xcatf / celeb-deepfakeforensics

Celeb-DF: A New Dataset for DeepFake Forensics

Celeb-DF: A Large-scale Challenging Dataset for DeepFake Forensics

Yuezun Li¹, Xin Yang¹, Pu Sun², Honggang Qi² and Siwei Lyu^1
1 University at Albany, State University of New York, USA; ² University of Chinese Academy of Sciences, China
[Project page]

Overview

Celeb-DF dataset includes 590 original videos collected from YouTube with subjects of different ages, ethic groups and genders, and 5639 corresponding DeepFake videos. See paper for details (full paper will be uploaded soon).

Download

If you would like to access the Celeb-DF dataset, please fill out this form . The download link will be sent to you once the form is accepted. If you have any questions, please send email to [email protected]

Please cite our paper in your publications if the Celeb-DF dataset is used in your research:

@inproceedings{Celeb_DF_cvpr20,
   author = {Yuezun Li, Xin Yang, Pu Sun, Honggang Qi and Siwei Lyu},
   title = {Celeb-DF: A Large-scale Challenging Dataset for DeepFake Forensics},
   booktitle= {IEEE Conference on Computer Vision and Patten Recognition (CVPR)},
   year = {2020}
}

Dataset Structure

Celeb-DF
|--- Celeb-real # 590 Celebrity videos downloaded from YouTube
|--- YouTube-real # 300 Additional videos downloaded from YouTube
|--- Celeb-synthesis # 5639 Synthesized videos from Celeb-real
|--- List_of_testing_videos.txt # 518 videos

Comparison with Other Datasets

See paper for details.

Average AUC Performance of SOTA Detection Methods on Each Dataset

See paper for details.

Privacy Statement

This dataset is released under the Terms to Use Celeb-DF, which is provided "as it is" and we are not responsible for any subsequence from using this dataset. All original videos of the Celeb-DF dataset are obtained from the Internet which are not property of the authors or the authors’ affiliated institutions. Neither the authors or the authors’ affiliated institution are responsible for the content nor the meaning of these videos. If you feel uncomfortable about your identity shown in this dataset, please contact us and we will remove corresponding information from the dataset.

Video Demo

Log

• The Celeb-DF (v2) is accepted to CVPR2020.
• Check here for previous Celeb-DF (v1) (795 DeepFake videos included).

Reference

_{[UADFV] Yang, Xin and Li, Yuezun and Lyu, Siwei. "Exposing Deep Fakes Using Inconsistent Head Poses", IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2019
[DeepFake-TIMIT] Korshunov, Pavel and Marcel, Sebastien. "DeepFakes: a New Threat to Face Recognition? Assessment and Detection", arXiv preprint arXiv:1812.08685, 2018
[FaceForensics++] Andreas Rössler and Davide Cozzolino and Luisa Verdoliva and Christian Riess and Justus Thies and Matthias Nießner. "FaceForensics++: Learning to Detect Manipulated Facial Images", IEEE International Conference on Computer Vision (ICCV), 2019
[DFD (Google)] Nicholas Dufour, Andrew Gully, Per Karlsson, Alexey Vic-tor Vorbyov, Thomas Leung, Jeremiah Childs, and ChristophBregler. Deepfakes detection dataset by google & jigsaw.
[DFDC (Facebook)] Brian Dolhansky, Russ Howes, Ben Pflaum, NicoleBaram, and Cristian Canton Ferrer. The deepfake detec-tion challenge (DFDC) preview dataset. arXiv preprintarXiv:1910.08854, 2019.}