aiqbench analysis
WB-WAM advances humanoid robotics by integrating heterogeneous whole-body motion data into generative video pre-training, effectively bridging the gap between limited robot demonstrations and complex loco-manipulation tasks. This approach leverages human motion transfer and mid-training techniques to achieve high success rates in both simulation and real-world environments while reducing data dependency.
WB-WAM: Heterogeneous Body-Hand Pre-training for Humanoid Loco-Manipulation
arXiv:2609.34199v1 Announce Type: new Abstract: Humanoid loco-manipulation demands coordinated body and hand behavior, while conventional robot pre-training data provide limited coverage of such whole-body motion. We present WB-WAM, a World Action Model that incorporates explicit whole-body action supervision into generative video pre-training. A shared physical action space integrates body, root, and dexterous hand annotations from heterogeneous sources, enabling joint video and action learning from 1880.2 hours of partially annotated video and motion data. The resulting priors are refined through PICO mid-training and adapted to robot tasks with auxiliary forward kinematics supervision. We construct WB-Datasets to support these stages with retargeted egocentric human demonstrations and robot trajectories, allowing task-aligned human motion to supplement limited robot data. Evaluations in simulation demonstrate strong whole-body task performance with 81.9% in HumanoidArena, while real-world experiments further validate WB-WAM with 84.0% mean success across five tasks. Moreover, task-aligned PICO mid-training improves downstream task performance while reducing the need for real-robot demonstrations. These results support heterogeneous whole-body pre-training and human motion transfer as a practical route to data-efficient humanoid loco-manipulation.
Abstract:Humanoid loco-manipulation demands coordinated body and hand behavior, while conventional robot pre-training data provide limited coverage of such whole-body motion. We present WB-WAM, a World Action Model that incorporates explicit whole-body action supervision into generative video pre-training. A shared physical action space integrates body, root, and dexterous hand annotations from heterogeneous sources, enabling joint video and action learning from 1880.2 hours of partially annotated video and motion data. The resulting priors are refined through PICO mid-training and adapted to robot tasks with auxiliary forward kinematics supervision. We construct WB-Datasets to support these stages with retargeted egocentric human demonstrations and robot trajectories, allowing task-aligned human motion to supplement limited robot data. Evaluations in simulation demonstrate strong whole-body task performance with 81.9% in HumanoidArena, while real-world experiments further validate WB-WAM with 84.0% mean success across five tasks. Moreover, task-aligned PICO mid-training improves downstream task performance while reducing the need for real-robot demonstrations. These results support heterogeneous whole-body pre-training and human motion transfer as a practical route to data-efficient humanoid loco-manipulation.
Robotics (cs.RO)
arXiv:2609.34199 [cs.RO]
(or arXiv:2609.34199v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2609.34199 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 261 words
Related coverage
InterMimicGen: Scaling Humanoid Loco-Manipulation through Self-Evolving Motion Imitation
arXiv · Oct 6, 2026
Humanoid Loco-Manipulation With Discrete VLA Model
arXiv · Sep 29, 2026
Towards a General Humanoid Loco-Manipulation Model via Egocentric Whole-Body Human Data Pretraining
arXiv · Oct 2, 2026
WholeBodyWAM: Generalizing Pre-trained World-Action Priors to Humanoid Loco-Manipulation via WBC-Grounded Coordination
arXiv · Sep 16, 2026