JamalLee/Omni-Fake-OOD
Omni-Fake-OOD Omni-Fake-OOD is the out-of-distribution benchmark split of Omni-Fake. Samples come from held-out generators and platforms not included in training, for measuring cross-domain generalization. It covers image, audio, video, and audio–video talking-head (AV-TH) with the same three-class labels as Omni-Fake-SET: real, fully synthetic, and tampered. Use together with Omni-Fake-SET (in-distribution training data). Paper: arXiv:2605.01638 Project page: Omni-Fake… See the full description on the dataset page: https://huggingface.co/datasets/JamalLee/Omni-Fake-OOD.
Video hybrid release: GenBuster .7z + tampered parquet
Upload data/Video/test-00020-of-00040.parquet with huggingface_hub
Upload data/Video/test-00007-of-00040.parquet with huggingface_hub
Upload data/Video/test-00005-of-00040.parquet with huggingface_hub
Upload data/Video/test-00016-of-00040.parquet with huggingface_hub
Upload data/Video/test-00000-of-00040.parquet with huggingface_hub
Upload data/Video/test-00039-of-00040.parquet with huggingface_hub
Upload data/Video/test-00038-of-00040.parquet with huggingface_hub
Upload data/Video/test-00037-of-00040.parquet with huggingface_hub
Upload data/Video/test-00036-of-00040.parquet with huggingface_hub
Upload data/Video/test-00034-of-00040.parquet with huggingface_hub
Upload data/Video/test-00035-of-00040.parquet with huggingface_hub
Upload data/Video/test-00001-of-00040.parquet with huggingface_hub
Upload data/Video/test-00033-of-00040.parquet with huggingface_hub
Upload data/Video/test-00032-of-00040.parquet with huggingface_hub
Upload data/Video/test-00031-of-00040.parquet with huggingface_hub
Upload data/Video/test-00029-of-00040.parquet with huggingface_hub
Upload data/Video/test-00030-of-00040.parquet with huggingface_hub
Upload data/Video/test-00027-of-00040.parquet with huggingface_hub
Upload data/Video/test-00028-of-00040.parquet with huggingface_hub
Upload data/Video/test-00026-of-00040.parquet with huggingface_hub
Upload data/Video/test-00015-of-00040.parquet with huggingface_hub
Upload data/Video/test-00025-of-00040.parquet with huggingface_hub
Upload data/Video/test-00024-of-00040.parquet with huggingface_hub
Upload data/Video/test-00023-of-00040.parquet with huggingface_hub
Upload data/Video/test-00022-of-00040.parquet with huggingface_hub
Upload data/Video/test-00021-of-00040.parquet with huggingface_hub
Upload data/Video/test-00009-of-00040.parquet with huggingface_hub
Upload data/Video/test-00019-of-00040.parquet with huggingface_hub
Upload data/Video/test-00018-of-00040.parquet with huggingface_hub
Upload data/Video/test-00017-of-00040.parquet with huggingface_hub
Upload data/Video/test-00014-of-00040.parquet with huggingface_hub
Upload data/Video/test-00012-of-00040.parquet with huggingface_hub
Upload data/Video/test-00013-of-00040.parquet with huggingface_hub
Upload data/Video/test-00010-of-00040.parquet with huggingface_hub
Upload data/Video/test-00008-of-00040.parquet with huggingface_hub
Upload data/Video/test-00011-of-00040.parquet with huggingface_hub
Upload data/Video/test-00002-of-00040.parquet with huggingface_hub
Upload data/Video/test-00006-of-00040.parquet with huggingface_hub
Upload data/Video/test-00004-of-00040.parquet with huggingface_hub
Upload data/Video/test-00003-of-00040.parquet with huggingface_hub
Remove bold from intro paragraph
Add Video modality to dataset card (three-class labels)
Remove Overview section from dataset card
Enable image preview in Dataset Viewer (default image config)
Update dataset card: standard Omni-Fake intro
Upload data/Audio/test-00046-of-00050.parquet with huggingface_hub
Upload data/Audio/test-00048-of-00050.parquet with huggingface_hub
Upload data/Audio/test-00049-of-00050.parquet with huggingface_hub
Upload data/Audio/test-00045-of-00050.parquet with huggingface_hub