AI Paper Atlas

Research map / Image generation / Image restoration

UniRestorer: Universal Image Restoration via Adaptively Estimating Image Degradation at Proper Granularity

Jingbo Lin, Zhilu Zhang, Wenbo Li, Renjing Pei, Hang Xu, Hongzhi Zhang, Wangmeng Zuo

ICLR 2026 · Image generation › Image restoration · approach: Efficient restoration architectures · 12 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 multi-granularity MoE with degradation and granularity estimation instead of prior all-in-one methodsUniRestorer · All-in-one 7 tasks averagePSNR: 28.27 -> 31.34Improved
use UniRestorer vs NAFNet on LIVE1 compression artifact removalUniRestorer · LIVE1 CARPSNR: 31.26 -> 30.57Hurt
add granularity router G_g on top of degradation router G_d (ablation)UniRestorer · Mixed-degradation Urban100 out-distPSNR: 18.60 -> 19.45Improved
add L_dg uncertainty loss and L_load load-balance loss (ablation)UniRestorer routers · Mixed-degradation Urban100 out-distPSNR: 18.37 -> 19.45Improved
use multiple granularities {1,4,8} instead of single level {8} (ablation)UniRestorer · Mixed-degradation Urban100PSNR out-dist: 18.35 -> 19.45Improved
add a fourth granularity level {1,4,8,16} (ablation)UniRestorer · Mixed-degradation Urban100PSNR out-dist: 19.45 -> 19.47No clear effect
increase number of DR clusters from 1 to 16 (ablation)UniRestorer single-level · Mixed-degradation Urban100PSNR in-dist: 22.06 -> 24.30Improved
increase number of DR clusters from 4 to 16 (ablation)UniRestorer single-level · Mixed-degradation Urban100 out-distPSNR: 18.49 -> 18.44No clear effect
use fine-grained DR extractor instead of DA-CLIP for clustering (ablation)UniRestorer · Rain200HPSNR: 31.65 -> 32.01Improved
cluster data with learned extractor rather than manual dataset-source split (ablation)UniRestorer · Rain200LPSNR: 41.43 -> 41.61Improved
use instruction mode against single-task modelsUniRestorer · Single-task comparison, LIVE1 q=10PSNR: 27.95 -> 27.94Mixed

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