用因果推理提升动作分析可解释性，更好识别疼痛保护行为。
CauSkelNet: Causal Representation Learning for Human Behaviour Analysis
- 基于因果推断构建关节间关系图，增强模型可解释性。
- 在EmoPain数据集上准确率与召回率显著优于传统GCN。
- 适合医疗健康领域中需要理解动作意图的研究者使用。
传统动作识别方法常因模型可解释性差、难以捕捉人类运动动态而受限。本文提出一种基于因果推断的表征学习框架，采用两阶段方法：先利用Peter-Clark（PC）算法和Kullback-Leibler（KL）散度识别并量化关节间的因果关系，再构建因果图卷积网络（Causal GCN），以捕捉关节交互特征，生成可解释且鲁棒的表征。在EmoPain数据集上的实验表明，该方法在准确率、F1分数和召回率上均优于传统GCN，尤其在检测保护性行为方面表现突出。本研究推动了人体运动分析的发展，为自适应智能医疗解决方案奠定基础。
原文摘要 · Abstract (English)
Traditional machine learning methods for movement recognition often struggle with limited model interpretability and a lack of insight into human movement dynamics. This study introduces a novel representation learning framework based on causal inference to address these challenges. Our two-stage approach combines the Peter-Clark (PC) algorithm and Kullback-Leibler (KL) divergence to identify and quantify causal relationships between human joints. By capturing joint interactions, the proposed causal Graph Convolutional Network (GCN) produces interpretable and robust representations. Experimental results on the EmoPain dataset demonstrate that the causal GCN outperforms traditional GCNs in accuracy, F1 score, and recall, particularly in detecting protective behaviors. This work contributes to advancing human motion analysis and lays a foundation for adaptive and intelligent healthcare solutions.
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。