Dynamic gesture recognition with Wi-Fi based on signal processing and machine learning

Wi-Fi signals have been typically acting as information carriers in modern communication system, but recent research has revealed their powerful capability in detecting and identifying various targets. With Wi-Fi, we can now "see" people´s location, activity, and even hand gestures. In this paper, a new method of dynamic gesture recognition using Wi-Fi based on signal processing and machine learning is proposed. In our work, power profiles of received Wi-Fi signals are acquired for signal processing. The discrete wavelet transform (DWT) is applied to extract features and eliminate noise. And a support vector machine (SVM) improved by dynamic time warping (DTW) algorithm is built to classify and recognize different gestures. The experimental result shows that, by applying the method, nine predefined dynamic gestures can be effectively recognized, with an average recognition rate up to 94.8%, using only a small amount of training samples.