.png){kind=icon}
.png){kind=icon}
Return
MapFM: Foundation Model-Driven HD Mapping with Multi-task Contextual Learning
.png){kind=icon}
PRE
AI
DOI：10.1007/978-3-032-08462-0_3.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
In autonomous driving, high-definition (HD) maps and semantic maps in birds-eye view (BEV) are essential for accurate localization, planning, and decision-making. This paper introduces an enhanced End-to-End model named MapFM for online vectorized HD map generation. We show significantly boost feature representation quality by incorporating powerful foundation model for encoding camera images. To further enrich the models understanding of the environment and improve prediction quality, we integrate auxiliary prediction heads for semantic segmentation in the BEV representation. This multi-task learning approach provides richer contextual supervision, leading to a more comprehensive scene representation and ultimately resulting in higher accuracy and improved quality of the predicted vectorized HD maps. We have significant increase (+2.7%) in mean average precision (mAP) compared to baseline on the nuScenes dataset. The source code is available at https://github.com/LIvanoff/MapFM.
Keywords:
High-Definition Map
Foundation model
Neural network
Multi-Task learning
Journal
H
IF:
0
Papers:
28
Citations:
0