Research map / Learning theory and algorithms / Prediction and decision losses
Noise-Blind Image Deblurring
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 noise-adaptive weighting to EPLL (ablation) | EPLL · Levin et al. 32 test images; known noise levels | PSNR (dB): 0.3–0.7dB on average | Helped |
| Add noise-adaptive weighting to EPLL (ablation) | EPLL · Sun et al. 640 test images | PSNR (dB): 32.05 -> 32.18; 29.60 -> 30.08; 28.25 -> 28.77; 27.34 -> 27.81 | Mixed |
| Add noise-adaptive weighting to EPLL (ablation) | EPLL · Berkeley 50 test images, large blurs | PSNR (dB): 25.38 -> 25.57; 23.53 -> 23.90; 22.54 -> 22.91; 21.91 -> 22.27 | Mixed |
| Use GradNet 7S instead of EPLL + NA | Noise-blind deblurring · Berkeley 50 test images, large blurs | PSNR (dB): 25.57 vs 25.57; 24.23 vs 23.90; 23.46 vs 22.91; 22.94 vs 22.27 | Mixed |
| Add exact MM preconditioning (ablation) | Gradient-descent deblurring with TV prior · Levin et al. dataset, 32 images | convergence speed: 6 and 6.5 times | Helped |
| Replace full GradNet with 1-stage or 4-stage GradNet (ablation) | GradNet · Figure 5, 1% noise example | PSNR | Mixed |
Closest papers by meaning
- Variational-EM-Based Deep Learning for Noise-Blind Image Deblurring (CVPR 2020 · similarity 0.89)
- Deblurring via Stochastic Refinement (CVPR 2022 · similarity 0.89)
- BlurDM: A Blur Diffusion Model for Image Deblurring (NeurIPS 2025 · similarity 0.87)
- Robust Blind Text Image Deblurring via Maximum Consensus Framework (AAAI 2024 · similarity 0.87)
- Unsupervised Blind Image Deblurring Based on Self-Enhancement (CVPR 2024 · similarity 0.87)
- Residual Diffusion Deblurring Model for Single Image Defocus Deblurring (AAAI 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.