Abstract
Dear Editor, Human activity recognition (HAR) using WiFi signals has been a significant task due to its potential applications in for example, healthcare services and smart homes. This letter deals with the WiFi channel state information (CSI)-based HAR task. To capture the dynamics of human activities well from CSI without using a huge number of training samples, we propose a recurrent model of convolution blocks and transformer encoders. Firstly, the model utilizes the convolution blocks to capture local variation and the self-attention mechanism in transformer encoders to characterize long-range dependencies. Secondly and more importantly, the recurrent architecture models the context information well within CSI signals and allows the network to deepen without scale increase, making it particularly suited to learning from a small amount of CSI samples.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.