Blind Deconvolution on Levin dataset realistic camera shake blur (alpha=40)
22.96PSNRPhD-Net
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| PhD-NetBlind deconvolution=false, End-To-End Trained=true2023.03 | 22.96 | 0.687 | 0.263 | 0.432 | |
| Kernel Trajectory NetworkBlind deconvolution=true, End-To-End Trained=false2023.03 | 22.32 | 0.647 | 0.273 | 0.444 | |
| MIMO-UNet+Blind deconvolution=true, End-To-End Trained=true2023.03 | 21.88 | 0.583 | 0.428 | 0.557 | |
| PURE-LETBlind deconvolution=false, End-To-End Trained=false2023.03 | 21.78 | 0.614 | 0.324 | 0.507 | |
| MPRNetBlind deconvolution=true, End-To-End Trained=true2023.03 | 21.54 | 0.567 | 0.447 | 0.479 | |
| Deep-DeblurBlind deconvolution=true, End-To-End Trained=true2023.03 | 21.2 | 0.545 | 0.494 | 0.493 | |
| SRNBlind deconvolution=true, End-To-End Trained=true2023.03 | 20.59 | 0.535 | 0.491 | 0.506 | |
| DHMPNBlind deconvolution=true, End-To-End Trained=true2023.03 | 20.5 | 0.532 | 0.494 | 0.506 | |
| P4IPBlind deconvolution=false, End-To-End Trained=false2023.03 | 17.32 | 0.362 | 0.487 | 0.56 |