AI Paper Atlas

Research map / Image generation / Image restoration

On-Demand Learning for Deep Image Restoration

Ruohan Gao, Kristen Grauman

ICCV 2017 · Image generation › Image restoration · approach: Priors for restoration networks · 81 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
replace rigid joint learning with on-demand allocation of sub-tasks (ablation)image restoration network · CelebA Image Deblurringmean PSNR: 29.40 dB -> 29.58 dBImproved
replace rigid joint learning with on-demand allocation of sub-tasks (ablation)image restoration network · CelebA Image Inpaintingmean PSNR: 32.11 dB -> 32.30 dBImproved
replace rigid joint learning with on-demand allocation of sub-tasks (ablation)image restoration network · SUN397 Image Deblurringmean PSNR: 28.53 dB -> 28.70 dBImproved
replace rigid joint learning with on-demand allocation of sub-tasks (ablation)image restoration network · SUN397 Image Denoisingmean PSNR: 25.69 dB -> 25.80 dBImproved
replace hard mining with on-demand learning (ablation)image restoration network · CelebA Image Deblurringmean PSNR: 27.33 dB -> 29.58 dBImproved
use staged curriculum instead of cumulative joint training (ablation)image restoration network · CelebA Image Deblurringmean PSNR: 29.31 dB -> 15.59 dBHurt
reverse curriculum order (staged anti-curriculum) versus joint learning (ablation)image restoration network · CelebA Image Deblurringmean PSNR: 29.40 dB -> 25.43 dBHurt
use cumulative curriculum instead of rigid joint learning (ablation)image restoration network · CelebA Image Deblurringmean PSNR: 29.40 dB -> 28.70 dBHurt
add 100,000 extra training images to rigid joint learning (200,000 total) (ablation)rigid joint learning · CelebAPSNR/L2No clear effect
replace center-only Pathak inpainter with on-demand inpainter on arbitrary blocksinpainting · CelebA arbitrary square blockPSNR: 11.80 dB -> 20.31 dBImproved
on-demand all-rounder versus Pathak center inpainter on central blocksinpainting · CelebA central square blockPSNR: 22.16 dB -> 20.74 dBHurt
train on a single corruption level (fixated) instead of all levels (ablation)fixated model · CelebA, all difficulty levelsqualitative restoration qualityHurt

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