AI Paper Atlas

Research map / Image generation / Image restoration

Rethinking Deep Image Prior for Denoising

Yeonsik Jo, Se Young Chun, Jonghyun Choi

ICCV 2021 · Image generation › Image restoration · approach: Self-supervised image denoising · 60 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 stochastic temporal ensembling (noise regularization with random level plus EMA) (ablation)SURE-based DIP · CSet9 (sigma=25)PSNR: 31.34 -> 31.54Improved
add stochastic temporal ensembling (ablation)SURE-based DIP · CSet9 (sigma=25)SSIM: 0.955 -> 0.953Hurt
add stochastic temporal ensembling (ablation)SURE-based DIP · CSet9 (sigma=25)LPIPS: 0.108 -> 0.107No clear effect
use hyperparameter-free MC divergence instead of epsilon-dependent approximation (DIP-SURE*)DIP with SURE · CSet9 (sigma=25)PSNR: 30.33 -> 31.34Improved
replace DIP early stopping with SURE-based objective and zero-crossing stopDIP · CSet9 (sigma=25)PSNR: 29.96 -> 31.54Improved
replace DIP early stopping with SURE-based objective and zero-crossing stopDIP · CSet9 (sigma=25)LPIPS: 0.152 -> 0.107Improved
use proposed method instead of Self2Self at low noiseSelf2Self · CSet9 (sigma=15)LPIPS: 0.135 -> 0.070Improved
use proposed method instead of Self2Self at high noiseSelf2Self · CSet9 (sigma=50)PSNR: 29.25 -> 28.90Hurt
increase number of Self2Self ensemble membersSelf2Self · CSet9PSNR vs LPIPSMixed
use proposed Poisson extension instead of DIP at high noiseDIP · Poisson (zeta=0.2)PSNR: 21.43 -> 22.85Improved
use proposed Poisson extension instead of Self2Self at low noiseSelf2Self · Poisson (zeta=0.01)PSNR: 32.18 -> 32.00Hurt
use proposed method instead of Self2Self for runtimeSelf2Self · CSet9inference time: 35 hours -> 4 hoursImproved

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