aiqbench analysis
This research introduces HumanVerse-500, a novel dataset leveraging egocentric human video to train $\lambda_0$, a vision-language-action policy for whole-body humanoid locomotion and manipulation. By transferring shared representations from human experience to robot control, the study addresses scalability challenges in coordinating complex body-hand interactions without relying on costly teleoperation. The work advances foundational machine learning methods for embodied AI, offering a scalable pathway to generalize diverse human behaviors to robotic embodiments.
Towards a General Humanoid Loco-Manipulation Model via Egocentric Whole-Body Human Data Pretraining
arXiv:2610.00438v1 Announce Type: new Abstract: Humanoid whole-body manipulation has advanced rapidly, enabling policies to coordinate locomotion, posture, bimanual interaction, and dexterous hand movements. Meanwhile, egocentric human videos provide diverse examples of everyday interactions across objects and scenes, offering scalable supervision without robot operation. However, existing supervision from these videos provides limited coverage of whole-body movement and coordination with hand-object interaction, while obtaining such supervision through humanoid teleoperation is also costly and difficult to scale. We therefore explore how human experience can support scalable learning of humanoid loco-manipulation. To support this study, we introduce HumanVerse-500, a 500-hour dataset of diverse human loco-manipulation behaviors in open-world environments, collected with a lightweight wearable system that synchronizes egocentric video with body and hand motion. Building on this dataset, we develop $\lambda_0$, a whole-body humanoid vision-language-action policy, through three-stage training that first learns interaction from diverse egocentric datasets, then coordinates body and hand motion using HumanVerse-500, and finally adapts the policy to downstream tasks and robot embodiments. Across these stages, $\lambda_0$ learns a shared representation space for human experience transfer, while domain-specific interfaces handle differences between human and robot states and actions. We evaluate $\lambda_0$ on SIMPLE and 4 real-world loco-manipulation tasks, achieving state-of-the-art performance, and further analyze its scaling behavior, generalization, and training-stage contributions to understand how human data support downstream whole-body humanoid control. We will release our code, models, and data to support further research.
Authors:Chongyang Xu, Zhao Wu, Jin Chen, Yiming Jiang, Jinhui Ye, Yuming Jiang, Shifeng Zhang, Ziliang Feng, Mu Xu, Yilun Chen, Li Lu, Steven C.H. Hoi
Abstract:Humanoid whole-body manipulation has advanced rapidly, enabling policies to coordinate locomotion, posture, bimanual interaction, and dexterous hand movements. Meanwhile, egocentric human videos provide diverse examples of everyday interactions across objects and scenes, offering scalable supervision without robot operation. However, existing supervision from these videos provides limited coverage of whole-body movement and coordination with hand-object interaction, while obtaining such supervision through humanoid teleoperation is also costly and difficult to scale. We therefore explore how human experience can support scalable learning of humanoid loco-manipulation. To support this study, we introduce HumanVerse-500, a 500-hour dataset of diverse human loco-manipulation behaviors in open-world environments, collected with a lightweight wearable system that synchronizes egocentric video with body and hand motion. Building on this dataset, we develop $\lambda_0$, a whole-body humanoid vision-language-action policy, through three-stage training that first learns interaction from diverse egocentric datasets, then coordinates body and hand motion using HumanVerse-500, and finally adapts the policy to downstream tasks and robot embodiments. Across these stages, $\lambda_0$ learns a shared representation space for human experience transfer, while domain-specific interfaces handle differences between human and robot states and actions. We evaluate $\lambda_0$ on SIMPLE and 4 real-world loco-manipulation tasks, achieving state-of-the-art performance, and further analyze its scaling behavior, generalization, and training-stage contributions to understand how human data support downstream whole-body humanoid control. We will release our code, models, and data to support further research.
Robotics (cs.RO)
arXiv:2610.00438 [cs.RO]
(or arXiv:2610.00438v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2610.00438 arXiv-issued DOI via DataCite (pending registration)
Full text under the publisher's open licence · 269 words
Related coverage
Dyna-2: General Purpose Physical Agents with Hour-Long Autonomy — Jason Ma, Dyna Robotics
Latent Space · Oct 13, 2026
$\lambda_0$: Scaling Open-World Egocentric Human Data for Humanoid Loco-Manipulation
arXiv · Oct 15, 2026
MotionWAM: Towards Foundation World Action Models for Real-Time Humanoid Loco-Manipulation
arXiv · Jun 9, 2026
WholeBodyWAM: Generalizing Pre-trained World-Action Priors to Humanoid Loco-Manipulation via WBC-Grounded Coordination
arXiv · Sep 16, 2026