Speech Separation On Libri10Mix
Metrics
SI-SDRi
Results
Performance results of various models on this benchmark
Comparison Table
| Model Name | SI-SDRi |
|---|---|
| sepit-approaching-a-single-channel-speech | 8.2 |
| separate-and-diffuse-using-a-pretrained | 9 |
| many-speakers-single-channel-speech | 7.78 |