AI Paper Atlas

Research map / Image generation / Image restoration

Generative Diffusion Prior for Unified Image Restoration and Enhancement

CVPR 2023 · Image generation › Image restoration · approach: Diffusion-based universal restoration · 275 citations (OpenAlex)

Read the paperCheck 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 changedWhereMeasuredEffect
Guide the predicted clean image x̃0 instead of the noisy state xt (ablation)GDP · ImageNet linear restoration tasksFID: 64.67 -> 38.24; 5.00 -> 2.44; 20.24 -> 16.58; 66.43 -> 37.60Helped
Add variance Σ to guided denoising (ablation)GDP-x0 · ImageNet linear restoration tasksFID: 38.24 -> 41.22; 2.44 -> 50.29; 16.58 -> 22.77; 37.60 -> 38.39Hurt
Apply guidance shift with the diffusion posterior coefficient (ablation)GDP-x0 · ImageNet linear restoration tasksFID: 38.24 -> 44.16; 2.44 -> 10.35; 16.58 -> 37.32; 37.60 -> 41.53Hurt
Use GDP-x0 guidance instead of GDP-xt guidance (ablation)GDP · LOL low-light enhancementFID: 238.92 -> 75.16Helped
Use GDP-x0 guidance instead of GDP-xt guidance (ablation)GDP · NTIRE HDR recoveryPSNR: 19.36 -> 24.88Helped

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy