Hand Gesture Recognition with the Fusion of ToF and UWB radar sensors

This paper proposed a fusion model of both radar and Time-of-Flight sensor data for the recognition of 12 daily used hand gestures. During the recognition process, a fusion model with the time-frequency analysis result of raw UWB radar data and the optical data after Mediapipe were constructed. A fusion of these two features to obtain further classification results. Finally, the fusion process yielded a classification accuracy of 95.42%.