Overview
Research area: Computer Vision Authors: Zekun Qi, Xuchuan Chen, Dairu Liu, Chenghuai Lin, Yunrui Lian, Sikai Liang, Zhikai Zhang, Yu Guan, Jilong Wang, Wenyao Zhang, Xinqiang Yu, He Wang, Li Yi Published: 2026-06-02 arXiv: 2606.03985
English Summary
We introduce Humanoid-GPT, a GPT-style Transformer with causal attention trained on a billion-scale motion corpus for whole-body control. Unlike prior shallow MLP trackers constrained by scarce data and an agility-generalization trade-off, Humanoid-GPT is pre-trained on a 2B-frame retargeted corpus that unifies all major mocap datasets with large-scale in-house recordings. Scaling both data and model capacity yields a single generative Transformer that tracks highly dynamic behaviors while achieving unprecedented zero-shot generalization to unseen motions and control tasks. Extensive experiments and scaling analyses show that our model establishes a new performance frontier, demonstrating robust zero-shot generalization to unseen tasks while simultaneously tracking highly dynamic and complex motions.
Key Contributions (translated from the Chinese abstract)
- A GPT-style Transformer with causal attention for humanoid whole-body control, trained at billion-scale.
- Pre-training on a 2-billion-frame retargeted motion corpus that unifies all major motion capture datasets plus large-scale in-house recordings.
- Overcomes the agility-generalization trade-off that limits prior shallow MLP-based trackers.
- A single generative model capable of tracking highly dynamic behaviors and generalizing zero-shot to unseen motions and control tasks.
- Extensive experiments and scaling analyses establishing a new performance frontier for motion tracking.