1. [Publications](/publications)
2. Augmenting Lane Perception and Topology Understanding with Standard Definition Navigation Maps
 
 # Augmenting Lane Perception and Topology Understanding with Standard Definition Navigation Maps

  ![Publication image](/sites/default/files/styles/wide/public/default_images/default.jpeg?itok=qUFsuJCP "Publication image")

 Autonomous driving has traditionally relied heavily on costly and labor-intensive High Definition (HD) maps, hindering scalability. In contrast, Standard Definition (SD) maps are more affordable and have worldwide coverage, offering a scalable alternative. In this work, we systematically explore the effect of SD maps for real-time lane-topology understanding. We propose a novel framework to integrate SD maps into online map prediction and propose a Transformer-based encoder, SD Map Encoder Representations from transFormers, to leverage priors in SD maps for the lane-topology prediction task. This enhancement consistently and significantly boosts (by up to 60%) lane detection and topology prediction on current state-of-the-art online map prediction methods without bells and whistles and can be immediately incorporated into any Transformer-based lane-topology method. Code is available at [this https URL](https://github.com/NVlabs/SMERF).



 ## Authors



Katie Z Luo (Cornell University)

[Xinshuo Weng](/person/xinshuo-weng)

[Yan Wang](/person/yan-wang)

Shuang Wu (NVIDIA)

Jie Li (NVIDIA)

Kilian Q Weinberger (Cornell University)

[Yue Wang](/person/yue-wang)

[Marco Pavone](/person/marco-pavone)

 

 

 ## Publication Date



Tuesday, November 7, 2023

 

 ## Published in



[Arxiv](https://arxiv.org/abs/2311.04079v1)

 

 ## Research Area



[Autonomous Vehicles](/research-area/autonomous-vehicles)