A Target Detection Method of Automotive Millimeter Wave Radar Based on Deep Learning

This paper introduces a research on automotive radar target detection using deep learning. Automotive radar faces the complex road environment, and the correct judgment of the location and attributes of the target around the vehicle is the key to safe driving. These works all use the Range-Doppler image data for detection. This paper finds four problems in traditional CFAR through experiments, and proposes a target detection method based on deep learning, which uses the YOLOv5 algorithm to better solve these problems.