Deepfake Detection

In this project, I focused on developing advanced methods to detect high-quality and unseen deepfakes, including those not encountered during training. Our approaches not only demonstrated superior performance on multiple datasets compared to state-of-the-art methods but also showed greater robustness to unseen common perturbations (e.g., Gaussian noise). Notably, our models require minimal computational overhead, underscoring their practical usefulness. This work has resulted in three publications, including two accepted at top-tier conferences in Computer Vision.

Please find more details of our algorithm with our 3 publications below:

(Nguyen et al., 2025) and (Nguyen et al., 2024) and (Nguyen et al., 2024)

References

2025

Vulnerability-Aware Spatio-Temporal Learning for Generalizable Deepfake Video Detection

Dat Nguyen, Marcella Astrid , Anis Kacem , and 2 more authors

In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) , Oct 2025

arXiv Code

2024

Fakeformer: Efficient vulnerability-driven transformers for generalisable deepfake detection

Dat Nguyen, Marcella Astrid , Enjie Ghorbel , and 1 more author

In arXiv preprint arXiv:2410.21964 , Oct 2024

arXiv Code

LAA-Net: Localized Artifact Attention Network for Quality-Agnostic and Generalizable Deepfake Detection

Dat Nguyen, Nesryne Mejri , Inder Pal Singh , and 5 more authors

In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , Jun 2024

arXiv Bib PDF Supp Code Poster

@inproceedings{Nguyen_2024_CVPR,
  author = {Nguyen, Dat and Mejri, Nesryne and Singh, Inder Pal and Kuleshova, Polina and Astrid, Marcella and Kacem, Anis and Ghorbel, Enjie and Aouada, Djamila},
  title = {LAA-Net: Localized Artifact Attention Network for Quality-Agnostic and Generalizable Deepfake Detection},
  booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
  month = jun,
  year = {2024},
  pages = {17395-17405},
}