NavMapFusion: Diffusion-based Fusion of Navigation Maps for Online Vectorized HD Map Construction

2025-12-03

AI Summary

This work addresses the challenge of leveraging outdated coarse-grained navigation graphs (e.g., OpenStreetMap) to assist real-time high-definition (HD) map construction in dynamic environments. Methodologically, it introduces a diffusion-based map fusion framework that, for the first time, models map discrepancies as learnable noise; it employs the coarse graph as a structural prior and vehicle-mounted sensor features as conditional inputs, enabling spatiotemporal alignment and vectorized reconstruction via iterative conditional denoising. Concurrently, the framework suppresses outdated road segments while enhancing geometrically consistent regions—balancing accuracy and real-time feasibility. Evaluated on the nuScenes benchmark, it achieves a 21.4% relative improvement in map accuracy within 100 meters, with greater gains at larger scales, while satisfying online processing constraints. This work establishes a novel paradigm for cost-effective, timely crowdsourced HD map updating.

Technology Category

• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Intelligent Robots: Localization, Mapping, and Navigation
• Search and Optimization: Learning to Search

Application Category

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

Abstract

Accurate environmental representations are essential for autonomous driving, providing the foundation for safe and efficient navigation. Traditionally, high-definition (HD) maps are providing this representation of the static road infrastructure to the autonomous system a priori. However, because the real world is constantly changing, such maps must be constructed online from on-board sensor data. Navigation-grade standard-definition (SD) maps are widely available, but their resolution is insufficient for direct deployment. Instead, they can be used as coarse prior to guide the online map construction process. We propose NavMapFusion, a diffusion-based framework that performs iterative denoising conditioned on high-fidelity sensor data and on low-fidelity navigation maps. This paper strives to answer: (1) How can coarse, potentially outdated navigation maps guide online map construction? (2) What advantages do diffusion models offer for map fusion? We demonstrate that diffusion-based map construction provides a robust framework for map fusion. Our key insight is that discrepancies between the prior map and online perception naturally correspond to noise within the diffusion process; consistent regions reinforce the map construction, whereas outdated segments are suppressed. On the nuScenes benchmark, NavMapFusion conditioned on coarse road lines from OpenStreetMap data reaches a 21.4% relative improvement on 100 m, and even stronger improvements on larger perception ranges, while maintaining real-time capabilities. By fusing low-fidelity priors with high-fidelity sensor data, the proposed method generates accurate and up-to-date environment representations, guiding towards safer and more reliable autonomous driving. The code is available at https://github.com/tmonnin/navmapfusion

Problem

• Fuses low-fidelity navigation maps with sensor data for online HD map construction
• Uses diffusion models to iteratively denoise and update outdated map segments
• Improves accuracy of real-time environment representation for autonomous driving

Innovation

• Diffusion model fuses low-fidelity navigation maps with sensor data
• Iterative denoising suppresses outdated map segments and reinforces consistent regions
• Real-time online HD map construction using coarse priors like OpenStreetMap