ROSCOSMOS commited on
Commit
3f2668d
·
verified ·
1 Parent(s): a45dfc6

Upload 11 files

Browse files
README.md CHANGED
@@ -1,3 +1,130 @@
1
  ---
2
  license: mit
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: mit
3
+ pretty_name: "Bridges"
4
+ tags: ["image", "computer-vision", "bridge", "bridges", "high-resolution"]
5
+ task_categories: ["image-classification"]
6
+ language: ["en"]
7
+ configs:
8
+ - config_name: default
9
+ data_files: "train/**/*.arrow"
10
+ features:
11
+ - name: image
12
+ dtype: image
13
+ - name: unique_id
14
+ dtype: string
15
+ - name: width
16
+ dtype: int32
17
+ - name: height
18
+ dtype: int32
19
+ - name: image_mode_on_disk
20
+ dtype: string
21
+ - name: original_file_format
22
+ dtype: string
23
+ - config_name: preview
24
+ data_files: "preview/**/*.arrow"
25
+ features:
26
+ - name: image
27
+ dtype: image
28
+ - name: unique_id
29
+ dtype: string
30
+ - name: width
31
+ dtype: int32
32
+ - name: height
33
+ dtype: int32
34
+ - name: original_file_format
35
+ dtype: string
36
+ - name: image_mode_on_disk
37
+ dtype: string
38
  ---
39
+
40
+ # Bridges
41
+
42
+ High resolution image subset from the Aesthetic-Train-V2 dataset, contains a collection of bridges from various parts of the world including many iconic landmark bridges.
43
+
44
+ ## Dataset Details
45
+
46
+ * **Curator:** Roscosmos
47
+ * **Version:** 1.0.0
48
+ * **Total Images:** 760
49
+ * **Average Image Size (on disk):** ~5.7 MB compressed
50
+ * **Primary Content:** Bridges
51
+ * **Standardization:** All images are standardized to RGB mode and saved at 95% quality for consistency.
52
+
53
+ ## Dataset Creation & Provenance
54
+
55
+ ### 1. Original Master Dataset
56
+ This dataset is a subset derived from:
57
+ **`zhang0jhon/Aesthetic-Train-V2`**
58
+ * **Link:** https://huggingface.co/datasets/zhang0jhon/Aesthetic-Train-V2
59
+ * **Providence:** Large-scale, high-resolution image dataset, refer to its original dataset card for full details.
60
+ * **Original License:** MIT
61
+
62
+ ### 2. Iterative Curation Methodology
63
+
64
+ CLIP retrieval / manual curation.
65
+
66
+ ## Dataset Structure & Content
67
+
68
+ This dataset offers the following configurations/subsets:
69
+ * **Default (Full `train` data) configuration:** Contains the full, high-resolution image data and associated metadata. This is the recommended configuration for model training and full data analysis. The default split for this configuration is `train`.
70
+ Each example (row) in the dataset contains the following fields:
71
+
72
+ * `image`: The actual image data. In the default (full) configuration, this is full-resolution. In the preview configuration, this is a viewer-compatible version.
73
+ * `unique_id`: A unique identifier assigned to each image.
74
+ * `width`: The width of the image in pixels (from the full-resolution image).
75
+ * `height`: The height of the image in pixels (from the full-resolution image).
76
+
77
+ ## Usage
78
+
79
+ To download and load this dataset from the Hugging Face Hub:
80
+
81
+ ```python
82
+
83
+ from datasets import load_dataset, Dataset, DatasetDict
84
+
85
+ # Login using e.g. `huggingface-cli login` to access this dataset
86
+
87
+ # To load the full, high-resolution dataset (recommended for training):
88
+ # This will load the 'default' configuration's 'train' split.
89
+ ds_main = load_dataset("ROSCOSMOS/Bridges", "default")
90
+
91
+ print("Main Dataset (default config) loaded successfully!")
92
+ print(ds_main)
93
+ print(f"Type of loaded object: {type(ds_main)}")
94
+
95
+ if isinstance(ds_main, Dataset):
96
+ print(f"Number of samples: {len(ds_main)}")
97
+ print(f"Features: {ds_main.features}")
98
+ elif isinstance(ds_main, DatasetDict):
99
+ print(f"Available splits: {list(ds_main.keys())}")
100
+ for split_name, dataset_obj in ds_main.items():
101
+ print(f" Split '{split_name}': {len(dataset_obj)} samples")
102
+ print(f" Features of '{split_name}': {dataset_obj.features}")
103
+
104
+
105
+ ```
106
+
107
+ ## Citation
108
+
109
+ ```bibtex
110
+ @inproceedings{zhang2025diffusion4k,
111
+ title={Diffusion-4K: Ultra-High-Resolution Image Synthesis with Latent Diffusion Models},
112
+ author={Zhang, Jinjin and Huang, Qiuyu and Liu, Junjie and Guo, Xiefan and Huang, Di},
113
+ year={2025},
114
+ booktitle={IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
115
+ }
116
+ @misc{zhang2025ultrahighresolutionimagesynthesis,
117
+ title={Ultra-High-Resolution Image Synthesis: Data, Method and Evaluation},
118
+ author={Zhang, Jinjin and Huang, Qiuyu and Liu, Junjie and Guo, Xiefan and Huang, Di},
119
+ year={2025},
120
+ note={arXiv:2506.01331},
121
+ }
122
+ ```
123
+
124
+ ## Disclaimer and Bias Considerations
125
+
126
+ Please consider any inherent biases from the original dataset and those potentially introduced by the automated filtering (e.g., CLIP's biases) and manual curation process.
127
+
128
+ ## Contact
129
+
130
+ N/A
train/data-00000-of-00010.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:68eb7c57ac59b03666b57c266f22b60838a687d740c72cf08ae8ab21fdf91229
3
+ size 445345824
train/data-00001-of-00010.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1a80a7398edee472ad093f2afa1dc90273c97910e27e5b57b2ed9ba05fd439db
3
+ size 480286488
train/data-00002-of-00010.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ebd7368a3bdad9727a7a667b64536e026bc08938de1ee980275cb3985f1d5a13
3
+ size 512299872
train/data-00003-of-00010.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:348bfdf2525a3a6d3b609bf354e4a8e55f84383dbf72d40425e6f75620298c3f
3
+ size 458174304
train/data-00004-of-00010.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:93cdcc1c5e812c5655d64721bb21bf88a50fb0d33e7a9218f4645472144b89c4
3
+ size 485587200
train/data-00005-of-00010.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fb8753dfec523956772ea5e8f12ba499117f19f6543d5a4fcc55cb1d3962f421
3
+ size 511336704
train/data-00006-of-00010.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aa1852c11c9eed8fce4809a5a81b37671ea3636039626b3c037ab7432011c674
3
+ size 465736664
train/data-00007-of-00010.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:da6f5e6774ebb05ed412c1c00ea8c77db44157fe56de3c370935e078e5965a33
3
+ size 428060912
train/data-00008-of-00010.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7b1dbab4b431eaa53789023cb1e861d9f88fe5d049637acc2f3532ea117737d7
3
+ size 412395600
train/data-00009-of-00010.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:39a72e01c5b82b35e834c5ee424b3e0835fde336afbfa13fe50021c4b94489ae
3
+ size 441543392