Enhanced Parquet Edition: NIST16

This repository packages the NIST16 evaluation assets distributed by ForgeryVCR into Hugging Face Parquet format. It contains 1,440 images: 876 authentic and 564 tampered, with image bytes embedded in Parquet and the supplied pixel masks.

These are the source release's 832-size preprocessed copies. The supplied image and mask bytes are preserved. This source supplies no NIST16 train/validation partition; the authentic and tampered archives form one full split.

Subsets

Config	Split	Rows	Image size	Pixel masks	Parquet files
nist16	full	1,440	832 × 832	Tampered images only	5

Columns

• image: the supplied RGB image, decoded with the Hugging Face Image feature.
• mask: the supplied grayscale pixel annotation. Black 0 marks background and white 255 marks edited regions; intermediate grayscale values are retained. Authentic images have no supplied mask and use null.
• label: image-level authenticity label, 0 = real, 1 = fake.
• manipulation: real or tampered; specific editing methods are not assigned by this release.
• sample_id, source_image_path: the image path relative to the extracted source release.
• source_mask_path: the mask path relative to the extracted source release, or null for authentic images.
• source_split: test, following the source release's evaluation directory.
• filename_source_id: the identifier from the supplied image filename.
• width, height: supplied image dimensions in pixels.
• has_mask: whether a pixel annotation is supplied.

Metadata Files

The nist16/stats.json file records split counts, authenticity counts, image dimensions, mask statistics, and Parquet shard paths and sizes.

The root inventory.json collects these statistics and records total row counts and encoded byte sizes. provenance/ retains the source manifests and release README.

Citation

Please acknowledge the original NIST/OpenMFC benchmark and the ForgeryVCR release used for this edition:

@inproceedings{wang2026forgeryvcr,
  title={ForgeryVCR: Visual-Centric Reasoning via Efficient Forensic Tools in MLLMs for Image Forgery Detection and Localization},
  author={Wang, Youqi and Chen, Shen and Wang, Haowei and Peng, Rongxuan and Yao, Taiping and Tan, Shunquan and Chen, Changsheng and Li, Bin and Ding, Shouhong},
  booktitle={Proceedings of the 34th ACM International Conference on Multimedia},
  year={2026}
}

Dataset Export

This Parquet edition reorganizes the source release and adds per-sample metadata. Refer to NIST/OpenMFC and the source release for the original data terms. Please also reference this dataset repository when using the packaged version.

@misc{thanejoss_nist16_2026,
  author={{ThaneJoss}},
  title={{NIST16}},
  year={2026},
  publisher={Hugging Face},
  url={https://huggingface.co/datasets/ThaneJoss/NIST16},
  note={Dataset available on Hugging Face}
}