Image quality system learns to spot and judge damage step by step
From Glance to Scrutiny: Progressive Distortion Reasoning for Fine-Grained Image Quality Assessment
Computer Vision and Pattern Recognition
Summary
Picture quality can be hard to judge, especially when damage varies in type and location. This study introduces a method called GS-IQA that looks at images first broadly to find damage, then closely examines what kind it is and how bad. It trains the system in stages, rewarding it only after it correctly finds and identifies damage before rating severity. To test their method well, the researchers made a detailed collection of images showing many damage types and levels. Their system not only finds and measures damage better than others but also helps improve overall image quality assessments.
Image quality assessmentMulti-modal large language modelsDistortion localizationReinforcement learningSeverity estimationProgressive reasoningPerceptual bottlenecksDistortion recognitionBenchmark datasetOrdinal severity levels
Authors
Aoting Zhang, Mingze Gao, Dongbao Yang, Longyi Chen, Daoxin Zhang, Yi Wu, Yao Hu, Yu Zhou
Abstract
Multi-modal large language models (MLLMs) have demonstrated significant potential in image quality assessment (IQA) by bridging visual perception with descriptive evaluations. However, existing approaches mainly focus on holistic quality prediction, often functioning as black boxes that provide limited insight into where distortions occur and how they affect perceived quality, hindering fine-grained analysis of localized and heterogeneous degradations. We propose GS-IQA, a framework that reformulates IQA as a progressive Where--What--How diagnosis, emulating the human perceptual process from an initial glance to closer scrutiny. Since a severity judgment is meaningful only for a correctly localized and recognized region, we realize this progression through a two-stage reinforcement learning paradigm that respects such dependencies: the glance stage uses a perception-gated reward to establish where degradations lie and what they are, activating severity feedback only once both are correct, while the scrutiny stage introduces online reward-conditioned degradation generation to synthesize hard examples targeted at the model's perceptual bottlenecks, sharpening its discrimination of subtle severity variations. To enable systematic evaluation, we construct Diag-Bench, a region-level IQA benchmark of about 25K curated samples spanning 12 distortion types and five ordinal severity levels. Extensive experiments show that GS-IQA consistently surpasses state-of-the-art methods in distortion localization, recognition, and severity estimation, and that its diagnostic representations transfer effectively to conventional global quality prediction across diverse external benchmarks. Code and data will be released.