Learning in ImaginationLand: Omnidirectional Policies through 3D Generative Models (OP-Gen)
Fuente:
arXiv
Salvato in:
| Autori principali: | , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866915484012642304 |
|---|---|
| author | Ren, Yifei Johns, Edward |
| author_facet | Ren, Yifei Johns, Edward |
| contents | Recent 3D generative models, which are capable of generating full object shapes from just a few images, now open up new opportunities in robotics. In this work, we show that 3D generative models can be used to augment a dataset from a single real-world demonstration, after which an omnidirectional policy can be learned within this imagined dataset. We found that this enables a robot to perform a task when initialised from states very far from those observed during the demonstration, including starting from the opposite side of the object relative to the real-world demonstration, significantly reducing the number of demonstrations required for policy learning. Through several real-world experiments across tasks such as grasping objects, opening a drawer, and placing trash into a bin, we study these omnidirectional policies by investigating the effect of various design choices on policy behaviour, and we show superior performance to recent baselines which use alternative methods for data augmentation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_06191 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Learning in ImaginationLand: Omnidirectional Policies through 3D Generative Models (OP-Gen) Ren, Yifei Johns, Edward Robotics Computer Vision and Pattern Recognition Machine Learning Recent 3D generative models, which are capable of generating full object shapes from just a few images, now open up new opportunities in robotics. In this work, we show that 3D generative models can be used to augment a dataset from a single real-world demonstration, after which an omnidirectional policy can be learned within this imagined dataset. We found that this enables a robot to perform a task when initialised from states very far from those observed during the demonstration, including starting from the opposite side of the object relative to the real-world demonstration, significantly reducing the number of demonstrations required for policy learning. Through several real-world experiments across tasks such as grasping objects, opening a drawer, and placing trash into a bin, we study these omnidirectional policies by investigating the effect of various design choices on policy behaviour, and we show superior performance to recent baselines which use alternative methods for data augmentation. |
| title | Learning in ImaginationLand: Omnidirectional Policies through 3D Generative Models (OP-Gen) |
| topic | Robotics Computer Vision and Pattern Recognition Machine Learning |
| url | https://arxiv.org/abs/2509.06191 |