Search papers, labs, and topics across Lattice.
Technical University of Berlin
1
0
2
Bridging the gap between audio embeddings and LLMs, this method achieves a remarkable 16.2% boost in deepfake voice detection accuracy on unseen domains.
