VLA-R: Vision-Language Action Retrieval toward Open-World End-to-End Autonomous Driving

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Seong, Hyunki, Moon, Seongwoo, Ahn, Hojin, Kang, Jehun, Shim, David Hyunchul
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866912712335818752
author Seong, Hyunki
Moon, Seongwoo
Ahn, Hojin
Kang, Jehun
Shim, David Hyunchul
author_facet Seong, Hyunki
Moon, Seongwoo
Ahn, Hojin
Kang, Jehun
Shim, David Hyunchul
contents Exploring open-world situations in an end-to-end manner is a promising yet challenging task due to the need for strong generalization capabilities. In particular, end-to-end autonomous driving in unstructured outdoor environments often encounters conditions that were unfamiliar during training. In this work, we present Vision-Language Action Retrieval (VLA-R), an open-world end-to-end autonomous driving (OW-E2EAD) framework that integrates open-world perception with a novel vision-action retrieval paradigm. We leverage a frozen vision-language model for open-world detection and segmentation to obtain multi-scale, prompt-guided, and interpretable perception features without domain-specific tuning. A Q-Former bottleneck aggregates fine-grained visual representations with language-aligned visual features, bridging perception and action domains. To learn transferable driving behaviors, we introduce a vision-action contrastive learning scheme that aligns vision-language and action embeddings for effective open-world reasoning and action retrieval. Our experiments on a real-world robotic platform demonstrate strong generalization and exploratory performance in unstructured, unseen environments, even with limited data. Demo videos are provided in the supplementary material.
format Preprint
id arxiv_https___arxiv_org_abs_2511_12405
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle VLA-R: Vision-Language Action Retrieval toward Open-World End-to-End Autonomous Driving
Seong, Hyunki
Moon, Seongwoo
Ahn, Hojin
Kang, Jehun
Shim, David Hyunchul
Computer Vision and Pattern Recognition
Exploring open-world situations in an end-to-end manner is a promising yet challenging task due to the need for strong generalization capabilities. In particular, end-to-end autonomous driving in unstructured outdoor environments often encounters conditions that were unfamiliar during training. In this work, we present Vision-Language Action Retrieval (VLA-R), an open-world end-to-end autonomous driving (OW-E2EAD) framework that integrates open-world perception with a novel vision-action retrieval paradigm. We leverage a frozen vision-language model for open-world detection and segmentation to obtain multi-scale, prompt-guided, and interpretable perception features without domain-specific tuning. A Q-Former bottleneck aggregates fine-grained visual representations with language-aligned visual features, bridging perception and action domains. To learn transferable driving behaviors, we introduce a vision-action contrastive learning scheme that aligns vision-language and action embeddings for effective open-world reasoning and action retrieval. Our experiments on a real-world robotic platform demonstrate strong generalization and exploratory performance in unstructured, unseen environments, even with limited data. Demo videos are provided in the supplementary material.
title VLA-R: Vision-Language Action Retrieval toward Open-World End-to-End Autonomous Driving
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2511.12405