Driving by Hybrid Navigation: An Online HD-SD Map Association Framework and Benchmark for Autonomous Vehicles

2025-07-10

AI Summary

In autonomous driving, the disconnect between globally standardized (SD) maps and locally high-definition (HD) maps hinders robust hybrid navigation. To address this, we propose OMA—the first benchmark for online HD-to-global SD map association—featuring a large-scale annotated dataset of 480K road segments and 260K lane paths, along with dedicated evaluation metrics. To tackle the joint geometric-topological modeling challenge, we introduce the Map Association Transformer (MAT), a novel architecture integrating path-aware attention and spatial attention to enable robust cross-scale matching of map elements. Extensive experiments demonstrate that MAT significantly improves map association accuracy and navigation robustness over prior methods. Our work establishes a reproducible baseline and standardized evaluation paradigm for end-to-end hybrid navigation, bridging the gap between global SD and local HD map representations.

Technology Category

• Intelligent Robots: Localization, Mapping, and Navigation
• Planning, Routing, and Scheduling: Mixed Discrete/Continuous Planning
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Autonomous vehicles rely on global standard-definition (SD) maps for road-level route planning and online local high-definition (HD) maps for lane-level navigation. However, recent work concentrates on construct online HD maps, often overlooking the association of global SD maps with online HD maps for hybrid navigation, making challenges in utilizing online HD maps in the real world. Observing the lack of the capability of autonomous vehicles in navigation, we introduce extbf{O}nline extbf{M}ap extbf{A}ssociation, the first benchmark for the association of hybrid navigation-oriented online maps, which enhances the planning capabilities of autonomous vehicles. Based on existing datasets, the OMA contains 480k of roads and 260k of lane paths and provides the corresponding metrics to evaluate the performance of the model. Additionally, we propose a novel framework, named Map Association Transformer, as the baseline method, using path-aware attention and spatial attention mechanisms to enable the understanding of geometric and topological correspondences. The code and dataset can be accessed at https://github.com/WallelWan/OMA-MAT.

Problem

• Associating global SD maps with online HD maps for hybrid navigation
• Enhancing autonomous vehicle planning capabilities through map association
• Understanding geometric and topological correspondences in hybrid navigation

Innovation

• Hybrid navigation with HD-SD map association
• Path-aware and spatial attention mechanisms
• First benchmark for hybrid navigation maps