Hugging Face's logo

Datasets:
srishti-kaushik
/
COCO-2017

Tasks:
Object Detection
Image Segmentation
Image-to-Text
Modalities:
Image
Tabular
Text
Formats:
parquet
Sub-tasks:
instance-segmentation
semantic-segmentation
panoptic-segmentation
Languages:
English
Size:
1M - 10M
Tags:
coco
detection
segmentation
panoptic
keypoints
captioning
Libraries:
Datasets
Dask
Polars
License:
Dataset card Data Studio Files
xet
 Community
1
COCO-2017 / docs /VERIFICATION.md
srishti-kaushik's picture
srishti-kaushik
Rewrite card: minimal front page, sample grid, docs split out
a7739e9 verified
preview code
|
Raw History Blame Contribute Delete
2.36 kB

Verification

This copy was checked against the official COCO 2017 release rather than assumed correct. Every check below passed.

Image and mask integrity

Check Result
JPEG bytes SHA-256 identical to the official files 287,360 / 287,360
Mask PNG bytes SHA-256 identical to the official files 246,574 / 246,574
JPEGs fully decoded (not header-checked) 287,360, 0 corrupt
Decoded pixel dimensions vs. dimensions declared in the JSON 0 mismatches
Row count per split vs. files on disk exact for all splits

Image counts match the official release: 118,287 train / 5,000 val / 40,670 test / 123,403 unlabeled.

Annotation conversion is lossless

Every annotation was reconstructed from Parquet back into its original COCO JSON shape and compared field by field against the source file.

Family Annotations compared Mismatches
instances 896,782 0
captions 616,767 0
person_keypoints 273,469 0
stuff 780,259 0
panoptic 1,386,712 segments 0

This covers bbox, area, category_id, iscrowd, keypoints, num_keypoints, caption, and every segmentation in all three of its formats, plus per-image metadata (file_name, width, height, license, coco_url, flickr_url, date_captured) and the category tables for all five families.

Round-trip behaviour

  • image and mask columns decode as PIL images through datasets.
  • Reconstructed segmentations decode through pycocotools (2,000 sampled, 100%).
  • Annotation image_ids and image image_ids cover each other exactly, in both directions, for every split.
  • Every panoptic_file_name resolves to a mask present in panoptic_masks_*.

Reproducing

The scripts that built and checked this repo are in scripts/:

python scripts/build_annotations.py   # JSONs  -> annotation configs
python scripts/build_images.py        # JPEGs  -> image configs
python scripts/build_masks.py         # PNGs   -> mask configs
python scripts/verify_release.py      # full verification (byte + round-trip)
python scripts/validate.py            # counts, joins, feature decoding

verify_release.py accepts images, masks, annotations, features or all. It compares against a local copy of the official release; paths are set at the top of each script.