通过对比解释图增强模型鲁棒性，有效应对分布偏移问题
eXplaining to Learn (eX2L): Regularization Using Contrastive Visual Explanation Pairs for Distribution Shifts
- 用双重分类器生成激活图对比，强制解耦标签与干扰特征
- 在硬挑战数据集上平均准确率82.24%，最差组准确率66.31%
- 方法可解释且适合需要鲁棒性的实际部署场景
尽管已有大量研究致力于缓解分布偏移问题，但许多现有算法表现不稳定，常无法在多样场景中超越基础的经验风险最小化（ERM）方法。此外，高算法复杂度往往削弱可解释性，且仅间接处理虚假相关。本文提出eXplaining to Learn（eX2L）：一种可解释的、基于解释的框架，在训练过程中通过惩罚主标签分类器与同时训练的混淆因子分类器生成的Grad-CAM激活图之间的相似性，实现对分类器潜在表示中混淆特征的解耦。在严格的Spurious Many-to-Many Hard Challenge基准测试中，eX2L达到平均准确率（AA）82.24% ± 3.87%、最差组准确率（WGA）66.31% ± 8.73%，分别优于当前最优（SOTA）方法5.49%和10.90%。eX2L还表明，通过在组级别显式解耦标签与噪声属性，可实现功能域不变性。
原文摘要 · Abstract (English)
Despite extensive research into mitigating distribution shifts, many existing algorithms yield inconsistent performance, often failing to outperform baseline Empirical Risk Minimization (ERM) across diverse scenarios. Furthermore, high algorithmic complexity frequently limits interpretability and offers only an indirect means of addressing spurious correlations. We propose eXplaining to Learn (eX2L): an interpretable, explanation-based framework that decorrelates confounding features from a classifier's latent representations during training. eX2L achieves this by penalizing the similarity between Grad-CAM activation maps generated by a primary label classifier and those from a concurrently trained confounder classifier. On the rigorous Spawrious Many-to-Many Hard Challenge benchmark, eX2L achieves an average accuracy (AA) of 82.24% +/- 3.87% and a worst-group accuracy (WGA) of 66.31% +/- 8.73%, outperforming the current state-of-the-art (SOTA) by 5.49% and 10.90%, respectively. Beyond its competitive performance, eX2L demonstrates that functional domain invariance can be achieved by explicitly decoupling label and nuisance attributes at the group level.
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。