AutoPlace: Robust Place Recognition with Single-chip Automotive Radar

TL;DR

This paper introduces AutoPlace, a robust place recognition method for autonomous vehicles using low-cost single-chip automotive radar, which outperforms existing visual, LiDAR, and spinning radar approaches in real-world scenarios.

Contribution

AutoPlace is the first tailored place recognition approach specifically designed for single-chip automotive radar, enhancing robustness and leveraging radar-specific features.

Findings

Outperforms baseline methods on nuScenes dataset

Effective dynamic points removal improves recognition accuracy

Radar-specific features enhance robustness in real-world conditions

Abstract

This paper presents a novel place recognition approach to autonomous vehicles by using low-cost, single-chip automotive radar. Aimed at improving recognition robustness and fully exploiting the rich information provided by this emerging automotive radar, our approach follows a principled pipeline that comprises (1) dynamic points removal from instant Doppler measurement, (2) spatial-temporal feature embedding on radar point clouds, and (3) retrieved candidates refinement from Radar Cross Section measurement. Extensive experimental results on the public nuScenes dataset demonstrate that existing visual/LiDAR/spinning radar place recognition approaches are less suitable for single-chip automotive radar. In contrast, our purpose-built approach for automotive radar consistently outperforms a variety of baseline methods via a comprehensive set of metrics, providing insights into the efficacy when used in a realistic system.