Decoupling Perception and Calibration: Label-Efficient Image Quality Assessment Framework

TL;DR

This paper introduces LEAF, a framework that distills perceptual quality priors from large language models into a lightweight regressor, enabling efficient image quality assessment with minimal human annotations.

Contribution

LEAF is the first framework to decouple perception and calibration in IQA, significantly reducing annotation needs while maintaining high correlation with human judgments.

Findings

Reduces human annotation requirements for IQA.

Maintains strong correlation with human MOS.

Effective on both user-generated and AI-generated images.

Abstract

Recent multimodal large language models (MLLMs) have demonstrated strong capabilities in image quality assessment (IQA) tasks. However, adapting such large-scale models is computationally expensive and still relies on substantial Mean Opinion Score (MOS) annotations. We argue that for MLLM-based IQA, the core bottleneck lies not in the quality perception capacity of MLLMs, but in MOS scale calibration. Therefore, we propose LEAF, a Label-Efficient Image Quality Assessment Framework that distills perceptual quality priors from an MLLM teacher into a lightweight student regressor, enabling MOS calibration with minimal human supervision. Specifically, the teacher conducts dense supervision through point-wise judgments and pair-wise preferences, with an estimate of decision reliability. Guided by these signals, the student learns the teacher's quality perception patterns through joint distillation and is calibrated on a small MOS subset to align with human annotations. Experiments on both user-generated and AI-generated IQA benchmarks demonstrate that our method significantly reduces the need for human annotations while maintaining strong MOS-aligned correlations, making lightweight IQA practical under limited annotation budgets.