Spoken Question Answering on WebQ S→T
41.8AccuracyVITA 1.5
Evaluation Results
| Method | Links | |
|---|---|---|
| VITA 1.5Type=System-level2026.07 | 41.8 | |
| StepAudio-2-miniType=Half-duplex2026.07 | 39.9 | |
| Freeze-OmniType=System-level2026.07 | 38.3 | |
| Lychee-FDType=Native2026.07 | 38.3 | |
| Lychee-FD (w/o Param-Sep)Type=Native2026.07 | 34.6 | |
| Lychee-FD (w/o Sem-Channel)Type=Native2026.07 | 34.1 | |
| SALMONN-omniType=Native2026.07 | 33.7 | |
| Fun-Audio-ChatType=Native2026.07 | 26.2 | |
| MoshiType=Native2026.07 | 25.3 | |
| FLM-audioType=Native2026.07 | 15.6 |