发表机构
University of Science and Technology Beijing; Sony China Research Laboratory; The Hong Kong University of Science and Technology; University of Exeter（北京科技大学; 索尼中国研究院; 香港科技大学; 埃克塞特大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
提出StructFlow-HPR，一种基于姿态条件流匹配的生成框架，用于增强5G CSI数据，在有限数据下提升人体姿态识别性能。
AI 中文摘要
随着对隐私保护和抗遮挡人体姿态识别（HPR）需求的日益增长，5G信道状态信息（CSI）通过融合通信与感知能力，提供了一种有前景的非接触式感知模态。然而，在实际5G系统中，收集大规模同步的CSI-姿态数据对仍然成本高昂。为解决这一局限，我们提出了StructFlow-HPR，一种用于生成式CSI增强的结构化姿态条件流匹配框架。StructFlow-HPR在姿态引导下学习从高斯噪声到真实CSI表示的连续潜在传输过程，同时通过重建保持的自编码器保留CSI的接收器-频率拓扑结构。进一步设计了一个姿态条件Transformer来建模潜在速度场，并通过常微分方程采样生成姿态对齐的CSI样本。在真实5G感知数据上的实验表明，StructFlow-HPR能够生成逼真的CSI-姿态数据对，并在数据有限条件下提升下游HPR性能。
英文摘要
With the growing demand for privacy-preserving and occlusion-resilient human pose recognition (HPR), 5G channel state information (CSI) offers a promising contactless sensing modality by integrating communication and sensing capabilities. However, collecting large-scale synchronized CSI-pose pairs remains costly in practical 5G systems. To address this limitation, we propose StructFlow-HPR, a structured pose-conditioned flow matching framework for generative CSI augmentation. StructFlow-HPR learns a continuous latent transport process from Gaussian noise to real CSI representations under pose guidance, while preserving the receiver-frequency topology of CSI through a reconstruction-preserving autoencoder. A pose-conditioned Transformer is further designed to model the latent velocity field and generate pose-aligned CSI samples via ordinary differential equation sampling. Experiments on real-world 5G sensing data show that StructFlow-HPR can produce realistic CSI-pose pairs and improve downstream HPR performance under limited-data conditions.
