Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2505.23171 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917185091272704 |
|---|---|
| author | Liu, Liu Wang, Xiaofeng Zhao, Guosheng Li, Keyu Qin, Wenkang Zhu, Jiagang Qiu, Jiaxiong Zhu, Zheng Huang, Guan Su, Zhizhong |
| author_facet | Liu, Liu Wang, Xiaofeng Zhao, Guosheng Li, Keyu Qin, Wenkang Zhu, Jiagang Qiu, Jiaxiong Zhu, Zheng Huang, Guan Su, Zhizhong |
| contents | The goal of general-purpose robotics is to create agents that can seamlessly adapt to and operate in diverse, unstructured human environments. Imitation learning has become a key paradigm for robotic manipulation, yet collecting large-scale and diverse demonstrations is prohibitively expensive. Simulators provide a cost-effective alternative, but the sim-to-real gap remains a major obstacle to scalability. We present RoboTransfer, a diffusion-based video generation framework for synthesizing robotic data. By leveraging cross-view feature interactions and globally consistent 3D geometry, RoboTransfer ensures multi-view geometric consistency while enabling fine-grained control over scene elements, such as background editing and object replacement. Extensive experiments demonstrate that RoboTransfer produces videos with superior geometric consistency and visual fidelity. Furthermore, policies trained on this synthetic data exhibit enhanced generalization to novel, unseen scenarios. Project page: https://horizonrobotics.github.io/robot_lab/robotransfer. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_23171 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | RoboTransfer: Controllable Geometry-Consistent Video Diffusion for Manipulation Policy Transfer Liu, Liu Wang, Xiaofeng Zhao, Guosheng Li, Keyu Qin, Wenkang Zhu, Jiagang Qiu, Jiaxiong Zhu, Zheng Huang, Guan Su, Zhizhong Computer Vision and Pattern Recognition The goal of general-purpose robotics is to create agents that can seamlessly adapt to and operate in diverse, unstructured human environments. Imitation learning has become a key paradigm for robotic manipulation, yet collecting large-scale and diverse demonstrations is prohibitively expensive. Simulators provide a cost-effective alternative, but the sim-to-real gap remains a major obstacle to scalability. We present RoboTransfer, a diffusion-based video generation framework for synthesizing robotic data. By leveraging cross-view feature interactions and globally consistent 3D geometry, RoboTransfer ensures multi-view geometric consistency while enabling fine-grained control over scene elements, such as background editing and object replacement. Extensive experiments demonstrate that RoboTransfer produces videos with superior geometric consistency and visual fidelity. Furthermore, policies trained on this synthetic data exhibit enhanced generalization to novel, unseen scenarios. Project page: https://horizonrobotics.github.io/robot_lab/robotransfer. |
| title | RoboTransfer: Controllable Geometry-Consistent Video Diffusion for Manipulation Policy Transfer |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2505.23171 |