| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| WildBench | STEP3-VL-10B | Score0.8604 | 19 | 9d ago | |
| DNS blind (test) | MOS4.8125 | 5 | 1mo ago | ||
| Subjective Evaluation Mandarin Accent | Accent Vector | Perceived Accuracy70.77 | 1 | 1mo ago | |
| Subjective Evaluation Hindi Accent | Accent Vector | Accuracy78.46 | 1 | 1mo ago | |
| Subjective Evaluation Spanish Accent | Accent Vector | Accuracy (%)53.85 | 1 | 1mo ago | |
| Subjective Evaluation French Accent | Accent Vector | Accuracy66.15 | 1 | 1mo ago | |
| Subjective Evaluation German Accent | Accent Vector | Accuracy53.85 | 1 | 1mo ago | |
| Subjective Evaluation British (England) Accent | Accent Vector | Accuracy78.46 | 1 | 1mo ago | |
| Subjective Evaluation US Accent | Accent Vector | Perceived Accuracy80 | 1 | 1mo ago | |
| Synthetic audiovisual pairs 1.0 (test) | - | - | 0 | 1mo ago |
