6-DoF Grasp Pose Evaluation and Optimization via Transfer Learning from NeRFs
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866914642280841216 |
|---|---|
| author | Sóti, Gergely Huang, Xi Wurll, Christian Hein, Björn |
| author_facet | Sóti, Gergely Huang, Xi Wurll, Christian Hein, Björn |
| contents | We address the problem of robotic grasping of known and unknown objects using implicit behavior cloning. We train a grasp evaluation model from a small number of demonstrations that outputs higher values for grasp candidates that are more likely to succeed in grasping. This evaluation model serves as an objective function, that we maximize to identify successful grasps. Key to our approach is the utilization of learned implicit representations of visual and geometric features derived from a pre-trained NeRF. Though trained exclusively in a simulated environment with simplified objects and 4-DoF top-down grasps, our evaluation model and optimization procedure demonstrate generalization to 6-DoF grasps and novel objects both in simulation and in real-world settings, without the need for additional data. Supplementary material is available at: https://gergely-soti.github.io/grasp |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2401_07935 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | 6-DoF Grasp Pose Evaluation and Optimization via Transfer Learning from NeRFs Sóti, Gergely Huang, Xi Wurll, Christian Hein, Björn Robotics We address the problem of robotic grasping of known and unknown objects using implicit behavior cloning. We train a grasp evaluation model from a small number of demonstrations that outputs higher values for grasp candidates that are more likely to succeed in grasping. This evaluation model serves as an objective function, that we maximize to identify successful grasps. Key to our approach is the utilization of learned implicit representations of visual and geometric features derived from a pre-trained NeRF. Though trained exclusively in a simulated environment with simplified objects and 4-DoF top-down grasps, our evaluation model and optimization procedure demonstrate generalization to 6-DoF grasps and novel objects both in simulation and in real-world settings, without the need for additional data. Supplementary material is available at: https://gergely-soti.github.io/grasp |
| title | 6-DoF Grasp Pose Evaluation and Optimization via Transfer Learning from NeRFs |
| topic | Robotics |
| url | https://arxiv.org/abs/2401.07935 |