Upload endoscopic_inbody_classification version 0.5.0

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+models/model.ts filter=lfs diff=lfs merge=lfs -text

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 [yyyy] [name of copyright owner]
+
+   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.

configs/evaluate.json

@@ -0,0 +1,47 @@
+{
+    "validate#postprocessing": {
+        "_target_": "Compose",
+        "transforms": [
+            {
+                "_target_": "AsDiscreted",
+                "keys": [
+                    "pred",
+                    "label"
+                ],
+                "argmax": [
+                    true,
+                    false
+                ],
+                "to_onehot": 2
+            }
+        ]
+    },
+    "validate#handlers": [
+        {
+            "_target_": "CheckpointLoader",
+            "load_path": "$@ckpt_dir + '/model.pt'",
+            "load_dict": {
+                "model": "@network"
+            }
+        },
+        {
+            "_target_": "StatsHandler",
+            "iteration_log": false
+        },
+        {
+            "_target_": "MetricsSaver",
+            "save_dir": "@output_dir",
+            "metrics": [
+                "val_accu"
+            ],
+            "metric_details": [
+                "val_accu"
+            ],
+            "batch_transform": "$lambda x: [xx['image'].meta for xx in x]",
+            "summary_ops": "*"
+        }
+    ],
+    "run": [
+        "$@validate#evaluator.run()"
+    ]
+}

configs/inference.json

@@ -0,0 +1,131 @@
+{
+    "imports": [
+        "$import json",
+        "$import os",
+        "$import numpy",
+        "$import torch"
+    ],
+    "bundle_root": ".",
+    "image_key": "image",
+    "load_pretrain": true,
+    "output_dir": "$@bundle_root + '/eval'",
+    "output_filename": "predictions.csv",
+    "dataset_dir": "/workspace/data/endoscopic_inbody_classification",
+    "test_json": "$@bundle_root+'/label/test_samples.json'",
+    "test_fp": "$open(@test_json,'r', encoding='utf8')",
+    "test_dict": "$json.load(@test_fp)",
+    "device": "$torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')",
+    "network_def": {
+        "_target_": "SEResNet50",
+        "spatial_dims": 2,
+        "in_channels": 3,
+        "num_classes": 2
+    },
+    "network": "$@network_def.to(@device)",
+    "preprocessing": {
+        "_target_": "Compose",
+        "transforms": [
+            {
+                "_target_": "LoadImaged",
+                "keys": "@image_key",
+                "image_only": false
+            },
+            {
+                "_target_": "EnsureChannelFirstd",
+                "keys": "@image_key",
+                "channel_dim": -1
+            },
+            {
+                "_target_": "Resized",
+                "keys": "@image_key",
+                "spatial_size": [
+                    256,
+                    256
+                ],
+                "mode": "bilinear"
+            },
+            {
+                "_target_": "CastToTyped",
+                "dtype": "$torch.float32",
+                "keys": "@image_key"
+            },
+            {
+                "_target_": "NormalizeIntensityd",
+                "nonzero": true,
+                "channel_wise": true,
+                "keys": "@image_key"
+            },
+            {
+                "_target_": "EnsureTyped",
+                "keys": "@image_key"
+            }
+        ]
+    },
+    "dataset": {
+        "_target_": "Dataset",
+        "data": "@test_dict",
+        "transform": "@preprocessing"
+    },
+    "dataloader": {
+        "_target_": "DataLoader",
+        "dataset": "@dataset",
+        "batch_size": 1,
+        "shuffle": false,
+        "num_workers": 4
+    },
+    "inferer": {
+        "_target_": "SimpleInferer"
+    },
+    "postprocessing": {
+        "_target_": "Compose",
+        "transforms": [
+            {
+                "_target_": "AsDiscreted",
+                "argmax": true,
+                "to_onehot": 2,
+                "keys": [
+                    "pred"
+                ]
+            },
+            {
+                "_target_": "SaveClassificationd",
+                "keys": "pred",
+                "meta_keys": "$@image_key + '_meta_dict'",
+                "output_dir": "@output_dir",
+                "filename": "@output_filename"
+            }
+        ]
+    },
+    "handlers": [
+        {
+            "_target_": "StatsHandler",
+            "iteration_log": true
+        }
+    ],
+    "evaluator": {
+        "_target_": "SupervisedEvaluator",
+        "device": "@device",
+        "val_data_loader": "@dataloader",
+        "network": "@network",
+        "inferer": "@inferer",
+        "postprocessing": "@postprocessing",
+        "val_handlers": "@handlers"
+    },
+    "checkpointloader": {
+        "_target_": "CheckpointLoader",
+        "load_path": "$@bundle_root + '/models/model.pt'",
+        "load_dict": {
+            "model": "@network"
+        }
+    },
+    "initialize": [
+        "$monai.utils.set_determinism(seed=123)",
+        "$@checkpointloader(@evaluator) if @load_pretrain else None"
+    ],
+    "run": [
+        "[email protected]()"
+    ],
+    "finalize": [
+        "$@test_fp.close()"
+    ]
+}

configs/inference_trt.json

@@ -0,0 +1,12 @@
+{
+    "imports": [
+        "$import os",
+        "$import json",
+        "$import torch_tensorrt"
+    ],
+    "network_def": "$torch.jit.load(@bundle_root + '/models/model_trt.ts')",
+    "evaluator#amp": false,
+    "initialize": [
+        "$monai.utils.set_determinism(seed=123)"
+    ]
+}

configs/logging.conf

@@ -0,0 +1,21 @@
+[loggers]
+keys=root
+
+[handlers]
+keys=consoleHandler
+
+[formatters]
+keys=fullFormatter
+
+[logger_root]
+level=INFO
+handlers=consoleHandler
+
+[handler_consoleHandler]
+class=StreamHandler
+level=INFO
+formatter=fullFormatter
+args=(sys.stdout,)
+
+[formatter_fullFormatter]
+format=%(asctime)s - %(name)s - %(levelname)s - %(message)s

configs/metadata.json

@@ -0,0 +1,104 @@
+{
+    "schema": "https://github.com/Project-MONAI/MONAI-extra-test-data/releases/download/0.8.1/meta_schema_20240725.json",
+    "version": "0.5.0",
+    "changelog": {
+        "0.5.0": "update to huggingface hosting and fix missing dependencies",
+        "0.4.9": "use monai 1.4 and update large files",
+        "0.4.8": "update to use monai 1.3.1",
+        "0.4.7": "add load_pretrain flag for infer",
+        "0.4.6": "add output for inference",
+        "0.4.5": "update with EnsureChannelFirstd and remove meta dict usage",
+        "0.4.4": "fix the wrong GPU index issue of multi-node",
+        "0.4.3": "add dataset dir example",
+        "0.4.2": "update ONNX-TensorRT descriptions",
+        "0.4.1": "update the model weights with the deterministic training",
+        "0.4.0": "add the ONNX-TensorRT way of model conversion",
+        "0.3.9": "fix mgpu finalize issue",
+        "0.3.8": "enable deterministic training",
+        "0.3.7": "adapt to BundleWorkflow interface",
+        "0.3.6": "add name tag",
+        "0.3.5": "fix a comment issue in the data_process script",
+        "0.3.4": "add note for multi-gpu training with example dataset",
+        "0.3.3": "enhance data preprocess script and readme file",
+        "0.3.2": "restructure readme to match updated template",
+        "0.3.1": "add workflow, train loss and validation accuracy figures",
+        "0.3.0": "update dataset processing",
+        "0.2.2": "update to use monai 1.0.1",
+        "0.2.1": "enhance readme on commands example",
+        "0.2.0": "update license files",
+        "0.1.0": "complete the first version model package",
+        "0.0.1": "initialize the model package structure"
+    },
+    "monai_version": "1.4.0",
+    "pytorch_version": "2.4.0",
+    "numpy_version": "1.24.4",
+    "required_packages_version": {
+        "nibabel": "5.2.1",
+        "pytorch-ignite": "0.4.11",
+        "pillow": "10.4.0",
+        "tensorboard": "2.17.0"
+    },
+    "supported_apps": {},
+    "name": "Endoscopic inbody classification",
+    "task": "Endoscopic inbody classification",
+    "description": "A pre-trained binary classification model for endoscopic inbody classification task",
+    "authors": "NVIDIA DLMED team",
+    "copyright": "Copyright (c) 2021-2022, NVIDIA CORPORATION",
+    "data_source": "private dataset",
+    "data_type": "RGB",
+    "image_classes": "three channel data, intensity [0-255]",
+    "label_classes": "0: inbody, 1: outbody",
+    "pred_classes": "vector whose length equals to 2, [1,0] means in body, [0,1] means out body",
+    "eval_metrics": {
+        "accuracy": 0.99
+    },
+    "references": [
+        "J. Hu, L. Shen and G. Sun, Squeeze-and-Excitation Networks, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2018, pp. 7132-7141. https://arxiv.org/pdf/1709.01507.pdf"
+    ],
+    "network_data_format": {
+        "inputs": {
+            "image": {
+                "type": "magnitude",
+                "format": "RGB",
+                "modality": "regular",
+                "num_channels": 3,
+                "spatial_shape": [
+                    256,
+                    256
+                ],
+                "dtype": "float32",
+                "value_range": [
+                    0,
+                    1
+                ],
+                "is_patch_data": false,
+                "channel_def": {
+                    "0": "R",
+                    "1": "G",
+                    "2": "B"
+                }
+            }
+        },
+        "outputs": {
+            "pred": {
+                "type": "probabilities",
+                "format": "classes",
+                "num_channels": 2,
+                "spatial_shape": [
+                    1,
+                    2
+                ],
+                "dtype": "float32",
+                "value_range": [
+                    0,
+                    1
+                ],
+                "is_patch_data": false,
+                "channel_def": {
+                    "0": "in body",
+                    "1": "out body"
+                }
+            }
+        }
+    }
+}

configs/multi_gpu_train.json

@@ -0,0 +1,41 @@
+{
+    "device": "$torch.device('cuda:' + os.environ['LOCAL_RANK'])",
+    "network": {
+        "_target_": "torch.nn.parallel.DistributedDataParallel",
+        "module": "$@network_def.to(@device)",
+        "device_ids": [
+            "@device"
+        ]
+    },
+    "train#sampler": {
+        "_target_": "DistributedSampler",
+        "dataset": "@train#dataset",
+        "even_divisible": true,
+        "shuffle": true
+    },
+    "train#dataloader#sampler": "@train#sampler",
+    "train#dataloader#shuffle": false,
+    "train#trainer#train_handlers": "$@train#handlers[: -2 if dist.get_rank() > 0 else None]",
+    "validate#sampler": {
+        "_target_": "DistributedSampler",
+        "dataset": "@validate#dataset",
+        "even_divisible": false,
+        "shuffle": false
+    },
+    "validate#dataloader#sampler": "@validate#sampler",
+    "validate#evaluator#val_handlers": "$None if dist.get_rank() > 0 else @validate#handlers",
+    "initialize": [
+        "$import torch.distributed as dist",
+        "$dist.is_initialized() or dist.init_process_group(backend='nccl')",
+        "$torch.cuda.set_device(@device)",
+        "$monai.utils.set_determinism(seed=123)"
+    ],
+    "run": [
+        "$@train#trainer.run()"
+    ],
+    "finalize": [
+        "$dist.is_initialized() and dist.destroy_process_group()",
+        "$@train_fp.close()",
+        "$@val_fp.close()"
+    ]
+}

configs/train.json

@@ -0,0 +1,264 @@
+{
+    "imports": [
+        "$import torch",
+        "$import json",
+        "$import ignite",
+        "$import os"
+    ],
+    "bundle_root": ".",
+    "ckpt_dir": "$@bundle_root + '/models'",
+    "output_dir": "$@bundle_root + '/eval'",
+    "dataset_dir": "/workspace/data/endoscopic_inbody_classification",
+    "train_json": "$@bundle_root+'/label/train_samples.json'",
+    "val_json": "$@bundle_root+'/label/val_samples.json'",
+    "train_fp": "$open(@train_json,'r', encoding='utf8')",
+    "train_dict": "$json.load(@train_fp)",
+    "val_fp": "$open(@val_json,'r', encoding='utf8')",
+    "val_dict": "$json.load(@val_fp)",
+    "val_interval": 1,
+    "device": "$torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')",
+    "network_def": {
+        "_target_": "SEResNet50",
+        "spatial_dims": 2,
+        "in_channels": 3,
+        "num_classes": 2
+    },
+    "network": "$@network_def.to(@device)",
+    "loss": {
+        "_target_": "torch.nn.CrossEntropyLoss",
+        "reduction": "sum"
+    },
+    "optimizer": {
+        "_target_": "torch.optim.Adam",
+        "params": "[email protected]()",
+        "lr": 0.001
+    },
+    "train": {
+        "deterministic_transforms": [
+            {
+                "_target_": "LoadImaged",
+                "keys": "image"
+            },
+            {
+                "_target_": "ToTensord",
+                "keys": "label"
+            },
+            {
+                "_target_": "EnsureChannelFirstd",
+                "keys": "image",
+                "channel_dim": -1
+            },
+            {
+                "_target_": "Resized",
+                "keys": "image",
+                "spatial_size": [
+                    256,
+                    256
+                ],
+                "mode": "bilinear"
+            },
+            {
+                "_target_": "CastToTyped",
+                "dtype": "$torch.float32",
+                "keys": "image"
+            },
+            {
+                "_target_": "NormalizeIntensityd",
+                "nonzero": true,
+                "channel_wise": true,
+                "keys": "image"
+            },
+            {
+                "_target_": "EnsureTyped",
+                "keys": "image"
+            }
+        ],
+        "random_transforms": [
+            {
+                "_target_": "RandRotated",
+                "range_x": 0.3,
+                "prob": 0.2,
+                "mode": "bilinear",
+                "keys": "image"
+            },
+            {
+                "_target_": "RandScaleIntensityd",
+                "factors": 0.3,
+                "prob": 0.5,
+                "keys": "image"
+            },
+            {
+                "_target_": "RandShiftIntensityd",
+                "offsets": 0.1,
+                "prob": 0.5,
+                "keys": "image"
+            },
+            {
+                "_target_": "RandGaussianNoised",
+                "std": 0.01,
+                "prob": 0.15,
+                "keys": "image"
+            },
+            {
+                "_target_": "RandFlipd",
+                "spatial_axis": 0,
+                "prob": 0.5,
+                "keys": "image"
+            },
+            {
+                "_target_": "RandFlipd",
+                "spatial_axis": 1,
+                "prob": 0.5,
+                "keys": "image"
+            }
+        ],
+        "preprocessing": {
+            "_target_": "Compose",
+            "transforms": "$@train#deterministic_transforms + @train#random_transforms"
+        },
+        "dataset": {
+            "_target_": "Dataset",
+            "data": "@train_dict",
+            "transform": "@train#preprocessing"
+        },
+        "dataloader": {
+            "_target_": "DataLoader",
+            "dataset": "@train#dataset",
+            "batch_size": 64,
+            "shuffle": true,
+            "num_workers": 4
+        },
+        "inferer": {
+            "_target_": "SimpleInferer"
+        },
+        "handlers": [
+            {
+                "_target_": "ValidationHandler",
+                "validator": "@validate#evaluator",
+                "epoch_level": true,
+                "interval": "@val_interval"
+            },
+            {
+                "_target_": "StatsHandler",
+                "tag_name": "train_loss",
+                "output_transform": "$monai.handlers.from_engine(['loss'], first=True)"
+            },
+            {
+                "_target_": "TensorBoardStatsHandler",
+                "log_dir": "@output_dir",
+                "tag_name": "train_loss",
+                "output_transform": "$monai.handlers.from_engine(['loss'], first=True)"
+            }
+        ],
+        "key_metric": {
+            "train_accu": {
+                "_target_": "ignite.metrics.Accuracy",
+                "output_transform": "$monai.handlers.from_engine(['pred', 'label'])"
+            }
+        },
+        "postprocessing": {
+            "_target_": "Compose",
+            "transforms": [
+                {
+                    "_target_": "AsDiscreted",
+                    "argmax": [
+                        true,
+                        false
+                    ],
+                    "to_onehot": [
+                        2,
+                        2
+                    ],
+                    "keys": [
+                        "pred",
+                        "label"
+                    ]
+                }
+            ]
+        },
+        "trainer": {
+            "_target_": "SupervisedTrainer",
+            "max_epochs": 25,
+            "device": "@device",
+            "train_data_loader": "@train#dataloader",
+            "network": "@network",
+            "loss_function": "@loss",
+            "optimizer": "@optimizer",
+            "inferer": "@train#inferer",
+            "postprocessing": "@train#postprocessing",
+            "key_train_metric": "@train#key_metric",
+            "train_handlers": "@train#handlers"
+        }
+    },
+    "validate": {
+        "preprocessing": {
+            "_target_": "Compose",
+            "transforms": "%train#deterministic_transforms"
+        },
+        "dataset": {
+            "_target_": "Dataset",
+            "data": "@val_dict",
+            "transform": "@validate#preprocessing"
+        },
+        "dataloader": {
+            "_target_": "DataLoader",
+            "dataset": "@validate#dataset",
+            "batch_size": 64,
+            "shuffle": false,
+            "num_workers": 4
+        },
+        "inferer": {
+            "_target_": "SimpleInferer"
+        },
+        "postprocessing": {
+            "_target_": "Compose",
+            "transforms": "%train#postprocessing"
+        },
+        "handlers": [
+            {
+                "_target_": "StatsHandler",
+                "iteration_log": false
+            },
+            {
+                "_target_": "TensorBoardStatsHandler",
+                "log_dir": "@output_dir",
+                "iteration_log": false
+            },
+            {
+                "_target_": "CheckpointSaver",
+                "save_dir": "@ckpt_dir",
+                "save_dict": {
+                    "model": "@network"
+                },
+                "save_key_metric": true,
+                "key_metric_filename": "model.pt"
+            }
+        ],
+        "key_metric": {
+            "val_accu": {
+                "_target_": "ignite.metrics.Accuracy",
+                "output_transform": "$monai.handlers.from_engine(['pred', 'label'])"
+            }
+        },
+        "evaluator": {
+            "_target_": "SupervisedEvaluator",
+            "device": "@device",
+            "val_data_loader": "@validate#dataloader",
+            "network": "@network",
+            "inferer": "@validate#inferer",
+            "postprocessing": "@validate#postprocessing",
+            "key_val_metric": "@validate#key_metric",
+            "val_handlers": "@validate#handlers"
+        }
+    },
+    "initialize": [
+        "$monai.utils.set_determinism(seed=123)"
+    ],
+    "run": [
+        "$@train#trainer.run()"
+    ],
+    "finalize": [
+        "$@train_fp.close()",
+        "$@val_fp.close()"
+    ]
+}

docs/README.md

@@ -0,0 +1,175 @@
+# Model Overview
+A pre-trained model for the endoscopic inbody classification task and trained using the SEResNet50 structure, whose details can be found in [1]. All datasets are from private samples of [Activ Surgical](https://www.activsurgical.com/). Samples in training and validation dataset are from the same 4 videos, while test samples are from different two videos.
+
+The [PyTorch model](https://drive.google.com/file/d/14CS-s1uv2q6WedYQGeFbZeEWIkoyNa-x/view?usp=sharing) and [torchscript model](https://drive.google.com/file/d/1fOoJ4n5DWKHrt9QXTZ2sXwr9C-YvVGCM/view?usp=sharing) are shared in google drive. Modify the `bundle_root` parameter specified in `configs/train.json` and `configs/inference.json` to reflect where models are downloaded. Expected directory path to place downloaded models is `models/` under `bundle_root`.
+
+![image](https://developer.download.nvidia.com/assets/Clara/Images/monai_endoscopic_inbody_classification_workflow.png)
+
+## Data
+The datasets used in this work were provided by [Activ Surgical](https://www.activsurgical.com/).
+
+Since datasets are private, we provide a [link](https://github.com/Project-MONAI/MONAI-extra-test-data/releases/download/0.8.1/inbody_outbody_samples.zip) of 20 samples (10 in-body and 10 out-body) to show what they look like.
+
+### Preprocessing
+After downloading this dataset, python script in `scripts` folder named `data_process` can be used to generate label json files by running the command below and modifying `datapath` to path of unziped downloaded data. Generated label json files will be stored in `label` folder under the bundle path.
+
+```
+python scripts/data_process.py --datapath /path/to/data/root
+```
+
+By default, label path parameter in `train.json` and `inference.json` of this bundle is point to the generated `label` folder under bundle path. If you move these generated label files to another place, please modify the `train_json`, `val_json` and `test_json` parameters specified in `configs/train.json` and `configs/inference.json` to where these label files are.
+
+The input label json should be a list made up by dicts which includes `image` and `label` keys. An example format is shown below.
+
+```
+[
+    {
+        "image":"/path/to/image/image_name0.jpg",
+        "label": 0
+    },
+    {
+        "image":"/path/to/image/image_name1.jpg",
+        "label": 0
+    },
+    {
+        "image":"/path/to/image/image_name2.jpg",
+        "label": 1
+    },
+    ....
+    {
+        "image":"/path/to/image/image_namek.jpg",
+        "label": 0
+    },
+]
+```
+
+## Training configuration
+The training as performed with the following:
+- GPU: At least 12GB of GPU memory
+- Actual Model Input: 256 x 256 x 3
+- Optimizer: Adam
+- Learning Rate: 1e-3
+
+### Input
+A three channel video frame
+
+### Output
+Two Channels
+- Label 0: in body
+- Label 1: out body
+
+## Performance
+Accuracy was used for evaluating the performance of the model. This model achieves an accuracy score of 0.99
+
+#### Training Loss
+![A graph showing the training loss over 25 epochs.](https://developer.download.nvidia.com/assets/Clara/Images/monai_endoscopic_inbody_classification_train_loss_v2.png)
+
+#### Validation Accuracy
+![A graph showing the validation accuracy over 25 epochs.](https://developer.download.nvidia.com/assets/Clara/Images/monai_endoscopic_inbody_classification_val_accuracy_v2.png)
+
+#### TensorRT speedup
+The `endoscopic_inbody_classification` bundle supports acceleration with TensorRT through the ONNX-TensorRT method. The table below displays the speedup ratios observed on an A100 80G GPU.
+
+| method | torch_fp32(ms) | torch_amp(ms) | trt_fp32(ms) | trt_fp16(ms) | speedup amp | speedup fp32 | speedup fp16 | amp vs fp16|
+| :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: |
+| model computation | 6.50 | 9.23 | 2.78 | 2.31 | 0.70 | 2.34 | 2.81 | 4.00 |
+| end2end | 23.54 | 23.78 | 7.37 | 7.14 | 0.99 | 3.19 | 3.30 | 3.33 |
+
+Where:
+- `model computation` means the speedup ratio of model's inference with a random input without preprocessing and postprocessing
+- `end2end` means run the bundle end-to-end with the TensorRT based model.
+- `torch_fp32` and `torch_amp` are for the PyTorch models with or without `amp` mode.
+- `trt_fp32` and `trt_fp16` are for the TensorRT based models converted in corresponding precision.
+- `speedup amp`, `speedup fp32` and `speedup fp16` are the speedup ratios of corresponding models versus the PyTorch float32 model
+- `amp vs fp16` is the speedup ratio between the PyTorch amp model and the TensorRT float16 based model.
+
+Currently, the only available method to accelerate this model is through ONNX-TensorRT. However, the Torch-TensorRT method is under development and will be available in the near future.
+
+This result is benchmarked under:
+ - TensorRT: 8.5.3+cuda11.8
+ - Torch-TensorRT Version: 1.4.0
+ - CPU Architecture: x86-64
+ - OS: ubuntu 20.04
+ - Python version:3.8.10
+ - CUDA version: 12.0
+ - GPU models and configuration: A100 80G
+
+## MONAI Bundle Commands
+In addition to the Pythonic APIs, a few command line interfaces (CLI) are provided to interact with the bundle. The CLI supports flexible use cases, such as overriding configs at runtime and predefining arguments in a file.
+
+For more details usage instructions, visit the [MONAI Bundle Configuration Page](https://docs.monai.io/en/latest/config_syntax.html).
+
+#### Execute training:
+
+```
+python -m monai.bundle run --config_file configs/train.json
+```
+
+Please note that if the default dataset path is not modified with the actual path in the bundle config files, you can also override it by using `--dataset_dir`:
+
+```
+python -m monai.bundle run --config_file configs/train.json --dataset_dir <actual dataset path>
+```
+
+#### Override the `train` config to execute multi-GPU training:
+
+```
+torchrun --standalone --nnodes=1 --nproc_per_node=2 -m monai.bundle run \
+    --config_file "['configs/train.json','configs/multi_gpu_train.json']"
+```
+
+Please note that the distributed training-related options depend on the actual running environment; thus, users may need to remove `--standalone`, modify `--nnodes`, or do some other necessary changes according to the machine used. For more details, please refer to [pytorch's official tutorial](https://pytorch.org/tutorials/intermediate/ddp_tutorial.html).
+
+In addition, if using the 20 samples example dataset, the preprocessing script will divide the samples to 16 training samples, 2 validation samples and 2 test samples. However, pytorch multi-gpu training requires number of samples in dataloader larger than gpu numbers. Therefore, please use no more than 2 gpus to run this bundle if using the 20 samples example dataset.
+
+#### Override the `train` config to execute evaluation with the trained model:
+
+```
+python -m monai.bundle run --config_file "['configs/train.json','configs/evaluate.json']"
+```
+
+#### Execute inference:
+
+```
+python -m monai.bundle run --config_file configs/inference.json
+```
+The classification result of every images in `test.json` will be printed to the screen.
+
+#### Export checkpoint to TorchScript file:
+
+```
+python -m monai.bundle ckpt_export network_def --filepath models/model.ts --ckpt_file models/model.pt --meta_file configs/metadata.json --config_file configs/inference.json
+```
+
+#### Export checkpoint to TensorRT based models with fp32 or fp16 precision:
+
+```bash
+python -m monai.bundle trt_export --net_id network_def \
+--filepath models/model_trt.ts --ckpt_file models/model.pt \
+--meta_file configs/metadata.json --config_file configs/inference.json \
+--precision <fp32/fp16>  --use_onnx "True" --use_trace "True"
+```
+
+#### Execute inference with the TensorRT model:
+
+```
+python -m monai.bundle run --config_file "['configs/inference.json', 'configs/inference_trt.json']"
+```
+
+# References
+[1] J. Hu, L. Shen and G. Sun, Squeeze-and-Excitation Networks, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2018, pp. 7132-7141. https://arxiv.org/pdf/1709.01507.pdf
+
+# License
+Copyright (c) MONAI Consortium
+
+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.

docs/data_license.txt

@@ -0,0 +1 @@
+Datasets used in this work were provided by Activ Surgical.

models/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dc8d874efa195e4416124aad4e1469e0b3ef873753e20948d762e7bfdb37b929
+size 104502013

models/model.ts

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ba8c62eca55d37044d6d7631b1b0444709ac9009c202839dc054ca38bf732edf
+size 104609651

scripts/data_process.py

@@ -0,0 +1,69 @@
+import argparse
+import json
+import os
+
+train_rate = 0.6
+val_rate = 0.2
+test_rate = 0.2
+
+
+def save_json(content, path, filename):
+    if not os.path.exists(path):
+        os.makedirs(path, exist_ok=True)
+    dst_file_name = os.path.join(path, filename)
+    with open(dst_file_name, "w+") as fp:
+        json.dump(content, fp, indent=4, separators=(",", ":"))
+
+
+def generate_labels(data_path, output_path):
+    """
+    Loading a model by name.
+
+    Args:
+        data_path: path to classification dataset, which must contain `inbody` and `outbody` directories.
+        output_path: path to save labels
+    """
+
+    data_list = [os.path.join(root, x) for root, _, filenames in os.walk(data_path) for x in filenames if "jpg" in x]
+    label_list = [int("outbody" in os.path.basename(os.path.dirname(x))) for x in data_list]
+    data_label_json = [{"image": x, "label": y} for x, y in zip(data_list, label_list)]
+    inbody_list = list(filter(lambda x: x["label"] == 0, data_label_json))
+    outbody_list = list(filter(lambda x: not (x["label"] == 0), data_label_json))
+    inbody_train_len = int(len(inbody_list) * train_rate)
+    outbody_train_len = int(len(outbody_list) * train_rate)
+    inbody_val_len = int(len(inbody_list) * (train_rate + val_rate))
+    outbody_val_len = int(len(outbody_list) * (train_rate + val_rate))
+    inbody_train_list = inbody_list[:inbody_train_len]
+    outbody_train_list = outbody_list[:outbody_train_len]
+    inbody_val_list = inbody_list[inbody_train_len:inbody_val_len]
+    outbody_val_list = outbody_list[outbody_train_len:outbody_val_len]
+    inbody_test_list = inbody_list[inbody_val_len:]
+    outbody_test_list = outbody_list[outbody_val_len:]
+    train_list = inbody_train_list + outbody_train_list
+    val_list = inbody_val_list + outbody_val_list
+    test_list = inbody_test_list + outbody_test_list
+    save_json(train_list, out_path, "train_samples.json")
+    save_json(val_list, out_path, "val_samples.json")
+    save_json(test_list, out_path, "test_samples.json")
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    # path to downloaded dataset.
+    parser.add_argument(
+        "--datapath",
+        type=str,
+        default=r"/workspace/data/endoscopic_inbody_classification",
+        help="The root path of the inbody classification dataset.",
+    )
+
+    # path to save label json.
+    parser.add_argument("--outpath", type=str, default=r"./label", help="The output path of labels.")
+
+    args = parser.parse_args()
+    data_path = args.datapath
+    out_path = args.outpath
+
+    if not os.path.exists(out_path):
+        os.makedirs(out_path, exist_ok=True)
+    generate_labels(data_path, out_path)

scripts/export_to_onnx.py

@@ -0,0 +1,118 @@
+import argparse
+import os
+
+import numpy as np
+import onnx
+import onnxruntime
+import torch
+from monai.networks.nets import SEResNet50
+
+device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+
+
+def load_model_and_export(modelname, outname, out_channels, height, width, multigpu=False, in_channels=3):
+    """
+    Loading a model by name.
+
+    Args:
+        modelname: a whole path name of the model that need to be loaded.
+        outname: a name for output onnx model.
+        out_channels: output channels, which usually equals to 1 + class_number.
+        height: input images' height.
+        width: input images' width.
+        multigpu: if the pre-trained model trained on a multigpu environment.
+        in_channels: input images' channel number.
+    """
+    isopen = os.path.exists(modelname)
+    if not isopen:
+        raise Exception("The specified model to load does not exist!")
+
+    model = SEResNet50(spatial_dims=2, in_channels=in_channels, num_classes=out_channels)
+
+    if multigpu:
+        model = torch.nn.DataParallel(model)
+    model = model.cuda()
+    model.load_state_dict(torch.load(modelname, map_location=device))  # if the model is trained on multi gpu
+    model = model.eval()
+
+    np.random.seed(0)
+    x = np.random.random((1, 3, width, height))
+    x = torch.tensor(x, dtype=torch.float32)
+    x = x.cuda()
+    torch_out = model(x)
+    input_names = ["INPUT__0"]
+    output_names = ["OUTPUT__0"]
+    # Export the model
+    if multigpu:
+        model_trans = model.module
+    else:
+        model_trans = model
+    torch.onnx.export(
+        model_trans,  # model to save
+        x,  # model input
+        outname,  # model save path
+        export_params=True,
+        verbose=True,
+        do_constant_folding=True,
+        input_names=input_names,
+        output_names=output_names,
+        opset_version=15,
+        dynamic_axes={"INPUT__0": {0: "batch_size"}, "OUTPUT__0": {0: "batch_size"}},
+    )
+    onnx_model = onnx.load(outname)
+    onnx.checker.check_model(onnx_model, full_check=True)
+    ort_session = onnxruntime.InferenceSession(outname)
+
+    def to_numpy(tensor):
+        return tensor.detach().cpu().numpy() if tensor.requires_grad else tensor.cpu().numpy()
+
+    # compute ONNX Runtime output prediction
+    ort_inputs = {ort_session.get_inputs()[0].name: to_numpy(x)}
+    ort_outs = ort_session.run(["OUTPUT__0"], ort_inputs)
+    numpy_torch_out = to_numpy(torch_out)
+    # compare ONNX Runtime and PyTorch results
+    np.testing.assert_allclose(numpy_torch_out, ort_outs[0], rtol=1e-03, atol=1e-05)
+    print("Exported model has been tested with ONNXRuntime, and the result looks good!")
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    # the original model for converting.
+    parser.add_argument(
+        "--model",
+        type=str,
+        default=r"/workspace/bundle/endoscopic_inbody_classification/models/model.pt",
+        help="Input an existing model weight",
+    )
+
+    # path to save the onnx model.
+    parser.add_argument(
+        "--outpath",
+        type=str,
+        default=r"/workspace/bundle/endoscopic_inbody_classification/models/model.onnx",
+        help="A path to save the onnx model.",
+    )
+
+    parser.add_argument("--width", type=int, default=256, help="Width for exporting onnx model.")
+
+    parser.add_argument("--height", type=int, default=256, help="Height for exporting onnx model.")
+
+    parser.add_argument(
+        "--out_channels", type=int, default=2, help="Number of expected out_channels in model for exporting to onnx."
+    )
+
+    parser.add_argument("--multigpu", type=bool, default=False, help="If loading model trained with multi gpu.")
+
+    args = parser.parse_args()
+    modelname = args.model
+    outname = args.outpath
+    out_channels = args.out_channels
+    height = args.height
+    width = args.width
+    multigpu = args.multigpu
+
+    if os.path.exists(outname):
+        raise Exception(
+            "The specified outpath already exists! Change the outpath to avoid overwriting your saved model. "
+        )
+    model = load_model_and_export(modelname, outname, out_channels, height, width, multigpu)