Mapping Roads with the Lane Model Transformer Network

September 23, 2024

Ph.D. student Thomas Monninger will present LMT-Net: The Lane Model Transformer at the 27th IEEE International Conference on Intelligent Transportation Systems (ITSC 2024), in Edmonton, Canada, this week.

High Definition (HD) maps may help autonomous vehicles to safely navigate the road network. Because of changes and constructions in the road network, a reliable map must be updated frequently involving enormous human annotation efforts. The Lane Model Transformer Network automates this task based on sparse observations of vehicles.  Aligned and aggregated observations are used by the transformer to predicts lane pairs and their connectivity. Our evaluation shows promising results and demonstrates superior performance compared to the implemented baseline on both highway and non-highway Operational Design Domain (ODD).
 
Michael Mink, Thomas Monninger, Steffen Staab.    LMT-Net: Lane Model Transformer Network for Automated HD Mapping from Sparse Vehicle Observations Conference: 27th IEEE International Conference on Intelligent Transportation Systems (ITSC), September 24- 27, 2024 in Edmonton, Canada. https://arxiv.org/abs/2409.12409 
To the top of the page