Automatic Speech Recognition on Artificial Dataset Additive Noise
4.39CERROSE
Evaluation results
| Method | Links | |
|---|---|---|
| ROSEASR Model=W2V22023.12 | 4.39 | |
| ROSE w/o attn.ASR Model=W2V22023.12 | 5.57 | |
| ROSE w/o LASRASR Model=W2V22023.12 | 6.29 | |
| SE-ConformerASR Model=W2V22023.12 | 6.54 | |
| BSRNNASR Model=W2V22023.12 | 6.73 | |
| DCCRN-EASR Model=W2V22023.12 | 6.88 | |
| MixedASR Model=W2V22023.12 | 7.03 | |
| ROSEASR Model=DS22023.12 | 12.75 | |
| ROSE w/o attn.ASR Model=DS22023.12 | 17.86 | |
| ROSE w/o LASRASR Model=DS22023.12 | 19.12 | |
| SE-ConformerASR Model=DS22023.12 | 21.43 | |
| BSRNNASR Model=DS22023.12 | 23.79 | |
| DCCRN-EASR Model=DS22023.12 | 26.69 | |
| MixedASR Model=DS22023.12 | 33.36 |