AI Paper Atlas

Research map / Image generation / Image restoration

Self-Supervised Image Prior Learning With GMM From a Single Noisy Image

Haosen Liu, Xuan Liu, Jiangbo Lu, Shan Tan

ICCV 2021 · Image generation › Image restoration · approach: Self-supervised image denoising · 10 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
learn the GMM prior from a single noisy image with SS-GMM instead of supervised EPLL training on clean imagesGMM image prior · Set12PSNR: over 0.3dB on averageImproved
SS-GMM versus EPLL on BSD68, whose EPLL model was trained on same-dataset imagesGMM image prior · BSD68PSNR: 31.22 -> 31.26 at sigma=15No clear effect
use SS-GMM instead of BM3D on BSD68denoising · BSD68PSNR: about 0.1dB-0.2dBImproved
use SS-GMM instead of BM3D on Set12denoising · Set12PSNR: 32.12 -> 32.18 at sigma=15; 29.73 -> 29.68 at sigma=25Mixed
use SS-GMM instead of NL-Bayesdenoising · Set12 and BSD68PSNR: 31.98 -> 32.18 on Set12 sigma=15Improved
use SS-GMM instead of DIPdenoising · BSD68PSNR: 28.00 -> 28.73 at sigma=25Improved
use SS-GMM instead of Self2Self at low noisedenoising · BSD68PSNR: 30.26 -> 31.26 at sigma=15Improved
use SS-GMM instead of Self2Self at high noisedenoising · BSD68PSNR: 25.92 -> 25.70 at sigma=50Hurt
replace the noise estimator of Chen et al. with the peak-region estimatornoise level estimation · Set12Bias, Std, sqrt(MSE)Improved

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