Imaging Radars - Learning for Enhanced Vision in Modern Radar Systems

Explore the cutting-edge advancements in imaging radar technology through this tinyML Asia 2021 conference talk by Ankit Sharma, System Architect at Steradian. Delve into the evolution of modern radar systems, which have progressed from 3-D to 5-D imaging capabilities, with the fifth dimension representing object classification. Discover how millimeter-wave radars are pushing the boundaries of imaging resolution and learn about the application of supervised and unsupervised machine learning techniques in radar signal processing. Examine the novel use cases for traffic enforcement and autonomous driving, where object type classification becomes crucial. Gain insights into a real-time classification algorithm that utilizes radar point-cloud data and electromagnetic wave reflectivity at 80GHz to categorize detected objects. The talk covers the hardware and software architecture, system stack, and concludes with a discussion on the accuracy measures of the proposed classification method.