Reference-based Super-Resolution on CUFED5
28.77Clean Output PSNRSSMTF-rec
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| SSMTF-recTraining Loss=Reconstruction-only2026.01 | 28.77 | 0.8553 | 19.31 | 0.629 | 9.46 | 0.2263 | 36.19 | 0.9249 | |
| DATSR-recTraining Loss=Reconstruction-only2026.01 | 28.49 | 0.851 | 18.35 | 0.564 | 10.14 | 0.287 | 36.05 | 0.9232 | |
| SSMTFTraining Loss=Full2026.01 | 28.13 | 0.8383 | 18.88 | 0.5569 | 9.25 | 0.2814 | 35.75 | 0.9193 | |
| DATSRTraining Loss=Full2026.01 | 27.76 | 0.8285 | 17.12 | 0.469 | 10.64 | 0.3595 | 36.15 | 0.9236 | |
| MASA-recTraining Loss=Reconstruction-only2026.01 | 27.35 | 0.814 | 24.55 | 0.7549 | 2.8 | 0.0591 | 37.03 | 0.9349 | |
| TTSR-recTraining Loss=Reconstruction-only2026.01 | 26.99 | 0.8003 | 23.55 | 0.7145 | 3.44 | 0.0858 | 36.28 | 0.9236 | |
| TTSRTraining Loss=Full2026.01 | 25.4 | 0.76 | 21.84 | 0.5599 | 3.56 | 0.2001 | 35.71 | 0.9138 | |
| MASATraining Loss=Full2026.01 | 24.65 | 0.7257 | 19.69 | 0.5811 | 4.96 | 0.1446 | 36.99 | 0.9316 |