Gaussian Image Denoising on CBSD68 (test)
28.72PSNR (sigma=50)DMID-d
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| DMID-dTraining strategy=Noise-blind (various noise levels)2023.07 | 28.72 | 34.45 | 31.86 | |
| RestormerTraining strategy=Noise-specific (separate models)2023.07 | 28.6 | 34.4 | 31.79 | |
| RestormerTraining strategy=Noise-blind (various noise levels)2023.07 | 28.59 | 34.39 | 31.78 | |
| SwinIRTraining strategy=Noise-specific (separate models)2023.07 | 28.56 | 34.42 | 31.78 | |
| DRUNetTraining strategy=Noise-blind (various noise levels)2023.07 | 28.51 | 34.3 | 31.69 | |
| CODETraining strategy=Noise-specific (separate models)2023.07 | 28.47 | 34.33 | 31.69 | |
| IPTTraining strategy=Noise-specific (separate models)2023.07 | 28.39 | — | — | |
| RDNTraining strategy=Noise-specific (separate models)2023.07 | 28.31 | — | — | |
| RNANTraining strategy=Noise-specific (separate models)2023.07 | 28.27 | — | — | |
| BRDNetTraining strategy=Noise-specific (separate models)2023.07 | 28.16 | 34.1 | 31.43 | |
| DSNetTraining strategy=Noise-blind (various noise levels)2023.07 | 28.05 | 33.91 | 31.28 | |
| FFDNetTraining strategy=Noise-blind (various noise levels)2023.07 | 27.96 | 33.87 | 31.21 | |
| DnCNNTraining strategy=Noise-blind (various noise levels)2023.07 | 27.95 | 33.9 | 31.24 | |
| CBM3DTraining strategy=Noise-blind (various noise levels)2023.07 | 27.38 | 33.52 | 30.71 |