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

2026-02-03

AI Summary

Existing LiDAR-based relocalization methods suffer from insufficient robustness in dynamic or ambiguous environments, primarily due to their reliance on single-frame inference and neglect of temporal consistency. This work proposes TempLoc, a novel framework that introduces, for the first time, a temporally aware point correspondence modeling mechanism to enable end-to-end 6-DoF pose estimation. TempLoc integrates global coordinate estimation, prior coordinate generation, and an uncertainty-guided fusion strategy, leveraging attention mechanisms and point-level uncertainty modeling to enforce cross-frame geometric consistency. Evaluated on the NCLT and Oxford RobotCar benchmarks, the method significantly outperforms state-of-the-art approaches, demonstrating that explicit modeling of temporal consistency effectively enhances both the robustness and accuracy of LiDAR relocalization.

Technology Category

• Intelligent Robots: Localization, Mapping, and Navigation
• Planning, Routing, and Scheduling: Temporal Planning
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Systems and Infrastructure for Web, Mobile and WoT: Location- and context-aware Web and WoT applications and services
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs

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.

Problem

• LiDAR relocalization
• sequential consistency
• 6-DoF pose estimation
• spatio-temporal consistency
• dynamic environments

Innovation

• LiDAR relocalization
• sequential consistency
• temporal-aware correspondence
• uncertainty-guided fusion
• attention mechanism