Abstract:Advanced driver assistance systems (ADASs) and autonomous vehicles rely on different types of sensors, such as camera, radar, ultrasonic, and LiDAR, to sense the surrounding environment. Compared with the other types of sensors, millimeter-wave automotive radar has advantages in terms of cost and reliability under bad weather conditions (e.g., snow, rain, and fog) and does not suffer from light condition variations (e.g., darkness). Typical radar devices used in today's commercial vehicles with ADAS features produce sparse point clouds in low angular resolution with a limited number of antennas. In this article, we present a machine-learning-aided signal processing chain to suppress the radar imaging blur effect introduced by the phase migration in time-division multiplexing multiple-input multiple-output radar, to generate low-level high-resolution radar bird's-eye view (BEV) spectra with rich object's features. Compared with radar point clouds, there is no information loss in radar BEV spectra. We then propose a temporal-fusion distance-tolerant single-stage object detection network, termed as TDRadarNet, and an enhanced version, TDRadarNet+, to robustly detect vehicles in both long and short ranges on radar BEVs. We introduce a first-of-its-kind multimodel dataset, containing 14 800 frames of high-resolution low-level radar BEV spectra with synchronized stereo camera RGB images and 3-D LiDAR point clouds. Our dataset achieves 0.39-m range resolution and 1.2 degrees degree azimuth angular resolution with 100-m maximum detectable range. Moreover, we create a subdataset, the Doppler Unfolding dataset, containing 244 140 beam vectors extracted from the 3-D radar data cube. With extensive testing and evaluation, we demonstrate that our Doppler unfolding network achieves 93.46% Doppler unfolding accuracy. Compared to YOLOv7, a state-of-the-art image-based object detection network, TDRadarNet, achieves a 70.3% average precision (AP) for vehicle detection, demonstrating a 21.0% improvement; TDRadarNet+ achieves a 73.9% AP, showing a 24.6% improvement in performance.

Discrimination of stationary from moving targets with recurrent neural networks in automotive Radar

Human Identification by Means of Optoelectronic Reservoir Computing

Targetless Rotational Auto-Calibration of Radar and Camera for Intelligent Transportation Systems

Simultaneous Clutter Detection and Semantic Segmentation of Moving Objects for Automotive Radar Data

Using Machine Learning to Detect Ghost Images in Automotive Radar

RadarTCN: Lightweight Online Classification Network for Automotive Radar Targets Based on TCN

A recurrent CNN for online object detection on raw radar frames

Segmentation-assisted Multi-frame Radar Target Detection Network in Clutter Traffic Scenarios

300 GHz Radar Object Recognition based on Deep Neural Networks and Transfer Learning

DeepHybrid: Deep Learning on Automotive Radar Spectra and Reflections for Object Classification

Machine Learning Approach for Obstacle Classification in Automotive Radar

Classification Of Automotive Targets Using Inverse Synthetic Aperture Radar Images

Machine Learning-Based Radar Indoor Human Target Loss Judgment

Radar-based Road User Classification and Novelty Detection with Recurrent Neural Network Ensembles

Automatic Target Recognition of Millimeter-Wave Radar Based on Deep Learning

Multi-Object Tracking based on Imaging Radar 3D Object Detection

See Further Than CFAR: a Data-Driven Radar Detector Trained by Lidar

Distant Vehicle Detection Using Radar and Vision

Spectranet: A High Resolution Imaging Radar Deep Neural Network for Autonomous Vehicles.

CenterRadarNet: Joint 3D Object Detection and Tracking Framework using 4D FMCW Radar

Deep-Neural-Network-Enabled Vehicle Detection Using High-Resolution Automotive Radar Imaging