Search papers, labs, and topics across Lattice.
1
0
2
4
Fusing speech and environmental sound representations with a novel matching head and cross-attention network significantly boosts deepfake audio detection, surpassing previous benchmarks.
