Implementation of Deep Learning-based Kick Gesture Recognition Using 60 GHz Radar Sensor

In the realm of automotive technology, hands-free trunk operation systems have emerged as a cornerstone of convenience and functionality. Predominantly reliant on ultra-sonic and camera-based sensors, these systems, however, falter in adverse weather conditions such as rain, snow, and fog, and are limited by short detection ranges and susceptibility to false alarms due to unintentional movements. Addressing these limitations, this paper proposes an innovative solution utilizing a 60 GHz frequency-modulated continuous wave radar sensor. This system is resilient in various weather conditions and has a robust detection mechanism for kick gestures. By employing two-dimensional fast Fourier transform for range and velocity analysis and an antenna array system for azimuth angle extraction, our radar sensor effectively discerns kick gestures. The data, once processed, is channeled through controller area network flexible data-rate communication to a deep learning classifier. This classifier, tested for accuracy, successfully identifies valid kick gestures with a remarkable 99.75% accuracy. Our findings present a significant advancement in sensor technology for smart trunk systems, paving the way for more reliable and efficient hands-free automotive operations.