AI Paper Atlas

Research map / Image generation / Image restoration

Relationship Quantification of Image Degradations

Boyun Li, Yuanbiao Gou, Wenxin Wang, Peng Hu, Wangmeng Zuo, Xi Peng

TPAMI 2025 · Image generation › Image restoration · approach: Efficient restoration architectures · 6 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
add 10% denoising data to dehazing training (ablation)MSBDN · RESIDE+ outdoor dehazingPSNR: 33.84 -> 33.96Improved
raise noise proportion to 90% (ablation)MSBDN · RESIDE+ outdoor dehazingPSNR: 33.84 -> 30.96Hurt
raise noise proportion to 30% (ablation)MSBDN · RESIDE+ outdoor dehazingPSNR: 33.84 -> 33.00Hurt
raise noise proportion from 10% to 20% (ablation)MSBDN · RESIDE+ outdoor dehazingPSNR: 33.96 -> 33.93No clear effect
add 10% rain data to dehazing trainingMSBDN · SOTS+ indoorPSNR: 32.72 -> 35.14Improved
add 10% rain data to dehazing trainingDehamer · SOTS+ outdoorPSNR: 35.18 -> 36.11Improved
add 10% snow data to dehazing trainingAECR-Net · SOTS+ indoorPSNR: 37.17 -> 37.29No clear effect
add auxiliary degradations (noise, rain, snow) at 10% averaged over seven baselinesseven dehazing baselines · SOTS+PSNR: +0.42dB, +0.77dB, +0.63dBImproved
add 10% haze auxiliary to rain restorationMSBDN · SOTS+ rain anchorPSNR: 38.12 -> 36.66Hurt
add 10% rain auxiliary to noise restorationMSBDN · SOTS+ noise anchorPSNR: 34.85 -> 34.83No clear effect
use DRI instead of Taskonomy affinity to predict benefit of 10% noisehaze anchor MSBDN · RESIDE+ hazesign of predictor vs PSNR gain: -1.00000 -> 0.00860Improved
compute DRI every 5 steps instead of every step (ablation)DRI · Pearson with final performancePearson coefficient: 0.9698 -> 0.9721Improved

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