UM3: Unsupervised Map to Map Matching

2025-08-22

AI Summary

Map-to-map matching faces three major challenges: absence of ground-truth correspondences, sparse node features, and poor scalability to large-scale maps. To address these, we propose the first fully unsupervised graph neural network framework. Our method introduces (1) pseudo-coordinate encoding to enrich node geometric representations and enable scale-invariant feature learning; (2) an adaptive feature-geometry similarity fusion mechanism jointly optimized with a geometric consistency loss to enhance robustness; and (3) a tiling-based overlapping partitioning strategy coupled with majority-voting post-processing to enable efficient parallel inference. Evaluated on real-world multi-source map datasets, our approach significantly outperforms existing supervised and unsupervised methods—achieving state-of-the-art accuracy, especially under high noise and at large scale. The results demonstrate its effectiveness, scalability, and practical utility for real-world map alignment tasks.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Intelligent Robots: Localization, Mapping, and Navigation
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Map-to-map matching is a critical task for aligning spatial data across heterogeneous sources, yet it remains challenging due to the lack of ground truth correspondences, sparse node features, and scalability demands. In this paper, we propose an unsupervised graph-based framework that addresses these challenges through three key innovations. First, our method is an unsupervised learning approach that requires no training data, which is crucial for large-scale map data where obtaining labeled training samples is challenging. Second, we introduce pseudo coordinates that capture the relative spatial layout of nodes within each map, which enhances feature discriminability and enables scale-invariant learning. Third, we design an mechanism to adaptively balance feature and geometric similarity, as well as a geometric-consistent loss function, ensuring robustness to noisy or incomplete coordinate data. At the implementation level, to handle large-scale maps, we develop a tile-based post-processing pipeline with overlapping regions and majority voting, which enables parallel processing while preserving boundary coherence. Experiments on real-world datasets demonstrate that our method achieves state-of-the-art accuracy in matching tasks, surpassing existing methods by a large margin, particularly in high-noise and large-scale scenarios. Our framework provides a scalable and practical solution for map alignment, offering a robust and efficient alternative to traditional approaches.

Problem

• Unsupervised map matching without training data
• Aligning spatial data across heterogeneous sources
• Handling large-scale maps with noisy coordinates

Innovation

• Unsupervised graph-based framework without training data
• Pseudo coordinates enable scale-invariant spatial learning
• Adaptive similarity balancing with geometric-consistent loss function