Research map / Image generation / Image restoration
Learning Discriminative Data Fitting Functions for Blind Image Deblurring
Jinshan Pan, Jiangxin Dong, Yu-Wing Tai, Zhixun Su, Ming-Hsuan Yang
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 |
|---|---|---|---|
| add learned data fitting weights instead of intensity only (ablation) | blind deblurring kernel estimation · synthetic blurred images | average PSNR (dB): 23.08 -> 31.88 | Improved |
| add learned data fitting weights instead of gradient only (ablation) | blind deblurring kernel estimation · synthetic blurred images | average PSNR (dB): 26.78 -> 31.88 | Improved |
| use learned weights instead of intensity for latent image and gradient for kernel (ablation) | blind deblurring kernel estimation · synthetic blurred images | average PSNR (dB): 30.52 -> 31.88 | Improved |
| learn intensity weight for kernel estimation (ablation) | learned data fitting function · kernel estimation | learned weight of zero-order term: 0 -> 0 | No clear effect |
| use learned weights separately for non-blind deconvolution | latent image restoration · non-blind deconvolution | learned weight of intensity term: 0.2095 | Improved |
| learn weights instead of fixed intensity or gradient terms | blind deblurring · Sun et al. benchmark (80 images) | success rate | Improved |
| use learned data fitting function in place of Xu et al. combination | blind deblurring run time · 255x255 image | run time (s): 3.10 -> 4.93 | Hurt |
Closest papers by meaning
- Learning Data Terms for Non-blind Deblurring (ECCV 2018 · similarity 0.91)
- Non-Blind Deblurring: Handling Kernel Uncertainty With CNNs (CVPR 2018 · similarity 0.90)
- Self-Supervised Non-Uniform Kernel Estimation With Flow-Based Motion Prior for Blind Image Deblurring (CVPR 2023 · similarity 0.89)
- Blind Deblurring for Saturated Images (CVPR 2021 · similarity 0.89)
- Blind image deblurring with noise-robust kernel estimation (ECCV 2024 · similarity 0.89)
- 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.
