Learning Affordances from Interactive Exploration using an Object-level Map
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_ | 1866913643718770688 |
|---|---|
| author | Wulkop, Paula Özdemir, Halil Umut Hüfner, Antonia Chung, Jen Jen Siegwart, Roland Ott, Lionel |
| author_facet | Wulkop, Paula Özdemir, Halil Umut Hüfner, Antonia Chung, Jen Jen Siegwart, Roland Ott, Lionel |
| contents | Many robotic tasks in real-world environments require physical interactions with an object such as pick up or push. For successful interactions, the robot needs to know the object's affordances, which are defined as the potential actions the robot can perform with the object. In order to learn a robot-specific affordance predictor, we propose an interactive exploration pipeline which allows the robot to collect interaction experiences while exploring an unknown environment. We integrate an object-level map in the exploration pipeline such that the robot can identify different object instances and track objects across diverse viewpoints. This results in denser and more accurate affordance annotations compared to state-of-the-art methods, which do not incorporate a map. We show that our affordance exploration approach makes exploration more efficient and results in more accurate affordance prediction models compared to baseline methods. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2501_06047 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Learning Affordances from Interactive Exploration using an Object-level Map Wulkop, Paula Özdemir, Halil Umut Hüfner, Antonia Chung, Jen Jen Siegwart, Roland Ott, Lionel Robotics Many robotic tasks in real-world environments require physical interactions with an object such as pick up or push. For successful interactions, the robot needs to know the object's affordances, which are defined as the potential actions the robot can perform with the object. In order to learn a robot-specific affordance predictor, we propose an interactive exploration pipeline which allows the robot to collect interaction experiences while exploring an unknown environment. We integrate an object-level map in the exploration pipeline such that the robot can identify different object instances and track objects across diverse viewpoints. This results in denser and more accurate affordance annotations compared to state-of-the-art methods, which do not incorporate a map. We show that our affordance exploration approach makes exploration more efficient and results in more accurate affordance prediction models compared to baseline methods. |
| title | Learning Affordances from Interactive Exploration using an Object-level Map |
| topic | Robotics |
| url | https://arxiv.org/abs/2501.06047 |