Hand Gesture Recognition with Ensemble Time-Frequency Signatures Using Enhanced Deep Convolutional Neural Network
Xiang Feng, Qun Song, Qingfang Guo, Duo Liu, Zhanfeng Zhao, Yinan Zhao · 2019
Hand gesture recognition using radar has been widely applied to control electronic appliances, military appliances and so on. In this paper, we investigate the feasibility of recognizing hand gestures using fused multiple time-frequency signatures, which ensembles micro-Doppler signatures, range-time signatures and angle-time signatures on spectrograms, with an Enhanced Deep Convolutional Neural Network (EDCNN). Several typical gestures included Tick, Double pushing, Rotating clockwise, and Rotating counterclockwise, were measured using Mm-wave radar and their spectrograms investigated. Therein EDCNN was employed to classify the spectrograms, with 80% of the data utilized for training and the remaining 20% for validation. Simulation said that the classification accuracy of the proposed method was found to be 96.2%.