Xinsheng Wang

Achieving high accuracy in multi-speaker transcription, SoulX-Transcriber outperforms existing models by effectively addressing speaker overlap and rapid turn-taking.

Adversarial training doesn't have to hurt speaker verification: by explicitly modeling language, you can disentangle speaker and language characteristics without sacrificing speaker discriminability.

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.