用反事实推理揭示图像质量感知的因果机制，提升模型可解释性。
Image Quality Assessment: Investigating Causal Perceptual Effects with Abductive Counterfactual Inference
- 基于反事实推理分析深层特征与感知失真的因果关系
- 在多个基准上表现优异，预测结果更贴近人类感知
- 不依赖网络结构，适用于各类深度模型，适合研究感知机制
现有的全参考图像质量评估（FR-IQA）方法难以捕捉人类对图像失真产生感知反应的复杂因果机制，限制了其在多样场景下的泛化能力。本文提出一种基于归纳反事实推理的FR-IQA方法，探究深层特征与感知失真之间的因果关系。首先，分析深层特征对感知的影响，将因果推理与特征对比结合，构建能有效处理多种失真类型的模型；其次，所提方法的感知因果分析独立于骨干网络架构，可适配多种深度网络。通过反事实实验验证了因果关系的有效性，确认了模型在感知相关性和评分可解释性上的优势。实验表明该方法在多个基准上具有鲁棒性与竞争力，性能表现优异。源代码已公开于 https://anonymous.4open.science/r/DeepCausalQuality-25BC。
原文摘要 · Abstract (English)
Existing full-reference image quality assessment (FR-IQA) methods often fail to capture the complex causal mechanisms that underlie human perceptual responses to image distortions, limiting their ability to generalize across diverse scenarios. In this paper, we propose an FR-IQA method based on abductive counterfactual inference to investigate the causal relationships between deep network features and perceptual distortions. First, we explore the causal effects of deep features on perception and integrate causal reasoning with feature comparison, constructing a model that effectively handles complex distortion types across different IQA scenarios. Second, the analysis of the perceptual causal correlations of our proposed method is independent of the backbone architecture and thus can be applied to a variety of deep networks. Through abductive counterfactual experiments, we validate the proposed causal relationships, confirming the model's superior perceptual relevance and interpretability of quality scores. The experimental results demonstrate the robustness and effectiveness of the method, providing competitive quality predictions across multiple benchmarks. The source code is available at https://anonymous.4open.science/r/DeepCausalQuality-25BC.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。