Enhancing mmWave Radar Point Cloud via Visual-inertial Supervision

TL;DR

This paper introduces mmEMP, a novel supervised learning method that enhances mmWave radar point clouds using low-cost visual-inertial data, improving scene understanding without relying on expensive LiDAR.

Contribution

The paper presents a new approach combining visual-inertial supervision with radar data enhancement, including a dynamic 3D reconstruction algorithm and a neural network to densify and denoise radar point clouds.

Findings

mmEMP achieves performance comparable to LiDAR-supervised methods.

Enhanced radar point clouds improve object detection, localization, and mapping.

The approach enables crowdsourcing training data from commercial vehicles.

Abstract

Complementary to prevalent LiDAR and camera systems, millimeter-wave (mmWave) radar is robust to adverse weather conditions like fog, rainstorms, and blizzards but offers sparse point clouds. Current techniques enhance the point cloud by the supervision of LiDAR's data. However, high-performance LiDAR is notably expensive and is not commonly available on vehicles. This paper presents mmEMP, a supervised learning approach that enhances radar point clouds using a low-cost camera and an inertial measurement unit (IMU), enabling crowdsourcing training data from commercial vehicles. Bringing the visual-inertial (VI) supervision is challenging due to the spatial agnostic of dynamic objects. Moreover, spurious radar points from the curse of RF multipath make robots misunderstand the scene. mmEMP first devises a dynamic 3D reconstruction algorithm that restores the 3D positions of dynamic features. Then, we design a neural network that densifies radar data and eliminates spurious radar points. We build a new dataset in the real world. Extensive experiments show that mmEMP achieves competitive performance compared with the SOTA approach training by LiDAR's data. In addition, we use the enhanced point cloud to perform object detection, localization, and mapping to demonstrate mmEMP's effectiveness.