Learning Human-to-Humanoid Real-Time Whole-Body Teleoperation

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: He, Tairan, Luo, Zhengyi, Xiao, Wenli, Zhang, Chong, Kitani, Kris, Liu, Changliu, Shi, Guanya
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866917607451394048
author He, Tairan
Luo, Zhengyi
Xiao, Wenli
Zhang, Chong
Kitani, Kris
Liu, Changliu
Shi, Guanya
author_facet He, Tairan
Luo, Zhengyi
Xiao, Wenli
Zhang, Chong
Kitani, Kris
Liu, Changliu
Shi, Guanya
contents We present Human to Humanoid (H2O), a reinforcement learning (RL) based framework that enables real-time whole-body teleoperation of a full-sized humanoid robot with only an RGB camera. To create a large-scale retargeted motion dataset of human movements for humanoid robots, we propose a scalable "sim-to-data" process to filter and pick feasible motions using a privileged motion imitator. Afterwards, we train a robust real-time humanoid motion imitator in simulation using these refined motions and transfer it to the real humanoid robot in a zero-shot manner. We successfully achieve teleoperation of dynamic whole-body motions in real-world scenarios, including walking, back jumping, kicking, turning, waving, pushing, boxing, etc. To the best of our knowledge, this is the first demonstration to achieve learning-based real-time whole-body humanoid teleoperation.
format Preprint
id arxiv_https___arxiv_org_abs_2403_04436
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Learning Human-to-Humanoid Real-Time Whole-Body Teleoperation
He, Tairan
Luo, Zhengyi
Xiao, Wenli
Zhang, Chong
Kitani, Kris
Liu, Changliu
Shi, Guanya
Robotics
Artificial Intelligence
Machine Learning
Systems and Control
We present Human to Humanoid (H2O), a reinforcement learning (RL) based framework that enables real-time whole-body teleoperation of a full-sized humanoid robot with only an RGB camera. To create a large-scale retargeted motion dataset of human movements for humanoid robots, we propose a scalable "sim-to-data" process to filter and pick feasible motions using a privileged motion imitator. Afterwards, we train a robust real-time humanoid motion imitator in simulation using these refined motions and transfer it to the real humanoid robot in a zero-shot manner. We successfully achieve teleoperation of dynamic whole-body motions in real-world scenarios, including walking, back jumping, kicking, turning, waving, pushing, boxing, etc. To the best of our knowledge, this is the first demonstration to achieve learning-based real-time whole-body humanoid teleoperation.
title Learning Human-to-Humanoid Real-Time Whole-Body Teleoperation
topic Robotics
Artificial Intelligence
Machine Learning
Systems and Control
url https://arxiv.org/abs/2403.04436