Speech Separation On Wsj0 5Mix
Metrics
SI-SDRi
Results
Performance results of various models on this benchmark
Model Name | SI-SDRi | Paper Title | Repository |
|---|---|---|---|
| TasTas | 11.14 | Toward Speech Separation in The Pre-Cocktail Party Problem with TasTas | - |
| Gated DualPathRNN | 10.56 | Voice Separation with an Unknown Number of Multiple Speakers | - |
| SepTDA | 21.0 | Boosting Unknown-number Speaker Separation with Transformer Decoder-based Attractor | - |
| Hungarian PIT | 13.22 | Many-Speakers Single Channel Speech Separation with Optimal Permutation Training | - |
| Conditional TasNet | 11.7 | Sequence to Multi-Sequence Learning via Conditional Chain Mapping for Mixture Signals | - |
| Multi-Decoder DPRNN | 5.9 | Multi-Decoder DPRNN: High Accuracy Source Counting and Separation | - |
0 of 6 row(s) selected.