albertvillanova HF staff commited on
Commit
65bc9e7
1 Parent(s): 33a5df5

Fix broken links (#5)

Browse files

- Update homepage URL in dataset script and card (03836b594f4494ec5032b4d52639e3f831848023)
- Update paper URL in dataset card (74b2b6abd56265ce44c593474997cfe125d19962)
- Update citation information (5ba9dda2e99dfa3546698490647aab9ce476e526)
- Remove legacy metadata dataset_infos.json (eb3ccf77a8e641ee9fe5a9a26adeec82102c8547)
- Remove script to generate dummy data (b26f59b18020ddb34fca357d9ca939861d793ab8)

README.md CHANGED
@@ -96,9 +96,9 @@ config_names:
96
 
97
  ## Dataset Description
98
 
99
- - **Homepage:** https://visualgenome.org/
100
  - **Repository:**
101
- - **Paper:** https://visualgenome.org/static/paper/Visual_Genome.pdf
102
  - **Leaderboard:**
103
  - **Point of Contact:** ranjaykrishna [at] gmail [dot] com
104
 
@@ -484,11 +484,15 @@ Visual Genome by Ranjay Krishna is licensed under a Creative Commons Attribution
484
  ### Citation Information
485
 
486
  ```bibtex
487
- @inproceedings{krishnavisualgenome,
488
  title={Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image Annotations},
489
- author={Krishna, Ranjay and Zhu, Yuke and Groth, Oliver and Johnson, Justin and Hata, Kenji and Kravitz, Joshua and Chen, Stephanie and Kalantidis, Yannis and Li, Li-Jia and Shamma, David A and Bernstein, Michael and Fei-Fei, Li},
490
- year = {2016},
491
- url = {https://arxiv.org/abs/1602.07332},
 
 
 
 
492
  }
493
  ```
494
 
 
96
 
97
  ## Dataset Description
98
 
99
+ - **Homepage:** https://homes.cs.washington.edu/~ranjay/visualgenome/
100
  - **Repository:**
101
+ - **Paper:** https://doi.org/10.1007/s11263-016-0981-7
102
  - **Leaderboard:**
103
  - **Point of Contact:** ranjaykrishna [at] gmail [dot] com
104
 
 
484
  ### Citation Information
485
 
486
  ```bibtex
487
+ @article{Krishna2016VisualGC,
488
  title={Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image Annotations},
489
+ author={Ranjay Krishna and Yuke Zhu and Oliver Groth and Justin Johnson and Kenji Hata and Joshua Kravitz and Stephanie Chen and Yannis Kalantidis and Li-Jia Li and David A. Shamma and Michael S. Bernstein and Li Fei-Fei},
490
+ journal={International Journal of Computer Vision},
491
+ year={2017},
492
+ volume={123},
493
+ pages={32-73},
494
+ url={https://doi.org/10.1007/s11263-016-0981-7},
495
+ doi={10.1007/s11263-016-0981-7}
496
  }
497
  ```
498
 
dataset_infos.json DELETED
@@ -1 +0,0 @@
1
- {"region_descriptions_v1.0.0": {"description": "Visual Genome enable to model objects and relationships between objects.\nThey collect dense annotations of objects, attributes, and relationships within each image.\nSpecifically, the dataset contains over 108K images where each image has an average of 35 objects, 26 attributes, and 21 pairwise relationships between objects.\n", "citation": "@inproceedings{krishnavisualgenome,\n title={Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image Annotations},\n author={Krishna, Ranjay and Zhu, Yuke and Groth, Oliver and Johnson, Justin and Hata, Kenji and Kravitz, Joshua and Chen, Stephanie and Kalantidis, Yannis and Li, Li-Jia and Shamma, David A and Bernstein, Michael and Fei-Fei, Li},\n year = {2016},\n url = {https://arxiv.org/abs/1602.07332},\n}\n", "homepage": "https://visualgenome.org/", "license": "Creative Commons Attribution 4.0 International License", "features": {"image": {"decode": true, "id": null, "_type": "Image"}, "image_id": {"dtype": "int32", "id": null, "_type": "Value"}, "url": {"dtype": "string", "id": null, "_type": "Value"}, "width": {"dtype": "int32", "id": null, "_type": "Value"}, "height": {"dtype": "int32", "id": null, "_type": "Value"}, "coco_id": {"dtype": "int64", "id": null, "_type": "Value"}, "flickr_id": {"dtype": "int64", "id": null, "_type": "Value"}, "regions": [{"region_id": {"dtype": "int32", "id": null, "_type": "Value"}, "image_id": {"dtype": "int32", "id": null, "_type": "Value"}, "phrase": {"dtype": "string", "id": null, "_type": "Value"}, "x": {"dtype": "int32", "id": null, "_type": "Value"}, "y": {"dtype": "int32", "id": null, "_type": "Value"}, "width": {"dtype": "int32", "id": null, "_type": "Value"}, "height": {"dtype": "int32", "id": null, "_type": "Value"}}]}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "visual_genome", "config_name": "region_descriptions_v1.0.0", "version": {"version_str": "1.0.0", "description": null, "major": 1, "minor": 0, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 260873884, "num_examples": 108077, "dataset_name": "visual_genome"}}, "download_checksums": {"https://visualgenome.org/static/data/dataset/image_data.json.zip": {"num_bytes": 1780854, "checksum": "b87a94918cb2ff4d952cf1dfeca0b9cf6cd6fd204c2f8704645653be1163681a"}, "https://visualgenome.org/static/data/dataset/region_descriptions_v1.json.zip": {"num_bytes": 99460401, "checksum": "9e54cd76082f7ce5168a1779fc8c3d0629720492ec0fa12f2d8339c3e0dc9734"}, "https://cs.stanford.edu/people/rak248/VG_100K_2/images.zip": {"num_bytes": 9731705982, "checksum": "51c682d2721f880150720bb416e0346a4c787e4c55d7f80dfd1bd3f73ba81646"}, "https://cs.stanford.edu/people/rak248/VG_100K_2/images2.zip": {"num_bytes": 5471658058, "checksum": "99da1a0ddf87011319ff3b05cf9176ffee2731cc3c52951162d9ef0d68e3cfb5"}}, "download_size": 15304605295, "post_processing_size": null, "dataset_size": 260873884, "size_in_bytes": 15565479179}}
 
 
fix_generated_dummy_data.py DELETED
@@ -1,48 +0,0 @@
1
- import json
2
- import re
3
- from pathlib import Path
4
- from zipfile import ZipFile
5
-
6
-
7
- def main():
8
- dummy_dir = Path(__file__).parent / "dummy"
9
- config_paths = list(dummy_dir.iterdir())
10
-
11
- for config in config_paths:
12
- versions = list(config.iterdir())
13
- assert len(versions) == 1, versions
14
- version = versions[0]
15
- zip_filepath = version / "dummy_data.zip"
16
-
17
- # We need to open the zip file
18
- with ZipFile(zip_filepath, "r") as zip_dir:
19
- with zip_dir.open("dummy_data/image_data.json.zip/image_data.json", "r") as fi:
20
- image_metadatas = json.load(fi)
21
-
22
- default_jpg_path = Path(__file__).parent / "huggingface.jpg"
23
- with ZipFile(zip_filepath, "a") as zip_dir:
24
- for image_metadata in image_metadatas:
25
- url = image_metadata["url"]
26
-
27
- matches = re.match(r"https://cs.stanford.edu/people/rak248/VG_100K(?:_(2))?/[0-9]+.jpg", url)
28
- assert matches is not None
29
-
30
- # Find where locally the images should be
31
- vg_version = matches.group(1)
32
- if vg_version is None:
33
- local_path = re.sub(
34
- "https://cs.stanford.edu/people/rak248/VG_100K", "dummy_data/images.zip/VG_100K", url
35
- )
36
- else:
37
- local_path = re.sub(
38
- f"https://cs.stanford.edu/people/rak248/VG_100K_{vg_version}",
39
- f"dummy_data/images{vg_version}.zip/VG_100K_{vg_version}",
40
- url,
41
- )
42
-
43
- # Write those images.
44
- zip_dir.write(filename=default_jpg_path, arcname=local_path)
45
-
46
-
47
- if __name__ == "__main__":
48
- main()
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
huggingface.jpg DELETED
Binary file (656 Bytes)
 
visual_genome.py CHANGED
@@ -27,11 +27,15 @@ import datasets
27
  logger = datasets.logging.get_logger(__name__)
28
 
29
  _CITATION = """\
30
- @inproceedings{krishnavisualgenome,
31
  title={Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image Annotations},
32
- author={Krishna, Ranjay and Zhu, Yuke and Groth, Oliver and Johnson, Justin and Hata, Kenji and Kravitz, Joshua and Chen, Stephanie and Kalantidis, Yannis and Li, Li-Jia and Shamma, David A and Bernstein, Michael and Fei-Fei, Li},
33
- year = {2016},
34
- url = {https://arxiv.org/abs/1602.07332},
 
 
 
 
35
  }
36
  """
37
 
@@ -41,7 +45,7 @@ They collect dense annotations of objects, attributes, and relationships within
41
  Specifically, the dataset contains over 108K images where each image has an average of 35 objects, 26 attributes, and 21 pairwise relationships between objects.
42
  """
43
 
44
- _HOMEPAGE = "https://visualgenome.org/"
45
 
46
  _LICENSE = "Creative Commons Attribution 4.0 International License"
47
 
 
27
  logger = datasets.logging.get_logger(__name__)
28
 
29
  _CITATION = """\
30
+ @article{Krishna2016VisualGC,
31
  title={Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image Annotations},
32
+ author={Ranjay Krishna and Yuke Zhu and Oliver Groth and Justin Johnson and Kenji Hata and Joshua Kravitz and Stephanie Chen and Yannis Kalantidis and Li-Jia Li and David A. Shamma and Michael S. Bernstein and Li Fei-Fei},
33
+ journal={International Journal of Computer Vision},
34
+ year={2017},
35
+ volume={123},
36
+ pages={32-73},
37
+ url={https://doi.org/10.1007/s11263-016-0981-7},
38
+ doi={10.1007/s11263-016-0981-7}
39
  }
40
  """
41
 
 
45
  Specifically, the dataset contains over 108K images where each image has an average of 35 objects, 26 attributes, and 21 pairwise relationships between objects.
46
  """
47
 
48
+ _HOMEPAGE = "https://homes.cs.washington.edu/~ranjay/visualgenome/"
49
 
50
  _LICENSE = "Creative Commons Attribution 4.0 International License"
51