AI Paper Atlas

Research map / Image generation / Image restoration

Ultra-High-Resolution Unpaired Stain Transformation via Kernelized Instance Normalization

Ming-Yang Ho, Min-Sheng Wu, Che-Ming Wu

ECCV 2022 · Image generation › Image restoration · approach: Efficient restoration architectures

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
replace patch-wise IN with KIN at test timeCUT · Glioma (testing set)FID: 105.66 -> 100.90Improved
replace patch-wise IN with KIN at test timeCUT · Kyoto summer2autumn (training set)FID: 77.59 -> 75.27Improved
replace patch-wise IN with KIN at test timeCUT · Kyoto summer2autumn (testing set)FID: 157.04 -> 167.15Hurt
replace patch-wise IN with KIN at test timeCUT · ANHIR breast FIDFID: 71.00 -> 72.59No clear effect
replace patch-wise IN with KIN at test timeCycleGAN · ANHIR breast FIDFID: 98.60 -> 96.09Improved
replace patch-wise IN with KIN at test timeCycleGAN · Glioma (testing set)FID: 142.28 -> 150.93Hurt
replace patch-wise IN with KIN at test timeL-LSeSim · ANHIR breast FIDFID: 65.82 -> 67.46No clear effect
replace TIN (global statistics) with KINCUT · ANHIR lung lesion FIDFID: 251.48 -> 56.38Improved
replace TIN (global statistics) with KINCUT · Glioma (testing set)FID: 200.56 -> 100.90Improved
replace TIN (global statistics) with KINCUT · ANHIR COAD FIDFID: 91.81 -> 93.68No clear effect
add KIN instead of patch-wise IN, human quality and fidelity evaluationCUT · ANHIR, glioma, Kyotohuman preferenceImproved
increase KIN kernel size from 1 to infinity (ablation)KIN · stain transformationqualitative outputMixed

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