[2508.08570] Superclass-Guided Representation Disentanglement for Spurious Correlation Mitigation

Computer Science > Computer Vision and Pattern Recognition arXiv:2508.08570 (cs) [Submitted on 12 Aug 2025 (v1), last revised 19 Mar 2026 (this version, v2)] Title:Superclass-Guided Representation Disentanglement for Spurious Correlation Mitigation Authors:Chenruo Liu, Hongjun Liu, Zeyu Lai, Yiqiu Shen, Chen Zhao, Qi Lei View a PDF of the paper titled Superclass-Guided Representation Disentanglement for Spurious Correlation Mitigation, by Chenruo Liu and 5 other authors View PDF HTML (experimental) Abstract:To enhance group robustness to spurious correlations, prior work often relies on auxiliary group annotations and assumes identical sets of groups across training and test domains. To overcome these limitations, we propose to leverage superclasses -- categories that lie higher in the semantic hierarchy than the task's actual labels -- as a more intrinsic signal than group labels for discerning spurious correlations. Our model incorporates superclass guidance from a pretrained vision-language model via gradient-based attention alignment, and then integrates feature disentanglement with a theoretically supported minimax-optimal feature-usage strategy. As a result, our approach attains robustness to more complex group structures and spurious correlations, without the need to annotate any training samples. Experiments across diverse domain generalization tasks show that our method significantly outperforms strong baselines and goes well beyond the vision-language model's gui...