Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Zhang, Arthur, Sikchi, Harshit, Zhang, Amy, Biswas, Joydeep
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2503.03921
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866909660477390848
author Zhang, Arthur
Sikchi, Harshit
Zhang, Amy
Biswas, Joydeep
author_facet Zhang, Arthur
Sikchi, Harshit
Zhang, Amy
Biswas, Joydeep
contents We introduce CREStE, a scalable learning-based mapless navigation framework to address the open-world generalization and robustness challenges of outdoor urban navigation. Key to achieving this is learning perceptual representations that generalize to open-set factors (e.g. novel semantic classes, terrains, dynamic entities) and inferring expert-aligned navigation costs from limited demonstrations. CREStE addresses both these issues, introducing 1) a visual foundation model (VFM) distillation objective for learning open-set structured bird's-eye-view perceptual representations, and 2) counterfactual inverse reinforcement learning (IRL), a novel active learning formulation that uses counterfactual trajectory demonstrations to reason about the most important cues when inferring navigation costs. We evaluate CREStE on the task of kilometer-scale mapless navigation in a variety of city, offroad, and residential environments and find that it outperforms all state-of-the-art approaches with 70% fewer human interventions, including a 2-kilometer mission in an unseen environment with just 1 intervention; showcasing its robustness and effectiveness for long-horizon mapless navigation. Videos and additional materials can be found on the project page: https://amrl.cs.utexas.edu/creste
format Preprint
id arxiv_https___arxiv_org_abs_2503_03921
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle CREStE: Scalable Mapless Navigation with Internet Scale Priors and Counterfactual Guidance
Zhang, Arthur
Sikchi, Harshit
Zhang, Amy
Biswas, Joydeep
Robotics
Artificial Intelligence
Computer Vision and Pattern Recognition
We introduce CREStE, a scalable learning-based mapless navigation framework to address the open-world generalization and robustness challenges of outdoor urban navigation. Key to achieving this is learning perceptual representations that generalize to open-set factors (e.g. novel semantic classes, terrains, dynamic entities) and inferring expert-aligned navigation costs from limited demonstrations. CREStE addresses both these issues, introducing 1) a visual foundation model (VFM) distillation objective for learning open-set structured bird's-eye-view perceptual representations, and 2) counterfactual inverse reinforcement learning (IRL), a novel active learning formulation that uses counterfactual trajectory demonstrations to reason about the most important cues when inferring navigation costs. We evaluate CREStE on the task of kilometer-scale mapless navigation in a variety of city, offroad, and residential environments and find that it outperforms all state-of-the-art approaches with 70% fewer human interventions, including a 2-kilometer mission in an unseen environment with just 1 intervention; showcasing its robustness and effectiveness for long-horizon mapless navigation. Videos and additional materials can be found on the project page: https://amrl.cs.utexas.edu/creste
title CREStE: Scalable Mapless Navigation with Internet Scale Priors and Counterfactual Guidance
topic Robotics
Artificial Intelligence
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2503.03921