UNFAKE: Unsupervised multi-type explainable deepFAKE detection

Project description:

Given the threat of deepfakes, significant efforts have been made for proposing deepfake detection methods.  Nevertheless, these methods remain not sufficiently mature for real-world deployment. as they usually specialize in detecting one type of deepfakes, which limits their generalization capability, and typically rely on very large models. Hence, UNFAKE aims to provide a more realistic deepfake detection framework that generalizes across different types of deepfakes by using an unsupervised explainable and low-weight learning framework to learn richer deep representations. 

  • Starting date: 01/09/2021
  • Duration: 36 months + 12
  • Funding source: FNR IF
  • Researchers: Nesryne Mejri,  Dr. Enjie Ghorbel, Prof. Djamila Aouada
  • Partners: POST