Research map / Image generation / Image restoration
XYDeblur: Divide and Conquer for Single Image Deblurring
Seo-Won Ji, Jeongmin Lee, Seung-Wook Kim, Jun-Pyo Hong, Seung-Jin Baek, Seung-Won Jung, Sung-Jea Ko
Check your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Replace PSS-NSC's U-Net base structure with XYDeblur | PSS-NSC · GoPro test | PSNR: 30.94(30.92) -> 31.27 | Improved |
| Replace DMPHN's U-Net base structure with XYDeblur | DMPHN · RealBlur test | PSNR: 26.97 -> 28.02 | Improved |
| Replace MSCNN's U-Net base structure with XYDeblur | MSCNN · GoPro test | PSNR: 29.22(29.08) -> 29.98 | Improved |
| Add a second independent decoder (ablation) | U-Net · GoPro test | PSNR: 30.61 -> 30.90 | Improved |
| Increase from two to three independent decoders (ablation) | U-Net2D · GoPro test | PSNR: 30.90 -> 30.99 | Improved |
| Increase from two to four independent decoders (ablation) | U-Net2D · GoPro test | PSNR: 30.90 -> 31.04 | Improved |
| Share decoder parameters after rotating one decoder's kernels 90 degrees (ablation) | U-Net2D · GoPro test | PSNR: 30.90 -> 30.97 | Improved |
| Share decoder parameters after rotating one decoder's kernels 90 degrees (ablation) | U-Net2D · GoPro test | SSIM: 0.9489 -> 0.9502 | Improved |
| Increase decoder channel count (ablation) | PSS-NSC · RealBlur test | PSNR: 28.78 -> 28.63 | Hurt |
| Increase decoder channel count (ablation) | DMPHN · RealBlur test | PSNR: 26.97 -> 26.86 | Hurt |
| Increase decoder channel count (ablation) | MSCNN · GoPro test | PSNR: 29.22(29.08) -> 29.34 | Improved |
Later papers that compared with this method
1 later accepted paper reports XYDeblur as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Learning to Deblur Polarized Images (IJCV 2025 · better on 3 of 3 shared results)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Deep Image Deblurring: A Survey (IJCV 2022 · similarity 0.90)
- Deblurring Using Analysis-Synthesis Networks Pair (CVPR 2020 · similarity 0.89)
- Efficient Dynamic Scene Deblurring Using Spatially Variant Deconvolution Network With Optical Flow Guided Training (CVPR 2020 · similarity 0.88)
- Rethinking Coarse-To-Fine Approach in Single Image Deblurring (ICCV 2021 · similarity 0.88)
- Blur2Blur: Blur Conversion for Unsupervised Image Deblurring on Unknown Domains (CVPR 2024 · similarity 0.87)
- DeblurDiff: Real-Word Image Deblurring with Generative Diffusion Models (NeurIPS 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
