Target-speaker Automatic Speech Recognition on NOTSOFAR Small-SDM 1
15.8tcpWERSE-DiCoW
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| SE-DiCoWDiarization Type=oracle2026.01 | 15.8 | — | — | |
| DiCoW v3.3Diarization Type=oracle2026.01 | 16 | — | — | |
| + new model init..Diarization Type=oracle2026.01 | 16.6 | — | — | |
| + flexible data seg.Diarization Type=oracle2026.01 | 17.6 | — | — | |
| DiCoW*Diarization Type=oracle2026.01 | 19.6 | — | — | |
| SOTADiarization Type=DiariZen2026.01 | 23.6 | 12.6 | 0.37 | |
| SE-DiCoWDiarization Type=DiariZen2026.01 | 26.1 | 12.6 | 0.37 | |
| DiCoW v3.3Diarization Type=DiariZen2026.01 | 26.6 | 12.6 | 0.37 | |
| DiCoW v3.3Model Category=Specialized MT-ASR Models, Diarization System=DiariZen2026.06 | 26.6 | — | — | |
| DixtralModel Category=General Purpose Spoken LMs, Diarization System=DiariZen2026.06 | 29.1 | — | — | |
| DiCoW*Diarization Type=DiariZen2026.01 | 29.8 | 12.6 | 0.37 | |
| VibeVoiceModel Category=Specialized MT-ASR Models2026.06 | 35.8 | — | — | |
| Gemini 3.0 FlashModel Category=General Purpose Spoken LMs2026.06 | 39.1 | — | — | |
| Voxtral MTv2Model Category=Specialized MT-ASR Models2026.06 | 54.4 | — | — |