AI Paper Atlas

Research map / Image generation / Image restoration

Unpaired Learning of Deep Image Denoising

Xiaohe Wu, Ming Liu, Yue Cao, Dongwei Ren, Wangmeng Zuo

ECCV 2020 · Image generation › Image restoration · approach: Self-supervised image denoising · 188 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 knowledge distillation to MWCNN with unpaired clean images on top of D-BSN (ablation)D-BSN · BSD68 gray AWGN sigma=25PSNR (dB): 29.12 -> 29.38Improved
add knowledge distillation to MWCNN with unpaired clean images on top of D-BSN (ablation)D-BSN · BSD68 gray heteroscedastic GaussianPSNR (dB): 29.16 -> 30.10Improved
add knowledge distillation to MWCNN with unpaired clean images on top of D-BSN (ablation)D-BSN · CBSD68 color AWGN sigma=25PSNR (dB): 30.61 -> 31.40Improved
add knowledge distillation to MWCNN with unpaired clean images on top of D-BSN (ablation)D-BSN · CBSD68 color heteroscedastic GaussianPSNR (dB): 30.56 -> 31.72Improved
add knowledge distillation to MWCNN with unpaired clean images on top of D-BSN (ablation)D-BSN · CBSD68 color multivariate GaussianPSNR (dB): 26.48 -> 26.81Improved
train MWCNN with unpaired data instead of noisy-noisy pairs (N2N)MWCNN (N2N) · CBSD68 color heteroscedastic GaussianPSNR (dB): 31.13 -> 31.72Improved
train MWCNN with unpaired data instead of paired supervision (N2C)MWCNN (N2C) · BSD68 gray AWGN sigma=25PSNR (dB): 29.41 -> 29.38No clear effect
train MWCNN with unpaired data instead of paired supervision (N2C)MWCNN (N2C) · CBSD68 color heteroscedastic GaussianPSNR (dB): 32.10 -> 31.72Hurt
use pixel-shuffle downsampling with factor 4 for real noisy photographs (ablation)D-BSN · real-world noisy photographsPSNRHurt
use MWCNN (unpaired) instead of GCBDGCBD · DND sRGB real noisePSNR (dB): 35.58 -> 37.93Improved
use MWCNN (unpaired) instead of N2S/N2VN2V · CC15 real noisePSNR (dB): 35.27 -> 35.90Improved
use MWCNN (unpaired) instead of CBDNetCBDNet · CBSD68 color heteroscedastic GaussianPSNR (dB): 30.89 -> 31.72Improved

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