FaceForensics++

Deepfakes have become a critical social problem, and detecting them is of utmost importance. The FaceForensics++ dataset offers fake video datasets. Most of the detection methods generally perform well only in detecting fake images generated in the same domain (i.e., training and testing data are created with the same method), but not in detecting those generated in a different domain (i.e., test deepfakes are generated with different methods).

Data and Resources

Cite this as

Rossler et al. (2024). Dataset: FaceForensics++. https://doi.org/10.57702/8ht5lray

DOI retrieved: December 2, 2024

Additional Info

Field Value
Created December 2, 2024
Last update December 2, 2024
Defined In https://doi.org/10.48550/arXiv.2103.16076
Citation
  • https://doi.org/10.48550/arXiv.2307.07928
  • https://doi.org/10.48550/arXiv.2308.08011
  • https://doi.org/10.48550/arXiv.2403.06592
  • https://doi.org/10.48550/arXiv.2103.14211
  • https://doi.org/10.48550/arXiv.2008.04848
  • https://doi.org/10.48550/arXiv.2403.05094
  • https://doi.org/10.48550/arXiv.2110.15561
  • https://doi.org/10.48550/arXiv.1901.08971
  • https://doi.org/10.48550/arXiv.2108.12715
  • https://doi.org/10.48550/arXiv.2103.02406
  • https://doi.org/10.48550/arXiv.2105.06117
  • https://doi.org/10.48550/arXiv.2404.05583
  • https://doi.org/10.48550/arXiv.2402.00541
  • https://doi.org/10.1145/3448017.3457387
Author Rossler et al.
Homepage https://github.com/ondyari/FaceForensics