AI Paper Atlas

Research map / Image generation / Image restoration

Learning Proximal Operators: Using Denoising Networks for Regularizing Inverse Imaging Problems

Tim Meinhardt, Michael Moller, Caner Hazirbas, Daniel Cremers

ICCV 2017 · Image generation › Image restoration · approach: Diffusion priors for denoising and deblurring · 378 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
use a DnCNN-S denoising network as the proximal operator instead of BM3D in demosaickingFlexISP-style PDHG · Bayer demosaicking, McMaster (18 images)PSNR (dB): 36.12 -> 37.12Improved
replace BM3D with the DnCNN-S network as the proximal operator in deconvolutionFlexISP-style PDHG · Deconvolution, 11 grayscale images, average over experiments a-ePSNR (dB): 24.997267 -> 25.11Improved
use one generic denoiser instead of MLP trained per experimentDeconvolution · Deconvolution, 11 grayscale images, average over experiments a-ePSNR (dB): 25.27 -> 25.11Hurt
use the generic denoiser instead of the specialized MLP on motion blurDeconvolution · Deconvolution experiment e (motion blur)PSNR (dB): 29.42 -> 30.17Improved
use the fixed denoiser versus a deep CNN trained for demosaickingDemosaicking · Bayer demosaicking, McMaster (18 images)PSNR (dB)Hurt
train the denoising network on a smaller noise level (sigma 0.001 instead of 0.02) (ablation)Ours · Bayer demosaicking, McMaster (18 images)PSNR (dB): 37.123543 -> 36.050674Hurt
train the denoising network on sigma 0.05 instead of 0.02 (ablation)Ours · Bayer demosaicking, McMaster (18 images)PSNR (dB): 37.123543 -> 36.082635Hurt
add explicit TV and cross-channel priors besides the denoiser (ablation)Ours · Bayer demosaicking, McMaster (18 images)PSNR (dB): about 0.4 dBImproved
train the denoiser on sigma 0.06 instead of 0.02 (ablation)Ours · Deconvolution, 11 grayscale images, average over experiments a-ePSNR (dB): 25.11 -> 25.3257Improved
train the denoiser on sigma 0.2 instead of 0.02 (ablation)Ours · Deconvolution, 11 grayscale images, average over experiments a-ePSNR (dB): 25.11 -> 24.2501Hurt
train the denoiser on sigma 0.01 instead of 0.02 (ablation)Ours · Deconvolution, 11 grayscale images, average over experiments a-ePSNR (dB): 25.11 -> 24.6253Hurt
use the neural denoiser in place of BM3D inside the optimization for runtimeDeconvolution · Deconvolution runtimeseconds: approx. 4s -> approx. 2.5s (37.5% improvement)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