基于CSI实例标准化的域泛化人体动作识别模型
Domain-generalization human activity recognition model based on CSI instance normalization
王杨 1许佳炜 1王傲 1夏慧娟 1赵传信 1季一木2
作者信息
- 1. 安徽师范大学计算机与信息学院,安徽 芜湖 241002
- 2. 南京邮电大学计算机学院,江苏 南京 210003
- 折叠
摘要
为了实现完全不依赖目标域数据的Wi-Fi跨域人体动作感知,提出了一种基于CSI实例标准化的域泛化人体动作识别模型INDG-Fi.INDG-Fi使用实例标准化去除CSI特征表示的领域信息,接着构建共享特征提取的动作分类器和域分类器,并通过动作偏向学习和对抗性的域学习,将编码层提取的特征偏向人体动作引起的信号特征,同时远离领域信号影响.为了让模型关注受人体动作影响更显著的子载波信号,在编码层中加入子载波注意力模块.实现结果表明,所提INDG-Fi在不可见的用户和位置的感知性能分别为97.99%和92.73%,能够实现鲁棒的跨域感知.
Abstract
To achieve Wi-Fi cross-domain human activity perception that was not dependent on target domain data,a domain-generalization human activity recognition model based on CSI instance normalization called INDG-Fi was pro-posed.The instance normalization standardization was utilized to remove domain information from the representation of CSI features by INDG-Fi.Then action classifiers and domain classifiers were constructed for shared feature extraction.By employing activity bias learning and adversarial domain learning,the model biased the features extracted from the en-coding layer towards signal variations caused by human actions while moving away from domain signals.To enhance the model's focus on subcarrier signals that were more significantly influenced by human actions,a subcarrier attention mod-ule was incorporated into the encoding layer.The implemented results demonstrate that the proposed INDG-Fi achieves perceptual accuracies of 97.99%and 92.73%for unseen users and locations,respectively,thus enabling robust cross-domain perception.
关键词
信道状态信息/无线感知/人体动作识别/域泛化Key words
channel state information/wireless sensing/human activity recognition/domain-generalization引用本文复制引用
基金项目
国家自然科学基金资助项目(61871412)
江苏省重点研发计划基金资助项目(BE2023004-2)
安徽省自然科学基金重点项目(KJ2019A0938)
安徽省自然科学基金重点项目(KJ2021A1314)
安徽省自然科学基金重点项目(KJ2019A0979)
出版年
2024
国家科技期刊平台
NSTL
万方数据