AI Paper Atlas

Research map / Image generation / Super-resolution

Structural Similarity in Deep Features: Unified Image Quality Assessment Robust to Geometrically Disparate Reference

Keke Zhang, Weiling Chen, Tiesong Zhao, Zhou Wang

TPAMI 2025 · Image generation › Super-resolution · approach: Perceptual and real-world SR

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 attention calibration and local (windowed) correlation to the deep structure comparison (ablation)DeepSSIM-Lite · LIVESRCC: 0.955 -> 0.964Improved
add attention calibration and local (windowed) correlation to the deep structure comparison (ablation)DeepSSIM-Lite · CSIQSRCC: 0.939 -> 0.950Improved
add attention calibration and local (windowed) correlation to the deep structure comparison (ablation)DeepSSIM-Lite · TID2013SRCC: 0.797 -> 0.838Improved
add attention calibration and local (windowed) correlation to the deep structure comparison (ablation)DeepSSIM-Lite · KADID-10kSRCC: 0.877 -> 0.903Improved
add attention calibration and local (windowed) correlation to the deep structure comparison (ablation)DeepSSIM-Lite · PIPAL (train set)SRCC: 0.683 -> 0.699Improved
add attention calibration and local (windowed) correlation to the deep structure comparison (ablation)DeepSSIM-Lite · CVIU (SRIQA, HR ref)SRCC: 0.871 -> 0.891Improved
add attention calibration and local (windowed) correlation to the deep structure comparison (ablation)DeepSSIM-Lite · SISAR (SRIQA, HR ref)SRCC: 0.711 -> 0.750Improved
add attention calibration and local (windowed) correlation to the deep structure comparison (ablation)DeepSSIM-Lite · RetargetMe (IRQA)Mean KRCC: 0.279 -> 0.471Improved
add attention calibration and local (windowed) correlation to the deep structure comparison (ablation)DeepSSIM-Lite · NRID (IRQA)Mean KRCC: 0.429 -> 0.619Improved
add attention calibration and local (windowed) correlation to the deep structure comparison (ablation)DeepSSIM-Lite · CUHK (IRQA)SRCC: 0.600 -> 0.641Improved
use low-resolution instead of high-resolution reference for super-resolution IQA (ablation)DeepSSIM · QADS (SRIQA)SRCC: 0.840 -> 0.726Hurt
use low-resolution instead of high-resolution reference for super-resolution IQA (ablation)DeepSSIM · SISAR (SRIQA)SRCC: 0.750 -> 0.677Hurt

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