Speech Enhancement On Whamr
Metrics
PESQ
SI-SDR
Results
Performance results of various models on this benchmark
Model Name | PESQ | SI-SDR | Paper Title | Repository |
|---|---|---|---|---|
| Non-Real-Time MultiScale+ | 1.52 | 5.33 | Phase-aware Single-stage Speech Denoising and Dereverberation with U-Net | |
| Unprocessed | 1.83 | -2.73 | - | - |
| SepFormer | 2.84 | - | Exploring Self-Attention Mechanisms for Speech Separation | |
| DTLN | 2.23 | 2.12 | Dual-Signal Transformation LSTM Network for Real-Time Noise Suppression |
0 of 4 row(s) selected.