Evaluations and comparisons of models using different loss function in terms of PESQ. The best socre is marked in bold.
| Metrix | PESQ | |||||||
|---|---|---|---|---|---|---|---|---|
| Test Noise | Babble | Cafeteria | ||||||
| Test SNR (dB) | -5 | 0 | 5 | Avg. | -5 | 0 | 5 | Avg. |
| LMse-LMag | 2.45 | 2.98 | 3.33 | 2.92 | 2.29 | 2.84 | 3.24 | 2.79 |
| LMag-LMag | 2.32 | 2.85 | 3.21 | 2.79 | 2.15 | 2.72 | 3.15 | 2.67 |
| LCom-Lcom | 2.24 | 2.85 | 3.25 | 2.78 | 2.04 | 2.68 | 3.19 | 2.64 |
| LMse-LMag-NC | 2.63 | 3.12 | 3.38 | 3.05 | 2.48 | 3.01 | 3.35 | 2.95 |
| LMag-LMag-NC | 2.64 | 3.13 | 3.38 | 3.05 | 2.47 | 3.00 | 3.35 | 2.94 |
| Lcom-Lcom-NC | 2.65 | 3.15 | 3.45 | 3.08 | 2.45 | 3.02 | 3.41 | 2.96 |
| Metrix | PESQ | |||||||
|---|---|---|---|---|---|---|---|---|
| Test Noise | Babble | Cafeteria | ||||||
| Test SNR (dB) | -5 | 0 | 5 | Avg. | -5 | 0 | 5 | Avg. |
| 2.32 | 2.85 | 3.21 | 2.79 | 2.15 | 2.72 | 3.15 | 2.67 | |
| 2.24 | 2.85 | 3.25 | 2.78 | 2.04 | 2.68 | 3.19 | 2.64 | |
| 2.63 | 3.12 | 3.38 | 3.05 | 3.01 | 3.35 | 2.95 | ||
| 2.64 | 3.13 | 3.38 | 3.05 | 2.47 | 3.00 | 3.35 | 2.94 | |
| 2.45 | ||||||||
Sharing content requires targeting cookies to be enabled. Please update your cookie preferences to use this feature.