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

2026-01-28

AI Summary

This work addresses the high cost and limited deployability of multimodal large language models (MLLMs) in image quality assessment, which typically rely on extensive human-annotated mean opinion scores (MOS). The authors propose LEAF, a novel framework that decouples perceptual modeling from MOS scale calibration for the first time. LEAF leverages a teacher MLLM to generate dense supervision signals—specifically, pointwise quality judgments and pairwise preferences—which are then used to train a lightweight student regressor via knowledge distillation. Only a minimal number of MOS annotations are required during the final calibration stage. Experiments on both user-generated and AI-generated image benchmarks demonstrate that LEAF drastically reduces annotation demands while maintaining strong correlation with human ratings.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Computer Vision: Multi-modal Vision
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information

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.

Problem

• Image Quality Assessment
• Mean Opinion Score
• Label Efficiency
• Multimodal Large Language Models
• Calibration

Innovation

• Label-Efficient Learning
• Knowledge Distillation
• Image Quality Assessment
• MOS Calibration
• Multimodal Large Language Models