Fine-grained MLLM-generated forgery descriptions plus dual global/fine-grained visual-text alignment yield more generalizable face-attack detectors than vision-only or coarse-text methods.
Mfclip: Multi-modal fine-grained clip for generalizable diffusion face forgery detection,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Unified Face Attack Detection via Fine-Grained Semantic Guidance
Fine-grained MLLM-generated forgery descriptions plus dual global/fine-grained visual-text alignment yield more generalizable face-attack detectors than vision-only or coarse-text methods.