Sanghun Jung, JoonHo Lee, Xiangyun Meng, Byron Boots, and Alexander Lambert
University of Washington
ICRA 2024
Abstract
Reliable estimation of terrain traversability is critical for the successful deployment of autonomous systems in wild, outdoor environments. Given the lack of large-scale annotated datasets for off-road navigation, strictly-supervised learning approaches remain limited in their generalization ability. To this end, we introduce a novel, image-based self-supervised learning method for traversability prediction, leveraging a state-of-the-art vision foundation model for improved out-of-distribution performance. Our method employs contrastive representation learning using both human driving data and instance-based segmentation masks during training. We show that this simple, yet effective, technique drastically outperforms recent methods in predicting traversability for both on- and off-trail driving scenarios. We compare our method with recent baselines on both a common benchmark as well as our own datasets, covering a diverse range of outdoor environments and varied terrain types. We also demonstrate the compatibility of resulting costmap predictions with a model-predictive controller. Finally, we evaluate our approach on zero- and few-shot tasks, demonstrating unprecedented performance for generalization to new environments.
| Status | Name | Date |
|---|---|---|
| ⬜️ | Code under preparation | 09/04/2025 |
| ⬜️ | Apply for release approval | - |
| ⬜️ | Code release | - |
To be updated
This research was developed with funding from the Defense Advanced Research Projects Agency (DARPA). The views, opinions and/or findings expressed are those of the author and should not be interpreted as representing the official views or policies of the Department of Defense or the U.S. Government.
@inproceedings{jung2024v,
title={V-strong: Visual self-supervised traversability learning for off-road navigation},
author={Jung, Sanghun and Lee, JoonHo and Meng, Xiangyun and Boots, Byron and Lambert, Alexander},
booktitle={2024 IEEE International Conference on Robotics and Automation (ICRA)},
pages={1766--1773},
year={2024},
organization={IEEE}
}