Speech Translation on CoVoST2 en→ja
47.57BLEUQwen2.5-Omni + SICL-AT2
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen2.5-Omni + SICL-AT2few-shot=true2026.01 | 47.57 | |
| Qwen2.5-Omni + SICL-AT3few-shot=true2026.01 | 47.19 | |
| Qwen2.5-Omni (Fine-tuned on RSR)few-shot=false2026.01 | 43.72 | |
| Qwen2.5-Omni + SICL-AT1few-shot=true2026.01 | 43.16 | |
| Qwen2.5-Omni + Vanilla SICLfew-shot=true2026.01 | 42.58 | |
| Qwen2.5-Omnifew-shot=false2026.01 | 41.7 | |
| MiMo-Audio + SICL-AT2few-shot=true2026.01 | 36.92 | |
| MiMo-Audio + SICL-AT3few-shot=true2026.01 | 36.84 | |
| Qwen2.5-Omni (Fine-tuned on CV-en) + Vanilla SICLfew-shot=true2026.01 | 33.28 | |
| Qwen2.5-Omni (Fine-tuned on CV-en)few-shot=false2026.01 | 31.8 | |
| MiMo-Audio + Vanilla SICLfew-shot=true2026.01 | 26.56 | |
| MiMo-Audiofew-shot=false2026.01 | 5.25 | |
| MiMo-Audio + SICL-AT1few-shot=true2026.01 | 1.4 |