Search papers, labs, and topics across Lattice.
1
0
3
By better aligning audio and visual features, MSCT significantly boosts deepfake detection, outperforming existing methods on the FakeAVCeleb dataset.
