Shallow Neural Networks for mmWave Radar Based Recognition of Vulnerable Road Users

Abstract

A millimetre-wave (mmWave) radar, having synergies with the multi-beam light detection and ranging (LiDAR) and cameras, has been considered as a must-have sensor in the connected and autonomous vehicles (CAV) in the future intelligent transportation systems (ITS). Besides the traditional target detection and ranging functions, the mmWave radar is expected to perform more intelligent tasks to improve the road safety, for example recognising the targets, especially the vulnerable road users like pedestrians and cyclists. This paper describes a simulation study of the micro-Doppler signatures of the pedestrian and cyclist based on mmWave vehicle radar and investigates the recognition capabilities through both the convolutional neural networks (CNN), recurrent neural networks (RNN) and mixed convolutional and recurrent approach respectively. The result demonstrates the usability of the mmWave radar Doppler information and complementary with the video and laser data streams in the CAV auto-piloting.