Causal Disentanglement for Full-Reference Image Quality Assessment

2026-04-23

AI Summary

This work addresses the limitations of existing full-reference image quality assessment methods, which rely on deep feature comparisons and struggle to model the causal relationship between image content and distortion, thereby exhibiting restricted generalization. To overcome this, the study introduces a causal disentanglement mechanism that separates content and distortion representations, treating distortion estimation as an intervention on latent representations. Inspired by human visual masking effects, a masking module is incorporated to extract content-modulated distortion features for quality prediction. The proposed method operates effectively under supervised, few-shot, and unsupervised settings and enables annotation-free cross-domain adaptation. It achieves state-of-the-art performance across standard benchmarks and diverse non-standard imaging domains—including underwater, medical, and neutron imaging—with particularly significant gains in unsupervised cross-domain scenarios over existing zero-shot models.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Machine Learning: Calibration & Uncertainty Quantification
• Reasoning under Uncertainty: Causality

Application Category

• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Existing deep network-based full-reference image quality assessment (FR-IQA) models typically work by performing pairwise comparisons of deep features from the reference and distorted images. In this paper, we approach this problem from a different perspective and propose a novel FR-IQA paradigm based on causal inference and decoupled representation learning. Unlike typical feature comparison-based FR-IQA models, our approach formulates degradation estimation as a causal disentanglement process guided by intervention on latent representations. We first decouple degradation and content representations by exploiting the content invariance between the reference and distorted images. Second, inspired by the human visual masking effect, we design a masking module to model the causal relationship between image content and degradation features, thereby extracting content-influenced degradation features from distorted images. Finally, quality scores are predicted from these degradation features using either supervised regression or label-free dimensionality reduction. Extensive experiments demonstrate that our method achieves highly competitive performance on standard IQA benchmarks across fully supervised, few-label, and label-free settings. Furthermore, we evaluate the approach on diverse non-standard natural image domains with scarce data, including underwater, radiographic, medical, neutron, and screen-content images. Benefiting from its ability to perform scenario-specific training and prediction without labeled IQA data, our method exhibits superior cross-domain generalization compared to existing training-free FR-IQA models.

Problem

• Full-Reference Image Quality Assessment
• Causal Disentanglement
• Cross-Domain Generalization
• Label-Free Learning
• Degradation Estimation

Innovation

• causal disentanglement
• decoupled representation learning
• visual masking effect
• label-free IQA
• cross-domain generalization