High-Definition Maps (HD maps) are essential for the precise navigation and decision-making of autonomous vehicles, yet their creation and upkeep present significant cost and timeliness challenges. The online construction of HD maps using on-board sensors has emerged as a promising solution; however, these methods can be impeded by incomplete data due to occlusions and inclement weather. This paper proposes the PriorDrive framework to addresses these limitations by harnessing the power of prior maps, significantly enhancing the robustness and accuracy of online HD map construction.
Our approach integrates a variety of prior maps, such as OpenStreetMap's Standard Definition Maps (SD maps), outdated HD maps from vendors, and locally constructed maps from historical vehicle data. To effectively encode this prior information into online mapping models, we introduce a Hybrid Prior Representation (HPQuery) that standardizes the representation of diverse map elements.
At the core of PriorDrive is the Unified Vector Encoder (UVE), which employs a dual encoding mechanism to process vector data. The intra-vector encoder captures fine-grained local features, while the inter-vector encoder integrates global context. Furthermore, we propose a segment-level and point-level pre-training strategy that enables the UVE to learn the prior distribution of vector data, thereby improving the encoder's generalizability and performance.
Through extensive testing on the nuScenes dataset, we demonstrate that PriorDrive is highly compatible with various online mapping models and substantially improves map prediction capabilities. The integration of prior maps through the PriorDrive framework offers a robust solution to the challenges of single-perception data, paving the way for more reliable autonomous vehicle navigation.
Qualitative results with and w/o online local prior. Prior maps can help restore obscured map elements and make predictions more complete and accurate.
Performance of various online mapping models with different prior maps on nuScenes within a 60m x 30m perception range. “C” denotes camera input. The best results using the same backbone are highlighted in bold. The SD map is sourced from OpenStreetMap. The online local map refers to the historical prediction results. As the nuScenes dataset lacks existing HD maps, we followed the approach of MapEX to create an HD map-EX*, simulating existing HD maps by removing pedestrian crossings and lane dividers, while retaining only the road boundaries.
If you find our work useful in your research, please cite our paper:
@article{zeng2024driving,
title={Driving with Prior Maps: Unified Vector Prior Encoding for Autonomous Vehicle Mapping},
author={Zeng, Shuang and Chang, Xinyuan and Liu, Xinran and Pan, Zheng and Wei, Xing},
journal={arXiv preprint arXiv:2409.05352},
year={2024}
}