AuthGuard trains a deepfake vision encoder with MLLM-generated text descriptions plus uncertainty-weighted contrastive learning, improving cross-dataset deepfake detection and adding interpretable LLM reasoning.
End-to-end reconstruction- classification learning for face forgery detection
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
AuthGuard: Generalizable Deepfake Detection via Language Guidance
AuthGuard trains a deepfake vision encoder with MLLM-generated text descriptions plus uncertainty-weighted contrastive learning, improving cross-dataset deepfake detection and adding interpretable LLM reasoning.