Constrained Stein Variational Gradient Descent for Robot Perception, Planning, and Identification
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866912406241804288 |
|---|---|
| author | Tabor, Griffin Hermans, Tucker |
| author_facet | Tabor, Griffin Hermans, Tucker |
| contents | Many core problems in robotics can be framed as constrained optimization problems. Often on these problems, the robotic system has uncertainty, or it would be advantageous to identify multiple high quality feasible solutions. To enable this, we present two novel frameworks for applying principles of constrained optimization to the new variational inference algorithm Stein variational gradient descent. Our general framework supports multiple types of constrained optimizers and can handle arbitrary constraints. We demonstrate on a variety of problems that we are able to learn to approximate distributions without violating constraints. Specifically, we show that we can build distributions of: robot motion plans that exactly avoid collisions, robot arm joint angles on the SE(3) manifold with exact table placement constraints, and object poses from point clouds with table placement constraints. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_00589 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Constrained Stein Variational Gradient Descent for Robot Perception, Planning, and Identification Tabor, Griffin Hermans, Tucker Robotics Machine Learning Many core problems in robotics can be framed as constrained optimization problems. Often on these problems, the robotic system has uncertainty, or it would be advantageous to identify multiple high quality feasible solutions. To enable this, we present two novel frameworks for applying principles of constrained optimization to the new variational inference algorithm Stein variational gradient descent. Our general framework supports multiple types of constrained optimizers and can handle arbitrary constraints. We demonstrate on a variety of problems that we are able to learn to approximate distributions without violating constraints. Specifically, we show that we can build distributions of: robot motion plans that exactly avoid collisions, robot arm joint angles on the SE(3) manifold with exact table placement constraints, and object poses from point clouds with table placement constraints. |
| title | Constrained Stein Variational Gradient Descent for Robot Perception, Planning, and Identification |
| topic | Robotics Machine Learning |
| url | https://arxiv.org/abs/2506.00589 |