AI Paper Atlas

Research map / Image generation / Image restoration

Scaling Laws For Deep Learning Based Image Reconstruction

Tobit Klug, Reinhard Heckel

ICLR 2023 · Image generation › Image restoration · approach: Diffusion priors for denoising and deblurring · 22 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
Scale SwinIR to 128.9M parameters and train on 100k imagesSwinIR · CBSD68, Kodak24, McMaster, Urban100PSNR: 0.05 to 0.22dBImproved
Increase denoising training data beyond a few thousand examplesU-Net and SwinIR · ImageNet denoising testPSNR scaling coefficient: 0.0048 to 0.0019; 0.0050 to 0.0017Mixed
Increase MRI training data beyond about 2.5k examplesU-Net · fastMRI multi-coil brain (4x acceleration)SSIM scaling coefficient: 0.0072 to 0.0026Mixed
Increase super-resolution training data beyond roughly 30k examplesU-Net · ImageNet super-resolution (factor 2)PSNR scaling coefficient: 0.0075 to 0.0029Mixed
Use a SwinIR architecture instead of a U-NetImageNet denoising · ImageNet denoising testPSNRImproved
Use a smaller noise variance, σz=15 instead of σz=25 (ablation)U-Net denoiser · ImageNet denoisingPSNR: about 1.3dBImproved
Fix the noise per training image instead of re-sampling it each epoch (ablation)U-Net denoiser · ImageNet denoisingPSNR: 0.3dBHurt
Add DIV2K, Flickr2K, BSD500, and WED images to the training data (ablation)U-Net denoiser · ImageNet denoising testPSNR: 32.252dB -> 32.239dBHurt
Use 128×128 patches instead of 256×256 patches (ablation)U-Net denoiser · ImageNet denoising testPSNRHurt
Increase MRI optimizer choice from RMSprop to Adam (ablation)U-Net MRI reconstructor · fastMRI multi-coil brain (4x acceleration)SSIMNo clear effect
Increase batch size beyond 10 (ablation)U-Net denoiser · ImageNet denoising testPSNRNo clear effect
Increase network size for the large-training-set MRI settings (ablation)U-Net MRI reconstructor · fastMRI multi-coil brain (4x acceleration)SSIMNo clear effect

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