AtomEgo: Exploring Ego-Robot Integration for Embodied Foundation Model Pretraining

原文：AtomEgo: Exploring Ego-Robot Integration for Embodied Foundation Model Pretraining

作者：Di Wu, Dongchen Zheng, Junhe Sheng, Zhongxing Wei, Songxin Zhang, Zejian Xie, Xiaoquan Sun, Junyang Zheng, Zhuoyang Song, Jiaxing Zhang, Jiayu Chen

来源：arXiv cs.RO（机器人）

正文

Computer Science > Robotics

arXiv:2609.21461v1 (cs)

[Submitted on 18 Sep 2026]

Title:AtomEgo: Exploring Ego-Robot Integration for Embodied Foundation Model Pretraining

Authors:Di Wu, Dongchen Zheng, Junhe Sheng, Zhongxing Wei, Songxin Zhang, Zejian Xie, Xiaoquan Sun, Junyang Zheng, Zhuoyang Song, Jiaxing Zhang, Jiayu Chen

View a PDF of the paper titled AtomEgo: Exploring Ego-Robot Integration for Embodied Foundation Model Pretraining, by Di Wu and 10 other authors

View PDF

HTML (experimental)

Abstract:Embodied foundation models are constrained by the limited scale and diversity of robot demonstrations, motivating the use of large-scale egocentric human interaction data. However, how to effectively incorporate such data into embodied-model pre-training remains unclear because of substantial embodiment and action-space gaps between humans and robots. We present AtomEgo, a systematic study of ego–robot co-training supported by a curated corpus of approximately 2,659 hours and a scalable data processing pipeline. Across vision–language–action and world–action model architectures, we investigate three representative paradigms: joint co-training with domain-specific action heads, progressive ego-to-robot transfer through embodiment alignment, and joint video–action modeling. We evaluate these paradigms through multi-task real-robot experiments and language-conditioned cross-embodiment representation analysis. Our results reveal a simple principle: Data Scale * Alignment Quality –> Capability Gain; egocentric data can improve generalization, but their value depends on how effectively they are aligned and utilized. This principle can provide practical guidance for scalable ego–robot pre-training.

Subjects:

Robotics (cs.RO); Artificial Intelligence (cs.AI)

Cite as:

arXiv:2609.21461 [cs.RO]

(or

arXiv:2609.21461v1 [cs.RO] for this version)

https://doi.org/10.48550/arXiv.2609.21461

Focus to learn more

arXiv-issued DOI via DataCite (pending registration)

主题

机器人 · 人工智能

由「前沿雷达」于 2026-09-18 采集。正文取自原文页面，已保留出处链接。标题与正文版权归原作者所有。