I compared diarization models on 15 clinical conversations: Nemotron 3, Pyannote, Sortformer and VibeVoice

I've been working on clinical speaker attribution at Omi and wanted to compare the current diarization models on the same audio. I used 15 mock doctor–patient consultations from PriMock57, about 2.4 hours. Full recordings, automatic speaker counts, without telling the models there are two peop

中文全文翻译正在补齐，可先查看下方英文原文。

I've been working on clinical speaker attribution at Omi and wanted to compare the current diarization models on the same audio. I used 15 mock doctor–patient consultations from PriMock57, about 2.4 hours. Full recordings, automatic speaker counts, without telling the models there are two peop