Research and Development of Gesture Recognition Technology Based on WiFi

As human-computer interaction continues to evolve towards a natural and non-contact mode, gesture recognition technology based on wifi has become a hot topic in the field of intelligent interaction due to its advantages such as relatively low cost, weak correlation with devices, and strong anti-interference ability. This article provides a systematic review of the core modules, typical solutions, key challenges and future applications of this technology. This study first compared traditional sensing technologies, highlighting the advantages of WiFi-based channel state information, namely CSI. It also provided an overview of key technologies such as CSI feature extraction and model design. Then, it investigated the principles, performance, and limitations of five representative methods. These five methods are HandFi, CSI-DeepNet, TransferSense, GESFI and C + SVM respectively. The solutions to challenges such as the cross-scenario generalization gap were summarized, and potential applications like hearing impairment assistance were explored. The research results show that the existing technologies have achieved high-precision and low-resource recognition, but there are some limitations in subtle gesture detection and cross-scenario adaptability. Future research should focus on refining fine-grained CSI feature extraction and exploring parametric-free domain adaptive methods.