الملخص
Humanoid whole-body manipulation requires coordinated whole-body dynamics, yet large-scale trajectories from a target robot are expensive to collect and difficult to scale. In contrast, whole-body motion from human and humanoid sources is abundantly available, although such data cannot be directly used as embodiment-specific robot actions. This work asks whether these scalable motion resources can instead provide a transferable predictive prior for humanoid world-action modeling. We introduce WholeBodyWAM, a humanoid world-action model that learns whole-body dynamics from large-scale heterogeneous motion before target-robot training. We curate UniMotion-4K, a motion corpus spanning more than 4K hours from human videos, native 3D motion datasets, and heterogeneous humanoid platforms, and canonicalize these diverse sources into a unified motion space. A language-conditioned Motion Expert is then pretrained to predict future whole-body motion without target-robot action supervision. During robot post-training, the pretrained Motion Expert is integrated with Video and Action Experts through asymmetric Mixture-of-Transformers (MoT) attention, enabling predictive scene dynamics and whole-body motion to jointly inform embodiment-specific action generation. Experiments show that WholeBodyWAM consistently benefits from increased motion-pretraining scale, improves future-motion prediction and downstream task performance, and transfers effectively to real-world humanoid manipulation. Moreover, the pretrained motion prior substantially improves data efficiency under limited target-robot demonstrations.
الكلمات المفتاحية
الموضوع
بيانات النشر
- المجلة
- غير متاح
- وصول مفتوح
- وصول مفتوح أخضر
اقتبس هذه المقالة
APA 7
Zhang, B., Zhang, Q., Bai, S., Wang, X., Li, M., Wang, Y., Zhang, L., Tang, J., Zhou, L., Sun, L., & Che, Z. (2026). WholeBodyWAM: Learning Whole-Body World Action Models with Scalable Motion Priors. https://omanscience.com/ar/articles/wholebodywam-learning-whole-body-world-action-models-with-scalable-motion-priors
MLA 9
Zhang, Bowei, et al. "WholeBodyWAM: Learning Whole-Body World Action Models with Scalable Motion Priors." https://omanscience.com/ar/articles/wholebodywam-learning-whole-body-world-action-models-with-scalable-motion-priors.
شيكاغو (المؤلف–التاريخ)
Zhang, Bowei, Qiyao Zhang, Shuanghao Bai, Xinhua Wang, Meng Li, Yilei Wang, Leiwang Zhang, Jian Tang, Lu Zhou, Lei Sun, and Zhengping Che. 2026. "WholeBodyWAM: Learning Whole-Body World Action Models with Scalable Motion Priors." https://omanscience.com/ar/articles/wholebodywam-learning-whole-body-world-action-models-with-scalable-motion-priors.
هارفارد
Zhang, B., Zhang, Q., Bai, S., Wang, X., Li, M., Wang, Y., Zhang, L., Tang, J., Zhou, L., Sun, L. and Che, Z. (2026) 'WholeBodyWAM: Learning Whole-Body World Action Models with Scalable Motion Priors', Available at: https://omanscience.com/ar/articles/wholebodywam-learning-whole-body-world-action-models-with-scalable-motion-priors.
فانكوفر
Zhang B, Zhang Q, Bai S, Wang X, Li M, Wang Y, et al. WholeBodyWAM: Learning Whole-Body World Action Models with Scalable Motion Priors. https://omanscience.com/ar/articles/wholebodywam-learning-whole-body-world-action-models-with-scalable-motion-priors
IEEE
B. Zhang, Q. Zhang, S. Bai, X. Wang, M. Li, Y. Wang, L. Zhang, J. Tang, L. Zhou, L. Sun, and Z. Che, "WholeBodyWAM: Learning Whole-Body World Action Models with Scalable Motion Priors," https://omanscience.com/ar/articles/wholebodywam-learning-whole-body-world-action-models-with-scalable-motion-priors.