Salvato in:
Dettagli Bibliografici
Autori principali: Chen, Sirui, Ye, Yufei, Cao, Zi-Ang, Lew, Jennifer, Xu, Pei, Liu, C. Karen
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:https://arxiv.org/abs/2508.03068
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866913979688812544
author Chen, Sirui
Ye, Yufei
Cao, Zi-Ang
Lew, Jennifer
Xu, Pei
Liu, C. Karen
author_facet Chen, Sirui
Ye, Yufei
Cao, Zi-Ang
Lew, Jennifer
Xu, Pei
Liu, C. Karen
contents We propose Hand-Eye Autonomous Delivery (HEAD), a framework that learns navigation, locomotion, and reaching skills for humanoids, directly from human motion and vision perception data. We take a modular approach where the high-level planner commands the target position and orientation of the hands and eyes of the humanoid, delivered by the low-level policy that controls the whole-body movements. Specifically, the low-level whole-body controller learns to track the three points (eyes, left hand, and right hand) from existing large-scale human motion capture data while high-level policy learns from human data collected by Aria glasses. Our modular approach decouples the ego-centric vision perception from physical actions, promoting efficient learning and scalability to novel scenes. We evaluate our method both in simulation and in the real-world, demonstrating humanoid's capabilities to navigate and reach in complex environments designed for humans.
format Preprint
id arxiv_https___arxiv_org_abs_2508_03068
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Hand-Eye Autonomous Delivery: Learning Humanoid Navigation, Locomotion and Reaching
Chen, Sirui
Ye, Yufei
Cao, Zi-Ang
Lew, Jennifer
Xu, Pei
Liu, C. Karen
Robotics
We propose Hand-Eye Autonomous Delivery (HEAD), a framework that learns navigation, locomotion, and reaching skills for humanoids, directly from human motion and vision perception data. We take a modular approach where the high-level planner commands the target position and orientation of the hands and eyes of the humanoid, delivered by the low-level policy that controls the whole-body movements. Specifically, the low-level whole-body controller learns to track the three points (eyes, left hand, and right hand) from existing large-scale human motion capture data while high-level policy learns from human data collected by Aria glasses. Our modular approach decouples the ego-centric vision perception from physical actions, promoting efficient learning and scalability to novel scenes. We evaluate our method both in simulation and in the real-world, demonstrating humanoid's capabilities to navigate and reach in complex environments designed for humans.
title Hand-Eye Autonomous Delivery: Learning Humanoid Navigation, Locomotion and Reaching
topic Robotics
url https://arxiv.org/abs/2508.03068