Yaoyao Chang

Streaming ASR Unifies Transcription and Speaker Attribution

VibeVoice-ASR-Streaming interleaves audio, lookahead, and prior text so one model can produce speaker-attributed transcripts as speech arrives.