Human Activity Detection Algorithm Based on TSACT Model

Human activity recognition technology based on WiFi has attracted considerable attention in health monitoring and medical treatment due to its characteristics, including the absence of additional equipment, strong privacy protection, and comprehensive coverage area. However, due to the high complexity involved in WiFi signal detection and human activity, the recognition rate of human activity based on WiFi signal detection is not ideal. To effectively explore the mapping relationship between WiFi signal variations and human activity, this paper proposes a TSACT model(Combined Time and Space Domain Information for Human Activity Detection). The TSACT model recognizes that CSI signals encompass both temporal and spatial feature information, and relying on information from only one dimension may result in the omission of critical data, thereby affecting the final activity recognition rate. Therefore, TSACT integrates temporal and spatial characteristics to identify human activities. In the conclusion of the paper, the recognition rates of human activities using temporal features, spatial features, and the TSACT model are compared. The proposed scheme achieves a higher recognition rate.