From Single Scan to Sequential Consistency: A New Paradigm for LIDAR Relocalization

TL;DR

TempLoc introduces a novel LiDAR relocalization framework that leverages sequential consistency and attention mechanisms to improve robustness and accuracy in dynamic environments.

Contribution

It proposes a comprehensive framework combining global coordinate estimation, attention-based correspondence, and uncertainty-guided fusion for enhanced LiDAR relocalization.

Findings

Outperforms state-of-the-art methods on NCLT and Oxford Robot-Car datasets.

Effectively models spatio-temporal consistency for more accurate localization.

Demonstrates robustness in dynamic and ambiguous scenarios.

Abstract

LiDAR relocalization aims to estimate the global 6-DoF pose of a sensor in the environment. However, existing regression-based approaches are prone to dynamic or ambiguous scenarios, as they either solely rely on single-frame inference or neglect the spatio-temporal consistency across scans. In this paper, we propose TempLoc, a new LiDAR relocalization framework that enhances the robustness of localization by effectively modeling sequential consistency. Specifically, a Global Coordinate Estimation module is first introduced to predict point-wise global coordinates and associated uncertainties for each LiDAR scan. A Prior Coordinate Generation module is then presented to estimate inter-frame point correspondences by the attention mechanism. Lastly, an Uncertainty-Guided Coordinate Fusion module is deployed to integrate both predictions of point correspondence in an end-to-end fashion, yielding a more temporally consistent and accurate global 6-DoF pose. Experimental results on the NCLT and Oxford Robot-Car benchmarks show that our TempLoc outperforms stateof-the-art methods by a large margin, demonstrating the effectiveness of temporal-aware correspondence modeling in LiDAR relocalization. Our code will be released soon.