4DRaL: Bridging 4D Radar with LiDAR for Place Recognition using Knowledge Distillation

TL;DR

This paper introduces 4DRaL, a framework that uses knowledge distillation from LiDAR models to improve 4D radar-based place recognition, achieving state-of-the-art results in all weather conditions.

Contribution

The novel 4DRaL framework enhances 4D radar place recognition by leveraging LiDAR models through multiple knowledge distillation modules, addressing noise and sparsity issues.

Findings

4DRaL outperforms existing methods in R2R and R2L place recognition tasks.

The framework maintains high accuracy under adverse weather conditions.

Knowledge distillation significantly improves 4D radar data utilization.

Abstract

Place recognition is crucial for loop closure detection and global localization in robotics. Although mainstream algorithms typically rely on cameras and LiDAR, these sensors are susceptible to adverse weather conditions. Fortunately, the recently developed 4D millimeter-wave radar (4D radar) offers a promising solution for all-weather place recognition. However, the inherent noise and sparsity in 4D radar data significantly limit its performance. Thus, in this paper, we propose a novel framework called 4DRaL that leverages knowledge distillation (KD) to enhance the place recognition performance of 4D radar. Its core is to adopt a high-performance LiDAR-to-LiDAR (L2L) place recognition model as a teacher to guide the training of a 4D radar-to-4D radar (R2R) place recognition model. 4DRaL comprises three key KD modules: a local image enhancement module to handle the sparsity of raw 4D radar points, a feature distribution distillation module that ensures the student model generates more discriminative features, and a response distillation module to maintain consistency in feature space between the teacher and student models. More importantly, 4DRaL can also be trained for 4D radar-to-LiDAR (R2L) place recognition through different module configurations. Experimental results prove that 4DRaL achieves state-of-the-art performance in both R2R and R2L tasks regardless of normal or adverse weather.