Vision-language models achieve at most 61.9% accuracy on identifying image distortion types and severities, falling short of human majority-vote performance at 65.7%.
A statistical evaluation of recent full reference image quality assessment algorithms.IEEE Transactions on Image Pro- cessing, 15(11):3440–3451
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.CV 2years
2026 2verdicts
UNVERDICTED 2representative citing papers
A self-supervised relational IQA system creates disentangled spatial distortion maps and contrastive quality scores from synthetic data alone, removing the need for human-labeled mean opinion scores.
citing papers explorer
-
DistortBench: Benchmarking Vision Language Models on Image Distortion Identification
Vision-language models achieve at most 61.9% accuracy on identifying image distortion types and severities, falling short of human majority-vote performance at 65.7%.
-
Pixel Perfect: Relational Image Quality Assessment with Spatially-Aware Distortions
A self-supervised relational IQA system creates disentangled spatial distortion maps and contrastive quality scores from synthetic data alone, removing the need for human-labeled mean opinion scores.