Search papers, labs, and topics across Lattice.
1
0
2
Forget MLLMs: self-supervised speech models are surprisingly good at capturing subtle differences in voice impressions, outperforming traditional acoustic features.
