e-6) ||
+ (fabs(crossProduct) < 1e-6 && dist[i] > dist[j])) {
+ auto q_tmp = q[i];
+ q[i] = q[j];
+ q[j] = q_tmp;
+ auto dist_tmp = dist[i];
+ dist[i] = dist[j];
+ dist[j] = dist_tmp;
+ }
+ }
+ }
+#else
+ // CPU version
+ std::sort(
+ q + 1, q + num_in, [](const Point& A, const Point& B) -> bool {
+ T temp = cross_2d(A, B);
+ if (fabs(temp) < 1e-6) {
+ return dot_2d(A, A) < dot_2d(B, B);
+ } else {
+ return temp > 0;
+ }
+ });
+ // compute distance to origin after sort, since the points are now different.
+ for (int i = 0; i < num_in; i++) {
+ dist[i] = dot_2d(q[i], q[i]);
+ }
+#endif
+
+ // Step 4:
+ // Make sure there are at least 2 points (that don't overlap with each other)
+ // in the stack
+ int k; // index of the non-overlapped second point
+ for (k = 1; k < num_in; k++) {
+ if (dist[k] > 1e-8) {
+ break;
+ }
+ }
+ if (k == num_in) {
+ // We reach the end, which means the convex hull is just one point
+ q[0] = p[t];
+ return 1;
+ }
+ q[1] = q[k];
+ int m = 2; // 2 points in the stack
+ // Step 5:
+ // Finally we can start the scanning process.
+ // When a non-convex relationship between the 3 points is found
+ // (either concave shape or duplicated points),
+ // we pop the previous point from the stack
+ // until the 3-point relationship is convex again, or
+ // until the stack only contains two points
+ for (int i = k + 1; i < num_in; i++) {
+ while (m > 1) {
+ auto q1 = q[i] - q[m - 2], q2 = q[m - 1] - q[m - 2];
+ // cross_2d() uses FMA and therefore computes round(round(q1.x*q2.y) -
+ // q2.x*q1.y) So it may not return 0 even when q1==q2. Therefore we
+ // compare round(q1.x*q2.y) and round(q2.x*q1.y) directly. (round means
+ // round to nearest floating point).
+ if (q1.x * q2.y >= q2.x * q1.y)
+ m--;
+ else
+ break;
+ }
+ // Using double also helps, but float can solve the issue for now.
+ // while (m > 1 && cross_2d(q[i] - q[m - 2], q[m - 1] - q[m - 2])
+ // >= 0) {
+ // m--;
+ // }
+ q[m++] = q[i];
+ }
+
+ // Step 6 (Optional):
+ // In general sense we need the original coordinates, so we
+ // need to shift the points back (reverting Step 2)
+ // But if we're only interested in getting the area/perimeter of the shape
+ // We can simply return.
+ if (!shift_to_zero) {
+ for (int i = 0; i < m; i++) {
+ q[i] += start;
+ }
+ }
+
+ return m;
+}
+
+template
+HOST_DEVICE_INLINE T polygon_area(const Point (&q)[24], const int& m) {
+ if (m <= 2) {
+ return 0;
+ }
+
+ T area = 0;
+ for (int i = 1; i < m - 1; i++) {
+ area += fabs(cross_2d(q[i] - q[0], q[i + 1] - q[0]));
+ }
+
+ return area / 2.0;
+}
+
+template
+HOST_DEVICE_INLINE T rotated_boxes_intersection(
+ const RotatedBox& box1,
+ const RotatedBox& box2) {
+ // There are up to 4 x 4 + 4 + 4 = 24 intersections (including dups) returned
+ // from rotated_rect_intersection_pts
+ Point intersectPts[24], orderedPts[24];
+
+ Point pts1[4];
+ Point pts2[4];
+ get_rotated_vertices(box1, pts1);
+ get_rotated_vertices(box2, pts2);
+
+ int num = get_intersection_points(pts1, pts2, intersectPts);
+
+ if (num <= 2) {
+ return 0.0;
+ }
+
+ // Convex Hull to order the intersection points in clockwise order and find
+ // the contour area.
+ int num_convex = convex_hull_graham(intersectPts, num, orderedPts, true);
+ return polygon_area(orderedPts, num_convex);
+}
+
+} // namespace
+
+template
+HOST_DEVICE_INLINE T
+single_box_iou_rotated(T const* const box1_raw, T const* const box2_raw) {
+ // shift center to the middle point to achieve higher precision in result
+ RotatedBox box1, box2;
+ auto center_shift_x = (box1_raw[0] + box2_raw[0]) / 2.0;
+ auto center_shift_y = (box1_raw[1] + box2_raw[1]) / 2.0;
+ box1.x_ctr = box1_raw[0] - center_shift_x;
+ box1.y_ctr = box1_raw[1] - center_shift_y;
+ box1.w = box1_raw[2];
+ box1.h = box1_raw[3];
+ box1.a = box1_raw[4];
+ box2.x_ctr = box2_raw[0] - center_shift_x;
+ box2.y_ctr = box2_raw[1] - center_shift_y;
+ box2.w = box2_raw[2];
+ box2.h = box2_raw[3];
+ box2.a = box2_raw[4];
+
+ T area1 = box1.w * box1.h;
+ T area2 = box2.w * box2.h;
+ if (area1 < 1e-14 || area2 < 1e-14) {
+ return 0.f;
+ }
+
+ T intersection = rotated_boxes_intersection(box1, box2);
+ T iou = intersection / (area1 + area2 - intersection);
+ return iou;
+}
+
+} // namespace detectron2
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/cocoeval/cocoeval.cpp b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/cocoeval/cocoeval.cpp
new file mode 100644
index 0000000000000000000000000000000000000000..0a5b7b907c06720fefc77b0dfd921b8ec3ecf2be
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/cocoeval/cocoeval.cpp
@@ -0,0 +1,507 @@
+// Copyright (c) Facebook, Inc. and its affiliates.
+#include "cocoeval.h"
+#include
+#include
+#include
+#include
+
+using namespace pybind11::literals;
+
+namespace detectron2 {
+
+namespace COCOeval {
+
+// Sort detections from highest score to lowest, such that
+// detection_instances[detection_sorted_indices[t]] >=
+// detection_instances[detection_sorted_indices[t+1]]. Use stable_sort to match
+// original COCO API
+void SortInstancesByDetectionScore(
+ const std::vector& detection_instances,
+ std::vector* detection_sorted_indices) {
+ detection_sorted_indices->resize(detection_instances.size());
+ std::iota(
+ detection_sorted_indices->begin(), detection_sorted_indices->end(), 0);
+ std::stable_sort(
+ detection_sorted_indices->begin(),
+ detection_sorted_indices->end(),
+ [&detection_instances](size_t j1, size_t j2) {
+ return detection_instances[j1].score > detection_instances[j2].score;
+ });
+}
+
+// Partition the ground truth objects based on whether or not to ignore them
+// based on area
+void SortInstancesByIgnore(
+ const std::array& area_range,
+ const std::vector& ground_truth_instances,
+ std::vector* ground_truth_sorted_indices,
+ std::vector* ignores) {
+ ignores->clear();
+ ignores->reserve(ground_truth_instances.size());
+ for (auto o : ground_truth_instances) {
+ ignores->push_back(
+ o.ignore || o.area < area_range[0] || o.area > area_range[1]);
+ }
+
+ ground_truth_sorted_indices->resize(ground_truth_instances.size());
+ std::iota(
+ ground_truth_sorted_indices->begin(),
+ ground_truth_sorted_indices->end(),
+ 0);
+ std::stable_sort(
+ ground_truth_sorted_indices->begin(),
+ ground_truth_sorted_indices->end(),
+ [&ignores](size_t j1, size_t j2) {
+ return (int)(*ignores)[j1] < (int)(*ignores)[j2];
+ });
+}
+
+// For each IOU threshold, greedily match each detected instance to a ground
+// truth instance (if possible) and store the results
+void MatchDetectionsToGroundTruth(
+ const std::vector& detection_instances,
+ const std::vector& detection_sorted_indices,
+ const std::vector& ground_truth_instances,
+ const std::vector& ground_truth_sorted_indices,
+ const std::vector& ignores,
+ const std::vector>& ious,
+ const std::vector& iou_thresholds,
+ const std::array& area_range,
+ ImageEvaluation* results) {
+ // Initialize memory to store return data matches and ignore
+ const int num_iou_thresholds = iou_thresholds.size();
+ const int num_ground_truth = ground_truth_sorted_indices.size();
+ const int num_detections = detection_sorted_indices.size();
+ std::vector ground_truth_matches(
+ num_iou_thresholds * num_ground_truth, 0);
+ std::vector& detection_matches = results->detection_matches;
+ std::vector& detection_ignores = results->detection_ignores;
+ std::vector& ground_truth_ignores = results->ground_truth_ignores;
+ detection_matches.resize(num_iou_thresholds * num_detections, 0);
+ detection_ignores.resize(num_iou_thresholds * num_detections, false);
+ ground_truth_ignores.resize(num_ground_truth);
+ for (auto g = 0; g < num_ground_truth; ++g) {
+ ground_truth_ignores[g] = ignores[ground_truth_sorted_indices[g]];
+ }
+
+ for (auto t = 0; t < num_iou_thresholds; ++t) {
+ for (auto d = 0; d < num_detections; ++d) {
+ // information about best match so far (match=-1 -> unmatched)
+ double best_iou = std::min(iou_thresholds[t], 1 - 1e-10);
+ int match = -1;
+ for (auto g = 0; g < num_ground_truth; ++g) {
+ // if this ground truth instance is already matched and not a
+ // crowd, it cannot be matched to another detection
+ if (ground_truth_matches[t * num_ground_truth + g] > 0 &&
+ !ground_truth_instances[ground_truth_sorted_indices[g]].is_crowd) {
+ continue;
+ }
+
+ // if detected instance matched to a regular ground truth
+ // instance, we can break on the first ground truth instance
+ // tagged as ignore (because they are sorted by the ignore tag)
+ if (match >= 0 && !ground_truth_ignores[match] &&
+ ground_truth_ignores[g]) {
+ break;
+ }
+
+ // if IOU overlap is the best so far, store the match appropriately
+ if (ious[d][ground_truth_sorted_indices[g]] >= best_iou) {
+ best_iou = ious[d][ground_truth_sorted_indices[g]];
+ match = g;
+ }
+ }
+ // if match was made, store id of match for both detection and
+ // ground truth
+ if (match >= 0) {
+ detection_ignores[t * num_detections + d] = ground_truth_ignores[match];
+ detection_matches[t * num_detections + d] =
+ ground_truth_instances[ground_truth_sorted_indices[match]].id;
+ ground_truth_matches[t * num_ground_truth + match] =
+ detection_instances[detection_sorted_indices[d]].id;
+ }
+
+ // set unmatched detections outside of area range to ignore
+ const InstanceAnnotation& detection =
+ detection_instances[detection_sorted_indices[d]];
+ detection_ignores[t * num_detections + d] =
+ detection_ignores[t * num_detections + d] ||
+ (detection_matches[t * num_detections + d] == 0 &&
+ (detection.area < area_range[0] || detection.area > area_range[1]));
+ }
+ }
+
+ // store detection score results
+ results->detection_scores.resize(detection_sorted_indices.size());
+ for (size_t d = 0; d < detection_sorted_indices.size(); ++d) {
+ results->detection_scores[d] =
+ detection_instances[detection_sorted_indices[d]].score;
+ }
+}
+
+std::vector EvaluateImages(
+ const std::vector>& area_ranges,
+ int max_detections,
+ const std::vector& iou_thresholds,
+ const ImageCategoryInstances>& image_category_ious,
+ const ImageCategoryInstances&
+ image_category_ground_truth_instances,
+ const ImageCategoryInstances&
+ image_category_detection_instances) {
+ const int num_area_ranges = area_ranges.size();
+ const int num_images = image_category_ground_truth_instances.size();
+ const int num_categories =
+ image_category_ious.size() > 0 ? image_category_ious[0].size() : 0;
+ std::vector detection_sorted_indices;
+ std::vector ground_truth_sorted_indices;
+ std::vector ignores;
+ std::vector results_all(
+ num_images * num_area_ranges * num_categories);
+
+ // Store results for each image, category, and area range combination. Results
+ // for each IOU threshold are packed into the same ImageEvaluation object
+ for (auto i = 0; i < num_images; ++i) {
+ for (auto c = 0; c < num_categories; ++c) {
+ const std::vector& ground_truth_instances =
+ image_category_ground_truth_instances[i][c];
+ const std::vector& detection_instances =
+ image_category_detection_instances[i][c];
+
+ SortInstancesByDetectionScore(
+ detection_instances, &detection_sorted_indices);
+ if ((int)detection_sorted_indices.size() > max_detections) {
+ detection_sorted_indices.resize(max_detections);
+ }
+
+ for (size_t a = 0; a < area_ranges.size(); ++a) {
+ SortInstancesByIgnore(
+ area_ranges[a],
+ ground_truth_instances,
+ &ground_truth_sorted_indices,
+ &ignores);
+
+ MatchDetectionsToGroundTruth(
+ detection_instances,
+ detection_sorted_indices,
+ ground_truth_instances,
+ ground_truth_sorted_indices,
+ ignores,
+ image_category_ious[i][c],
+ iou_thresholds,
+ area_ranges[a],
+ &results_all
+ [c * num_area_ranges * num_images + a * num_images + i]);
+ }
+ }
+ }
+
+ return results_all;
+}
+
+// Convert a python list to a vector
+template
+std::vector list_to_vec(const py::list& l) {
+ std::vector v(py::len(l));
+ for (int i = 0; i < (int)py::len(l); ++i) {
+ v[i] = l[i].cast();
+ }
+ return v;
+}
+
+// Helper function to Accumulate()
+// Considers the evaluation results applicable to a particular category, area
+// range, and max_detections parameter setting, which begin at
+// evaluations[evaluation_index]. Extracts a sorted list of length n of all
+// applicable detection instances concatenated across all images in the dataset,
+// which are represented by the outputs evaluation_indices, detection_scores,
+// image_detection_indices, and detection_sorted_indices--all of which are
+// length n. evaluation_indices[i] stores the applicable index into
+// evaluations[] for instance i, which has detection score detection_score[i],
+// and is the image_detection_indices[i]'th of the list of detections
+// for the image containing i. detection_sorted_indices[] defines a sorted
+// permutation of the 3 other outputs
+int BuildSortedDetectionList(
+ const std::vector& evaluations,
+ const int64_t evaluation_index,
+ const int64_t num_images,
+ const int max_detections,
+ std::vector* evaluation_indices,
+ std::vector* detection_scores,
+ std::vector* detection_sorted_indices,
+ std::vector* image_detection_indices) {
+ assert(evaluations.size() >= evaluation_index + num_images);
+
+ // Extract a list of object instances of the applicable category, area
+ // range, and max detections requirements such that they can be sorted
+ image_detection_indices->clear();
+ evaluation_indices->clear();
+ detection_scores->clear();
+ image_detection_indices->reserve(num_images * max_detections);
+ evaluation_indices->reserve(num_images * max_detections);
+ detection_scores->reserve(num_images * max_detections);
+ int num_valid_ground_truth = 0;
+ for (auto i = 0; i < num_images; ++i) {
+ const ImageEvaluation& evaluation = evaluations[evaluation_index + i];
+
+ for (int d = 0;
+ d < (int)evaluation.detection_scores.size() && d < max_detections;
+ ++d) { // detected instances
+ evaluation_indices->push_back(evaluation_index + i);
+ image_detection_indices->push_back(d);
+ detection_scores->push_back(evaluation.detection_scores[d]);
+ }
+ for (auto ground_truth_ignore : evaluation.ground_truth_ignores) {
+ if (!ground_truth_ignore) {
+ ++num_valid_ground_truth;
+ }
+ }
+ }
+
+ // Sort detections by decreasing score, using stable sort to match
+ // python implementation
+ detection_sorted_indices->resize(detection_scores->size());
+ std::iota(
+ detection_sorted_indices->begin(), detection_sorted_indices->end(), 0);
+ std::stable_sort(
+ detection_sorted_indices->begin(),
+ detection_sorted_indices->end(),
+ [&detection_scores](size_t j1, size_t j2) {
+ return (*detection_scores)[j1] > (*detection_scores)[j2];
+ });
+
+ return num_valid_ground_truth;
+}
+
+// Helper function to Accumulate()
+// Compute a precision recall curve given a sorted list of detected instances
+// encoded in evaluations, evaluation_indices, detection_scores,
+// detection_sorted_indices, image_detection_indices (see
+// BuildSortedDetectionList()). Using vectors precisions and recalls
+// and temporary storage, output the results into precisions_out, recalls_out,
+// and scores_out, which are large buffers containing many precion/recall curves
+// for all possible parameter settings, with precisions_out_index and
+// recalls_out_index defining the applicable indices to store results.
+void ComputePrecisionRecallCurve(
+ const int64_t precisions_out_index,
+ const int64_t precisions_out_stride,
+ const int64_t recalls_out_index,
+ const std::vector& recall_thresholds,
+ const int iou_threshold_index,
+ const int num_iou_thresholds,
+ const int num_valid_ground_truth,
+ const std::vector& evaluations,
+ const std::vector& evaluation_indices,
+ const std::vector& detection_scores,
+ const std::vector& detection_sorted_indices,
+ const std::vector& image_detection_indices,
+ std::vector* precisions,
+ std::vector* recalls,
+ std::vector* precisions_out,
+ std::vector* scores_out,
+ std::vector* recalls_out) {
+ assert(recalls_out->size() > recalls_out_index);
+
+ // Compute precision/recall for each instance in the sorted list of detections
+ int64_t true_positives_sum = 0, false_positives_sum = 0;
+ precisions->clear();
+ recalls->clear();
+ precisions->reserve(detection_sorted_indices.size());
+ recalls->reserve(detection_sorted_indices.size());
+ assert(!evaluations.empty() || detection_sorted_indices.empty());
+ for (auto detection_sorted_index : detection_sorted_indices) {
+ const ImageEvaluation& evaluation =
+ evaluations[evaluation_indices[detection_sorted_index]];
+ const auto num_detections =
+ evaluation.detection_matches.size() / num_iou_thresholds;
+ const auto detection_index = iou_threshold_index * num_detections +
+ image_detection_indices[detection_sorted_index];
+ assert(evaluation.detection_matches.size() > detection_index);
+ assert(evaluation.detection_ignores.size() > detection_index);
+ const int64_t detection_match =
+ evaluation.detection_matches[detection_index];
+ const bool detection_ignores =
+ evaluation.detection_ignores[detection_index];
+ const auto true_positive = detection_match > 0 && !detection_ignores;
+ const auto false_positive = detection_match == 0 && !detection_ignores;
+ if (true_positive) {
+ ++true_positives_sum;
+ }
+ if (false_positive) {
+ ++false_positives_sum;
+ }
+
+ const double recall =
+ static_cast(true_positives_sum) / num_valid_ground_truth;
+ recalls->push_back(recall);
+ const int64_t num_valid_detections =
+ true_positives_sum + false_positives_sum;
+ const double precision = num_valid_detections > 0
+ ? static_cast(true_positives_sum) / num_valid_detections
+ : 0.0;
+ precisions->push_back(precision);
+ }
+
+ (*recalls_out)[recalls_out_index] = !recalls->empty() ? recalls->back() : 0;
+
+ for (int64_t i = static_cast(precisions->size()) - 1; i > 0; --i) {
+ if ((*precisions)[i] > (*precisions)[i - 1]) {
+ (*precisions)[i - 1] = (*precisions)[i];
+ }
+ }
+
+ // Sample the per instance precision/recall list at each recall threshold
+ for (size_t r = 0; r < recall_thresholds.size(); ++r) {
+ // first index in recalls >= recall_thresholds[r]
+ std::vector::iterator low = std::lower_bound(
+ recalls->begin(), recalls->end(), recall_thresholds[r]);
+ size_t precisions_index = low - recalls->begin();
+
+ const auto results_ind = precisions_out_index + r * precisions_out_stride;
+ assert(results_ind < precisions_out->size());
+ assert(results_ind < scores_out->size());
+ if (precisions_index < precisions->size()) {
+ (*precisions_out)[results_ind] = (*precisions)[precisions_index];
+ (*scores_out)[results_ind] =
+ detection_scores[detection_sorted_indices[precisions_index]];
+ } else {
+ (*precisions_out)[results_ind] = 0;
+ (*scores_out)[results_ind] = 0;
+ }
+ }
+}
+py::dict Accumulate(
+ const py::object& params,
+ const std::vector& evaluations) {
+ const std::vector recall_thresholds =
+ list_to_vec(params.attr("recThrs"));
+ const std::vector max_detections =
+ list_to_vec(params.attr("maxDets"));
+ const int num_iou_thresholds = py::len(params.attr("iouThrs"));
+ const int num_recall_thresholds = py::len(params.attr("recThrs"));
+ const int num_categories = params.attr("useCats").cast() == 1
+ ? py::len(params.attr("catIds"))
+ : 1;
+ const int num_area_ranges = py::len(params.attr("areaRng"));
+ const int num_max_detections = py::len(params.attr("maxDets"));
+ const int num_images = py::len(params.attr("imgIds"));
+
+ std::vector precisions_out(
+ num_iou_thresholds * num_recall_thresholds * num_categories *
+ num_area_ranges * num_max_detections,
+ -1);
+ std::vector recalls_out(
+ num_iou_thresholds * num_categories * num_area_ranges *
+ num_max_detections,
+ -1);
+ std::vector scores_out(
+ num_iou_thresholds * num_recall_thresholds * num_categories *
+ num_area_ranges * num_max_detections,
+ -1);
+
+ // Consider the list of all detected instances in the entire dataset in one
+ // large list. evaluation_indices, detection_scores,
+ // image_detection_indices, and detection_sorted_indices all have the same
+ // length as this list, such that each entry corresponds to one detected
+ // instance
+ std::vector evaluation_indices; // indices into evaluations[]
+ std::vector detection_scores; // detection scores of each instance
+ std::vector detection_sorted_indices; // sorted indices of all
+ // instances in the dataset
+ std::vector
+ image_detection_indices; // indices into the list of detected instances in
+ // the same image as each instance
+ std::vector precisions, recalls;
+
+ for (auto c = 0; c < num_categories; ++c) {
+ for (auto a = 0; a < num_area_ranges; ++a) {
+ for (auto m = 0; m < num_max_detections; ++m) {
+ // The COCO PythonAPI assumes evaluations[] (the return value of
+ // COCOeval::EvaluateImages() is one long list storing results for each
+ // combination of category, area range, and image id, with categories in
+ // the outermost loop and images in the innermost loop.
+ const int64_t evaluations_index =
+ c * num_area_ranges * num_images + a * num_images;
+ int num_valid_ground_truth = BuildSortedDetectionList(
+ evaluations,
+ evaluations_index,
+ num_images,
+ max_detections[m],
+ &evaluation_indices,
+ &detection_scores,
+ &detection_sorted_indices,
+ &image_detection_indices);
+
+ if (num_valid_ground_truth == 0) {
+ continue;
+ }
+
+ for (auto t = 0; t < num_iou_thresholds; ++t) {
+ // recalls_out is a flattened vectors representing a
+ // num_iou_thresholds X num_categories X num_area_ranges X
+ // num_max_detections matrix
+ const int64_t recalls_out_index =
+ t * num_categories * num_area_ranges * num_max_detections +
+ c * num_area_ranges * num_max_detections +
+ a * num_max_detections + m;
+
+ // precisions_out and scores_out are flattened vectors
+ // representing a num_iou_thresholds X num_recall_thresholds X
+ // num_categories X num_area_ranges X num_max_detections matrix
+ const int64_t precisions_out_stride =
+ num_categories * num_area_ranges * num_max_detections;
+ const int64_t precisions_out_index = t * num_recall_thresholds *
+ num_categories * num_area_ranges * num_max_detections +
+ c * num_area_ranges * num_max_detections +
+ a * num_max_detections + m;
+
+ ComputePrecisionRecallCurve(
+ precisions_out_index,
+ precisions_out_stride,
+ recalls_out_index,
+ recall_thresholds,
+ t,
+ num_iou_thresholds,
+ num_valid_ground_truth,
+ evaluations,
+ evaluation_indices,
+ detection_scores,
+ detection_sorted_indices,
+ image_detection_indices,
+ &precisions,
+ &recalls,
+ &precisions_out,
+ &scores_out,
+ &recalls_out);
+ }
+ }
+ }
+ }
+
+ time_t rawtime;
+ struct tm local_time;
+ std::array buffer;
+ time(&rawtime);
+#ifdef _WIN32
+ localtime_s(&local_time, &rawtime);
+#else
+ localtime_r(&rawtime, &local_time);
+#endif
+ strftime(
+ buffer.data(), 200, "%Y-%m-%d %H:%num_max_detections:%S", &local_time);
+ return py::dict(
+ "params"_a = params,
+ "counts"_a = std::vector(
+ {num_iou_thresholds,
+ num_recall_thresholds,
+ num_categories,
+ num_area_ranges,
+ num_max_detections}),
+ "date"_a = buffer,
+ "precision"_a = precisions_out,
+ "recall"_a = recalls_out,
+ "scores"_a = scores_out);
+}
+
+} // namespace COCOeval
+
+} // namespace detectron2
diff --git a/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/cocoeval/cocoeval.h b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/cocoeval/cocoeval.h
new file mode 100644
index 0000000000000000000000000000000000000000..db246e49a026b7cd989b305f4d3d98100be3c912
--- /dev/null
+++ b/ControlNet-v1-1-nightly-main/annotator/oneformer/detectron2/layers/csrc/cocoeval/cocoeval.h
@@ -0,0 +1,88 @@
+// Copyright (c) Facebook, Inc. and its affiliates.
+#pragma once
+
+#include
+#include
+#include
+#include
+#include
+
+namespace py = pybind11;
+
+namespace detectron2 {
+
+namespace COCOeval {
+
+// Annotation data for a single object instance in an image
+struct InstanceAnnotation {
+ InstanceAnnotation(
+ uint64_t id,
+ double score,
+ double area,
+ bool is_crowd,
+ bool ignore)
+ : id{id}, score{score}, area{area}, is_crowd{is_crowd}, ignore{ignore} {}
+ uint64_t id;
+ double score = 0.;
+ double area = 0.;
+ bool is_crowd = false;
+ bool ignore = false;
+};
+
+// Stores intermediate results for evaluating detection results for a single
+// image that has D detected instances and G ground truth instances. This stores
+// matches between detected and ground truth instances
+struct ImageEvaluation {
+ // For each of the D detected instances, the id of the matched ground truth
+ // instance, or 0 if unmatched
+ std::vector detection_matches;
+
+ // The detection score of each of the D detected instances
+ std::vector