Speech Recognition on Simulated FLEURS (Wearer: English, Partner: Italian) (test)
13.3Wearer WER [%]SS+SLM
Evaluation results
| Method | Links | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| SS+SLM2026.02 | 13.3 | 1.6 | 6.4 | 5.3 | 0 | 10.2 | 1 | 5.2 | 3.9 | 0 | |
| Multi-channel ASR2026.02 | 16.1 | 1.4 | 5.8 | 8.9 | 0 | 22.3 | 0.8 | 14.6 | 6.8 | 0 | |
| JSTAR2026.02 | 16.5 | 1.7 | 5.6 | 9.1 | 0 | 17.6 | 1 | 7.9 | 8.6 | 0 | |
| Multi-channel SOT+SLM2026.02 | 16.8 | 4.3 | 6.3 | 5.7 | 0 | 15.4 | 1.3 | 6.8 | 5.1 | 2.3 |