Humanoid Policy ~ Human Policy
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866909824837484544 |
|---|---|
| author | Qiu, Ri-Zhao Yang, Shiqi Cheng, Xuxin Chawla, Chaitanya Li, Jialong He, Tairan Yan, Ge Yoon, David J. Hoque, Ryan Paulsen, Lars Yang, Ge Zhang, Jian Yi, Sha Shi, Guanya Wang, Xiaolong |
| author_facet | Qiu, Ri-Zhao Yang, Shiqi Cheng, Xuxin Chawla, Chaitanya Li, Jialong He, Tairan Yan, Ge Yoon, David J. Hoque, Ryan Paulsen, Lars Yang, Ge Zhang, Jian Yi, Sha Shi, Guanya Wang, Xiaolong |
| contents | Training manipulation policies for humanoid robots with diverse data enhances their robustness and generalization across tasks and platforms. However, learning solely from robot demonstrations is labor-intensive, requiring expensive tele-operated data collection which is difficult to scale. This paper investigates a more scalable data source, egocentric human demonstrations, to serve as cross-embodiment training data for robot learning. We mitigate the embodiment gap between humanoids and humans from both the data and modeling perspectives. We collect an egocentric task-oriented dataset (PH2D) that is directly aligned with humanoid manipulation demonstrations. We then train a human-humanoid behavior policy, which we term Human Action Transformer (HAT). The state-action space of HAT is unified for both humans and humanoid robots and can be differentiably retargeted to robot actions. Co-trained with smaller-scale robot data, HAT directly models humanoid robots and humans as different embodiments without additional supervision. We show that human data improves both generalization and robustness of HAT with significantly better data collection efficiency. Code and data: https://human-as-robot.github.io/ |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_13441 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Humanoid Policy ~ Human Policy Qiu, Ri-Zhao Yang, Shiqi Cheng, Xuxin Chawla, Chaitanya Li, Jialong He, Tairan Yan, Ge Yoon, David J. Hoque, Ryan Paulsen, Lars Yang, Ge Zhang, Jian Yi, Sha Shi, Guanya Wang, Xiaolong Robotics Artificial Intelligence Computer Vision and Pattern Recognition Training manipulation policies for humanoid robots with diverse data enhances their robustness and generalization across tasks and platforms. However, learning solely from robot demonstrations is labor-intensive, requiring expensive tele-operated data collection which is difficult to scale. This paper investigates a more scalable data source, egocentric human demonstrations, to serve as cross-embodiment training data for robot learning. We mitigate the embodiment gap between humanoids and humans from both the data and modeling perspectives. We collect an egocentric task-oriented dataset (PH2D) that is directly aligned with humanoid manipulation demonstrations. We then train a human-humanoid behavior policy, which we term Human Action Transformer (HAT). The state-action space of HAT is unified for both humans and humanoid robots and can be differentiably retargeted to robot actions. Co-trained with smaller-scale robot data, HAT directly models humanoid robots and humans as different embodiments without additional supervision. We show that human data improves both generalization and robustness of HAT with significantly better data collection efficiency. Code and data: https://human-as-robot.github.io/ |
| title | Humanoid Policy ~ Human Policy |
| topic | Robotics Artificial Intelligence Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2503.13441 |