Development of a human gesture dataset using 802.11ax WiFi CSI

• Main Author: Ma, Zhenduo
• Other Authors: Xie Lihua
• Format: Thesis-Master by Coursework
• Language: English
• Published: Nanyang Technological University 2024
• Subjects: Computer and Information Science; WiFi CSI; Deep learning; Embedded system
• Online Access: https://hdl.handle.net/10356/177796
• Institution: Nanyang Technological University

This study explores the capturing of WiFi Channel State Information (CSI) using the AX-CSI tool and constructs a unique dataset comprising hand-drawn numerical gestures. High-quality CSI data are captured from 802.11ax WiFi signals using this tool, and the gestures are accurately classified through deep learning models, including CNNs, LSTMs, and MLPs. The research demonstrates the potential of WiFi CSI as a powerful, non-invasive tool for human activity monitoring and interaction, offering promising avenues for innovation in smart environment applications. This highlights the utility and effectiveness of the AX-CSI tool in enhancing gesture recognition technologies within smart environments.