Search papers, labs, and topics across Lattice.
6
0
7
6
ClASH (Controlled Lexical-Acoustic Separation Harness), a bilingual counterfactual diagnostic framework that evaluates each utterance under original, lexical-preserving, prosody-preserving, and approximately neutralised conditions, is introduced.
Despite advances in large audio language models, the best performer in temporal audio grounding only achieves a mere 31.2 mIoU, revealing critical limitations in current capabilities.
Heavy speaker overlap drastically hinders speech recognition accuracy in smart glasses, revealing critical limitations in current audio-language models.
Achieving high accuracy in multi-speaker transcription, SoulX-Transcriber outperforms existing models by effectively addressing speaker overlap and rapid turn-taking.
3D reconstruction models can self-improve at test time by "hallucinating" better versions of themselves from more complete viewpoints, sidestepping the need for costly 3D ground truth.
Achieve human-like full-duplex voice interactions with SoulX-Duplug, a plug-and-play module that slashes latency and improves turn management by acting as a semantic VAD.
