Research map / Image generation / Image restoration
Deblurring via Stochastic Refinement
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 |
|---|---|---|---|
| use predict-and-refine stochastic diffusion | blind image deblurring · GoPro | LPIPS: 0.117 -> 0.059 | Helped |
| use predict-and-refine stochastic diffusion | blind image deblurring · HIDE (GoPro-trained) | LPIPS: 0.114 -> 0.089 | Helped |
| average multiple generated samples (Ours-SA) | predict-and-refine diffusion · GoPro | PSNR: 31.66 -> 33.23 | Helped |
| average multiple generated samples (Ours-SA) | predict-and-refine diffusion · GoPro | FID: 4.04 -> 17.46 | Hurt |
| increase U-Net channels from 16 to 32 (ablation) | initial-predictor regression baseline · GoPro | LPIPS: 0.137 -> 0.113 | Helped |
| increase U-Net channels from 16 to 32 (ablation) | initial-predictor regression baseline · GoPro | PSNR: 29.93 -> 31.05 | Helped |
| increase U-Net channels from 32 to 64 (ablation) | initial-predictor regression baseline · GoPro | LPIPS: 0.113 -> 0.103 | Helped |
| increase U-Net channels from 32 to 64 (ablation) | initial-predictor regression baseline · GoPro | PSNR: 31.05 -> 31.63 | Helped |
| increase batch size from 32 to 64 (ablation) | initial-predictor regression baseline · GoPro | LPIPS: 0.103 -> 0.099 | Helped |
| increase batch size from 32 to 64 (ablation) | initial-predictor regression baseline · GoPro | PSNR: 31.63 -> 31.85 | Helped |
| increase batch size from 64 to 128 (ablation) | initial-predictor regression baseline · GoPro | LPIPS: 0.099 -> 0.087 | Helped |
| increase batch size from 64 to 128 (ablation) | initial-predictor regression baseline · GoPro | PSNR: 31.85 -> 32.56 | Helped |
Closest papers by meaning
- Residual Diffusion Deblurring Model for Single Image Defocus Deblurring (AAAI 2025 · similarity 0.91)
- BlurDM: A Blur Diffusion Model for Image Deblurring (NeurIPS 2025 · similarity 0.90)
- DeblurDiff: Real-Word Image Deblurring with Generative Diffusion Models (NeurIPS 2025 · similarity 0.90)
- Resfusion: Denoising Diffusion Probabilistic Models for Image Restoration Based on Prior Residual Noise (NeurIPS 2024 · similarity 0.90)
- DEFT: Efficient Fine-tuning of Diffusion Models by Learning the Generalised $h$-transform (NeurIPS 2024 · similarity 0.90)
- A Variational Perspective on Solving Inverse Problems with Diffusion Models (ICLR 2024 · similarity 0.89)
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.