Understanding how people move using modern civilian radar
Human ambient intelligence is a concept that emerged over 20 years ago, but which remains elusive. Meanwhile, modern day civilian radar systems are becoming increasingly low-cost, low-power, and more capable, enabling their application to this domain. Deep learning has paved the way for great advancements in computer vision, but faces new challenges when applied to the RF domain. This talk will discuss how radar is different and current trends in how researchers are trying to solve these challenges. New ideas for physics-aware machine learning targeting RF signal classification will be presented, along with a challenge dataset specific to the classification of activities of daily living (ADL) using an RF sensor network.