Radio-frequency (RF) sensing for deep awareness of human physical status - part II

Abstract

The application of radar signal processing techniques, specifically with ultra-wideband (UWB) radar is investigated in this study. The intention is to non-invasively detect and classify human emotions using UWB radar. Features related to breathing and heart rate are extracted from the signal. Machine learning techniques are then applied to classify between emotions defined by the arousal-valence theory. The dataset was created by exposing subjects to different videos meant to incite different responses while being monitored by a UWB radar.