Reference-based Super-Resolution on WR-SR
27.89PSNR (Clean)SSMTF-rec
Evaluation results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| SSMTF-recTraining Loss=Reconstruction-only2026.01 | 27.89 | 0.7929 | 19.89 | 0.668 | 8 | 0.1249 | 36.71 | 0.9224 | |
| DATSR-recTraining Loss=Reconstruction-only2026.01 | 27.83 | 0.7916 | 18.41 | 0.6122 | 9.42 | 0.1794 | 36.52 | 0.92 | |
| MASA-recTraining Loss=Reconstruction-only2026.01 | 27.72 | 0.7836 | 25.68 | 0.7594 | 2.04 | 0.0242 | 37.47 | 0.9342 | |
| TTSR-recTraining Loss=Reconstruction-only2026.01 | 27.53 | 0.7803 | 23.55 | 0.668 | 3.98 | 0.1123 | 36.29 | 0.9163 | |
| SSMTFTraining Loss=Full2026.01 | 27.51 | 0.7767 | 19.8 | 0.6292 | 7.71 | 0.1475 | 36.25 | 0.9164 | |
| DATSRTraining Loss=Full2026.01 | 27.39 | 0.7732 | 16.76 | 0.5353 | 10.63 | 0.2379 | 36.63 | 0.9214 | |
| TTSRTraining Loss=Full2026.01 | 26.38 | 0.748 | 21.6 | 0.4809 | 4.78 | 0.2671 | 35.8 | 0.9078 | |
| MASATraining Loss=Full2026.01 | 25.33 | 0.7027 | 20.51 | 0.5795 | 4.82 | 0.1232 | 37.3 | 0.9294 |