Contrastive Learning for Unsupervised Radar Place Recognition

TL;DR

This paper introduces an unsupervised contrastive learning method for radar place recognition that leverages temporal data augmentation, achieving state-of-the-art accuracy on urban radar datasets and demonstrating improved understanding of complex re-localisation scenarios.

Contribution

The paper presents a novel unsupervised contrastive learning approach tailored for radar place recognition, exploiting temporal data for augmentation and outperforming existing methods.

Findings

Achieves 98.38% accuracy in re-localisation tasks.

Outperforms previous radar place recognition methods.

Better handles out-of-lane loop closures at arbitrary orientations.

Abstract

We learn, in an unsupervised way, an embedding from sequences of radar images that is suitable for solving the place recognition problem with complex radar data. Our method is based on invariant instance feature learning but is tailored for the task of re-localisation by exploiting for data augmentation the temporal successivity of data as collected by a mobile platform moving through the scene smoothly. We experiment across two prominent urban radar datasets totalling over 400 km of driving and show that we achieve a new radar place recognition state-of-the-art. Specifically, the proposed system proves correct for 98.38% of the queries that it is presented with over a challenging re-localisation sequence, using only the single nearest neighbour in the learned metric space. We also find that our learned model shows better understanding of out-of-lane loop closures at arbitrary orientation than non-learned radar scan descriptors.