eXplaining to Learn (eX2L): 基于对比视觉解释对的正则化以应对分布偏移
eXplaining to Learn (eX2L): Regularization Using Contrastive Visual Explanation Pairs for Distribution Shifts
AI总结：
eX2L通过对比视觉解释对正则化，解耦分类器的潜在表示中的混淆特征，提升模型在分布偏移下的性能，实现功能域不变性。
摘要：
尽管已有大量研究致力于缓解分布偏移，但许多现有算法性能不一致，常无法在不同场景中超越基准经验风险最小化（ERM）。此外，高算法复杂度限制了可解释性，并仅能间接解决虚假相关性。我们提出eXplaining to Learn（eX2L）：一个可解释、基于解释的框架，在训练过程中解耦分类器潜在表示中的混淆特征。eX2L通过惩罚主标签分类器生成的Grad-CAM激活图与同时训练的混淆分类器生成的激活图之间的相似性来实现此目标。在严格的Spawrious Many-to-Many Hard Challenge基准测试中，eX2L实现了平均准确率（AA）为82.24%±3.87%和最差组准确率（WGA）为66.31%±8.73%，分别超越当前最佳（SOTA）5.49%和10.90%。除了其竞争性性能外，eX2L还证明了通过在组层面显式解耦标签和噪声属性可以实现功能域不变性。
英文摘要：
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 and necessitating newer algorithms which can handle scenarios where existing algorithms currently underperform. 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 synthetic data 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 enforced by explicitly decoupling label and nuisance attributes at the group level.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
