AI Paper Atlas

Research map / Image generation / Image restoration

DiffIR: Efficient Diffusion Model for Image Restoration

Bin Xia, Yulun Zhang, Shiyin Wang, Yitong Wang, Xinglong Wu, Yapeng Tian, Wenming Yang, Luc Van Gool

ICCV 2023 · Image generation › Image restoration · approach: Diffusion-based universal restoration · 479 citations (OpenAlex)

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 changedWhereMeasuredEffect
use diffusion model to estimate the compact IR prior (vs no DM) (ablation)DiffIR_S2 · CelebA-HQ inpaintingFID: 5.6782 -> 5.1440Improved
joint optimization of denoising network, CPEN_S2 and DIRformer instead of traditional DM optimization (ablation)DiffIR_S2 · CelebA-HQ inpaintingFID: 5.9766 -> 5.1440Improved
traditional DM optimization with DM compared with no DM (ablation)DiffIR_S2 · CelebA-HQ inpaintingFID: 5.6782 -> 5.9766Hurt
insert variance noise in reverse diffusion process (ablation)DiffIR_S2 · CelebA-HQ inpaintingFID: 5.1440 -> 5.1937Hurt
use L1-style L_diff instead of L2 for IPR estimation (ablation)DiffIR_S2 · CelebA-HQ inpaintingFID: 5.1837 -> 5.1440Improved
use L_diff instead of KL loss for IPR estimation (ablation)DiffIR_S2 · CelebA-HQ inpaintingFID: 5.2365 -> 5.1440Improved
increase diffusion iterations up to 3-4 (ablation)DiffIR_S2 · CelebA-HQ inpaintingperformanceImproved
increase diffusion iterations beyond 4 (ablation)DiffIR_S2 · CelebA-HQ inpaintingperformanceNo clear effect
estimate a compact IPR with DM instead of whole images, versus RePaintinpainting · CelebA-HQ wide masksFID: 5.4881 -> 5.1440 with 4.3% parameters and 0.1% computeImproved
use DiffIR_S2 instead of Restormer for deblurringmotion deblurring · GoProPSNR: 32.92 -> 33.20Improved
use ground-truth-guided stage-1 prior (DiffIR_S1) vs estimated prior (S2) (ablation)DiffIR · CelebA-HQ inpaintingFID: 4.8045 -> 5.1440Mixed

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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