MSCT applies multi-scale self-attention and differential cross-modal attention within a transformer encoder to improve feature extraction and alignment for audio-visual deepfake detection, showing competitive results on FakeAVCeleb.
Emotions don’t lie: An audio-visual deepfake detection method using affec- tive cues,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
MSCT: Differential Cross-Modal Attention for Deepfake Detection
MSCT applies multi-scale self-attention and differential cross-modal attention within a transformer encoder to improve feature extraction and alignment for audio-visual deepfake detection, showing competitive results on FakeAVCeleb.