Research map / Image generation / Image restoration
Blind Image Deconvolution by Automatic Gradient Activation
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 |
|---|---|---|---|
| Select a subset of latent-image gradients automatically for kernel estimation instead of relying on a global sparse-gradient prior | Blind image deconvolution · Köhler et al. dataset | PSNR | Helped |
| Select a subset of latent-image gradients automatically for kernel estimation instead of relying on a global sparse-gradient prior | Blind image deconvolution · Köhler et al. dataset | SSIM | Helped |
| Set λ to 0.005 rather than 0.001 or 0.01 (ablation) | Gradient activation method · Levin synthetic dataset, noise-free | SSD error ratio | Helped |
| Use early stopping to avoid activating too many gradients (ablation) | Gradient activation method · Non-sparse-gradient toy image | SSD error | Helped |
| Discard gradients corresponding to high-intensity pixels (ablation) | Gradient activation method · Images with saturation and light streaks | blur-kernel estimation quality | Helped |
| Discard isolated activated gradients with fewer than 4 elements (ablation) | Gradient activation method · Images with impulse noise | blur-kernel estimation quality | Helped |
Closest papers by meaning
- Deep Learning for Handling Kernel/model Uncertainty in Image Deconvolution (CVPR 2020 · similarity 0.89)
- Neural Blind Deconvolution Using Deep Priors (CVPR 2020 · similarity 0.89)
- Robust Kernel Estimation With Outliers Handling for Image Deblurring (CVPR 2016 · similarity 0.88)
- Learning Fully Convolutional Networks for Iterative Non-Blind Deconvolution (CVPR 2017 · similarity 0.88)
- Blind Image Restoration via Fast Diffusion Inversion (NeurIPS 2024 · similarity 0.87)
- A Restoration Network as an Implicit Prior (ICLR 2024 · similarity 0.86)
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.