Pixel2Catch: Multi-Agent Sim-to-Real Transfer for Agile Manipulation with a Single RGB Camera
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866918357309063168 |
|---|---|
| author | Kim, Seongyong Cho, Junhyeon Lee, Kang-Won Lim, Soo-Chul |
| author_facet | Kim, Seongyong Cho, Junhyeon Lee, Kang-Won Lim, Soo-Chul |
| contents | To catch a thrown object, a robot must be able to perceive the object's motion and generate control actions in a timely manner. Rather than explicitly estimating the object's 3D position, this work focuses on a novel approach that recognizes object motion using pixel-level visual information extracted from a single RGB image. Such visual cues capture changes in the object's position and scale, allowing the policy to reason about the object's motion. Furthermore, to achieve stable learning in a high-DoF system composed of a robot arm equipped with a multi-fingered hand, we design a heterogeneous multi-agent reinforcement learning framework that defines the arm and hand as independent agents with distinct roles. Each agent is trained cooperatively using role-specific observations and rewards, and the learned policies are successfully transferred from simulation to the real world. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_22733 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Pixel2Catch: Multi-Agent Sim-to-Real Transfer for Agile Manipulation with a Single RGB Camera Kim, Seongyong Cho, Junhyeon Lee, Kang-Won Lim, Soo-Chul Robotics To catch a thrown object, a robot must be able to perceive the object's motion and generate control actions in a timely manner. Rather than explicitly estimating the object's 3D position, this work focuses on a novel approach that recognizes object motion using pixel-level visual information extracted from a single RGB image. Such visual cues capture changes in the object's position and scale, allowing the policy to reason about the object's motion. Furthermore, to achieve stable learning in a high-DoF system composed of a robot arm equipped with a multi-fingered hand, we design a heterogeneous multi-agent reinforcement learning framework that defines the arm and hand as independent agents with distinct roles. Each agent is trained cooperatively using role-specific observations and rewards, and the learned policies are successfully transferred from simulation to the real world. |
| title | Pixel2Catch: Multi-Agent Sim-to-Real Transfer for Agile Manipulation with a Single RGB Camera |
| topic | Robotics |
| url | https://arxiv.org/abs/2602.22733 |