Research map / Image generation / Image restoration
Learning Single Image Defocus Deblurring with Misaligned Training Pairs
Yu Li, Dongwei Ren, Xinya Shu, Wangmeng Zuo
Read the paperCheck 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 |
|---|---|---|---|
| Apply JDRL to MPRNet | MPRNet · SDD test set | PSNR: 26.28 -> 26.88 | Improved |
| Apply JDRL to UNet | UNet · SDD test set | PSNR: 24.62 -> 25.82 | Improved |
| Apply JDRL to MPRNet | MPRNet · DPDD test set (Deformed GT) | PSNR: 26.03 -> 26.21 | Improved |
| Apply JDRL to IFAN | IFAN · DPDD test set (Deformed GT) | PSNR: 25.86 -> 26.09 | Improved |
| Apply JDRL to IFAN | IFAN · RealDOF test set (Deformed GT) | PSNR: 25.39 -> 25.61 | Improved |
| Apply JDRL to MPRNet | MPRNet · RealDOF test set (Deformed GT) | PSNR: 24.66 -> 24.90 | Improved |
| Add cycle deformation (ablation) | JDRL with UNet · SDD test set | PSNR: 24.80 -> 25.82 | Improved |
| Add calibration mask (ablation) | JDRL with UNet · SDD test set | PSNR: 25.78 -> 25.82 | No clear effect |
| Add reblurring module (ablation) | JDRL with UNet · SDD test set | PSNR: 25.58 -> 25.82 | Improved |
| Add isotropic kernel design (ablation) | JDRL with UNet · SDD test set | PSNR: 25.68 -> 25.82 | Improved |
| Add weight prediction network (ablation) | JDRL with UNet · SDD test set | PSNR: 25.74 -> 25.82 | Improved |
| Replace JDRL with contextual bilateral loss | UNet · SDD test set | PSNR: 24.58 vs. 25.82 | Hurt |
Closest papers by meaning
- Reblurring-Guided Single Image Defocus Deblurring: A Learning Framework with Misaligned Training Pairs (IJCV 2025 · similarity 0.96)
- Residual Diffusion Deblurring Model for Single Image Defocus Deblurring (AAAI 2025 · similarity 0.92)
- Learnable Blur Kernel for Single-Image Defocus Deblurring in the Wild (AAAI 2023 · similarity 0.92)
- Prior and Prediction Inverse Kernel Transformer for Single Image Defocus Deblurring (AAAI 2024 · similarity 0.90)
- Quad-Pixel Image Defocus Deblurring: A New Benchmark and Model (CVPR 2025 · similarity 0.89)
- LDP: Language-driven Dual-Pixel Image Defocus Deblurring Network (CVPR 2024 · 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.
