AI Tools & Models HOTFeaturedAll
arXiv cs.AIScore60

Exposing and Mitigating Temporal Attack in Deepfake Video Detection

Summary

The study reveals that spatiotemporal deepfake detectors, despite high AUC scores, are vulnerable to temporal evasion attacks, and it proposes mitigations for this weakness.

Why this matters

Covers model · Relevant to this topic

Open original ↗

Source: arXiv cs.AI · arxiv.org

Topics

Related items