Pre-training Auto-regressive Robotic Models with 4D Representations
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913843519684608 |
|---|---|
| author | Niu, Dantong Sharma, Yuvan Xue, Haoru Biamby, Giscard Zhang, Junyi Ji, Ziteng Darrell, Trevor Herzig, Roei |
| author_facet | Niu, Dantong Sharma, Yuvan Xue, Haoru Biamby, Giscard Zhang, Junyi Ji, Ziteng Darrell, Trevor Herzig, Roei |
| contents | Foundation models pre-trained on massive unlabeled datasets have revolutionized natural language and computer vision, exhibiting remarkable generalization capabilities, thus highlighting the importance of pre-training. Yet, efforts in robotics have struggled to achieve similar success, limited by either the need for costly robotic annotations or the lack of representations that effectively model the physical world. In this paper, we introduce ARM4R, an Auto-regressive Robotic Model that leverages low-level 4D Representations learned from human video data to yield a better pre-trained robotic model. Specifically, we focus on utilizing 3D point tracking representations from videos derived by lifting 2D representations into 3D space via monocular depth estimation across time. These 4D representations maintain a shared geometric structure between the points and robot state representations up to a linear transformation, enabling efficient transfer learning from human video data to low-level robotic control. Our experiments show that ARM4R can transfer efficiently from human video data to robotics and consistently improves performance on tasks across various robot environments and configurations. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_13142 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Pre-training Auto-regressive Robotic Models with 4D Representations Niu, Dantong Sharma, Yuvan Xue, Haoru Biamby, Giscard Zhang, Junyi Ji, Ziteng Darrell, Trevor Herzig, Roei Robotics Artificial Intelligence Foundation models pre-trained on massive unlabeled datasets have revolutionized natural language and computer vision, exhibiting remarkable generalization capabilities, thus highlighting the importance of pre-training. Yet, efforts in robotics have struggled to achieve similar success, limited by either the need for costly robotic annotations or the lack of representations that effectively model the physical world. In this paper, we introduce ARM4R, an Auto-regressive Robotic Model that leverages low-level 4D Representations learned from human video data to yield a better pre-trained robotic model. Specifically, we focus on utilizing 3D point tracking representations from videos derived by lifting 2D representations into 3D space via monocular depth estimation across time. These 4D representations maintain a shared geometric structure between the points and robot state representations up to a linear transformation, enabling efficient transfer learning from human video data to low-level robotic control. Our experiments show that ARM4R can transfer efficiently from human video data to robotics and consistently improves performance on tasks across various robot environments and configurations. |
| title | Pre-training Auto-regressive Robotic Models with 4D Representations |
| topic | Robotics Artificial Intelligence |
| url | https://arxiv.org/abs/2502.13142 |