GaitGCI：用于步态识别的生成式反事实干预
GaitGCI: Generative Counterfactual Intervention for Gait Recognition
AI总结：
针对步态识别受混杂因素干扰的问题，提出 GaitGCI 框架，通过 CIL 与 DCDC 生成并区分事实/反事实注意力，以低成本提升模型判别性、可解释性和跨场景识别性能。
摘要：
步态是最有前景的生物特征之一，旨在根据行人的行走模式识别其身份。然而，主流方法容易受到混杂因素影响，导致网络难以关注反映有效行走模式的区域。为解决步态识别中的这一根本问题，我们提出了一个名为 GaitGCI 的生成式反事实干预框架，由反事实干预学习（CIL）和多样性约束动态卷积（DCDC）组成。CIL 通过最大化事实/反事实注意力之间的似然差异来消除混杂因素的影响，而 DCDC 自适应地生成逐样本的事实/反事实注意力，以高效感知逐样本属性。借助矩阵分解和多样性约束，DCDC 保证模型兼具效率与效果。大量实验表明，所提出的 GaitGCI：1）能够有效关注反映步态模式的、具有判别性且可解释的区域；2）与模型无关，可插入现有模型中提升性能，且几乎不产生额外成本；3）在任意场景（实验室内和真实野外环境）中高效达到最先进性能。
英文摘要：
Gait is one of the most promising biometrics that aims to identify pedestrians from their walking patterns. However, prevailing methods are susceptible to confounders, resulting in the networks hardly focusing on the regions that reflect effective walking patterns. To address this fundamental problem in gait recognition, we propose a Generative Counterfactual Intervention framework, dubbed GaitGCI, consisting of Counterfactual Intervention Learning (CIL) and Diversity-Constrained Dynamic Convolution (DCDC). CIL eliminates the impacts of confounders by maximizing the likelihood difference between factual/counterfactual attention while DCDC adaptively generates sample-wise factual/counterfactual attention to efficiently perceive the sample-wise properties. With matrix decomposition and diversity constraint, DCDC guarantees the model to be efficient and effective. Extensive experiments indicate that proposed GaitGCI: 1) could effectively focus on the discriminative and interpretable regions that reflect gait pattern; 2) is model-agnostic and could be plugged into existing models to improve performance with nearly no extra cost; 3) efficiently achieves state-of-the-art performance on arbitrary scenarios (in-the-lab and in-the-wild).
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。
