Search papers, labs, and topics across Lattice.
Brno University of Technology
4
0
4
SphereVBx achieves superior clustering accuracy in speaker diarization while significantly simplifying the process, making it a game-changer for EEND-VC applications.
Dixtral achieves up to 29% absolute improvement in speaker-attributed transcription accuracy by leveraging diarization masks without risking catastrophic forgetting.
Integrating visual cues into a long-context ASR system slashes word error rate by 16% in multi-talker conversational recordings, proving the power of AV fusion.
Achieve single-pass alignment of multi-talker speech – a feat previously impossible – by modeling overlaps as shuffles.
