Automatic Speech Recognition on ExtremeNoise
16.17WERQwen3-ASR
Evaluation results
| Method | Links | |
|---|---|---|
| Qwen3-ASRParameters=1.7B2026.01 | 16.17 | |
| Doubao-ASR2026.01 | 17.04 | |
| Qwen3-ASRParameters=0.6B2026.01 | 17.88 | |
| Gemini-2.5Version=Pro2026.01 | 29.06 | |
| GPT-4oMode=Transcribe2026.01 | 36.11 | |
| Fun-ASRModel=MLT-Nano2026.01 | 36.55 | |
| WhisperVersion=large-v32026.01 | 63.17 |