Plug In, Grade Right: Psychology-Inspired AGIQA
Zhicheng Liao, Baoliang Chen, Hanwei Zhu, Lingyu Zhu, Shiqi Wang, Weisi Lin
TL;DR
This work tackles semantic drift in AGIQA by introducing AGRM-based Quality Grading (AGQG), a psychometrics-inspired framework that decouples image ability from quality difficulty using two branches. By constraining grade thresholds with an arithmetic sequence $\beta_k=\beta_1+(k-1)\gamma$ and enforcing $\gamma>\frac{2\ln 2}{D\alpha}$, AGQG guarantees a unimodal, interpretable distribution of grade probabilities and a unique optimal grade for any image. The module is designed as a plug-and-play component that can augment existing AGIQA models (e.g., CLIP-IQA, CLIP-AGIQAtang2025clip, IPCE, TIER), delivering consistent intra- and cross-dataset performance gains while maintaining minimal computational overhead. Across three benchmark datasets (AGIQA-1K, AGIQA-3K, AIGCIQA2023) and multiple quality dimensions (perceptual quality, authenticity, and prompt alignment), AGQG reduces semantic drift, improves alignment with human judgments, and demonstrates strong generalization, signaling its potential as a foundational piece for robust future IQA systems. The approach also provides theoretical guarantees of unimodality and interpretable thresholds, bridging psychometrics and vision-language evaluation in a practical, scalable way.
Abstract
Existing AGIQA models typically estimate image quality by measuring and aggregating the similarities between image embeddings and text embeddings derived from multi-grade quality descriptions. Although effective, we observe that such similarity distributions across grades usually exhibit multimodal patterns. For instance, an image embedding may show high similarity to both "excellent" and "poor" grade descriptions while deviating from the "good" one. We refer to this phenomenon as "semantic drift", where semantic inconsistencies between text embeddings and their intended descriptions undermine the reliability of text-image shared-space learning. To mitigate this issue, we draw inspiration from psychometrics and propose an improved Graded Response Model (GRM) for AGIQA. The GRM is a classical assessment model that categorizes a subject's ability across grades using test items with various difficulty levels. This paradigm aligns remarkably well with human quality rating, where image quality can be interpreted as an image's ability to meet various quality grades. Building on this philosophy, we design a two-branch quality grading module: one branch estimates image ability while the other constructs multiple difficulty levels. To ensure monotonicity in difficulty levels, we further model difficulty generation in an arithmetic manner, which inherently enforces a unimodal and interpretable quality distribution. Our Arithmetic GRM based Quality Grading (AGQG) module enjoys a plug-and-play advantage, consistently improving performance when integrated into various state-of-the-art AGIQA frameworks. Moreover, it also generalizes effectively to both natural and screen content image quality assessment, revealing its potential as a key component in future IQA models.