Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Hsu, Cheng-Chun, Wen, Bowen, Xu, Jie, Narang, Yashraj, Wang, Xiaolong, Zhu, Yuke, Biswas, Joydeep, Birchfield, Stan
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2411.00965
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866913834947575808
author Hsu, Cheng-Chun
Wen, Bowen
Xu, Jie
Narang, Yashraj
Wang, Xiaolong
Zhu, Yuke
Biswas, Joydeep
Birchfield, Stan
author_facet Hsu, Cheng-Chun
Wen, Bowen
Xu, Jie
Narang, Yashraj
Wang, Xiaolong
Zhu, Yuke
Biswas, Joydeep
Birchfield, Stan
contents We introduce SPOT, an object-centric imitation learning framework. The key idea is to capture each task by an object-centric representation, specifically the SE(3) object pose trajectory relative to the target. This approach decouples embodiment actions from sensory inputs, facilitating learning from various demonstration types, including both action-based and action-less human hand demonstrations, as well as cross-embodiment generalization. Additionally, object pose trajectories inherently capture planning constraints from demonstrations without the need for manually-crafted rules. To guide the robot in executing the task, the object trajectory is used to condition a diffusion policy. We systematically evaluate our method on simulation and real-world tasks. In real-world evaluation, using only eight demonstrations shot on an iPhone, our approach completed all tasks while fully complying with task constraints. Project page: https://nvlabs.github.io/object_centric_diffusion
format Preprint
id arxiv_https___arxiv_org_abs_2411_00965
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle SPOT: SE(3) Pose Trajectory Diffusion for Object-Centric Manipulation
Hsu, Cheng-Chun
Wen, Bowen
Xu, Jie
Narang, Yashraj
Wang, Xiaolong
Zhu, Yuke
Biswas, Joydeep
Birchfield, Stan
Robotics
We introduce SPOT, an object-centric imitation learning framework. The key idea is to capture each task by an object-centric representation, specifically the SE(3) object pose trajectory relative to the target. This approach decouples embodiment actions from sensory inputs, facilitating learning from various demonstration types, including both action-based and action-less human hand demonstrations, as well as cross-embodiment generalization. Additionally, object pose trajectories inherently capture planning constraints from demonstrations without the need for manually-crafted rules. To guide the robot in executing the task, the object trajectory is used to condition a diffusion policy. We systematically evaluate our method on simulation and real-world tasks. In real-world evaluation, using only eight demonstrations shot on an iPhone, our approach completed all tasks while fully complying with task constraints. Project page: https://nvlabs.github.io/object_centric_diffusion
title SPOT: SE(3) Pose Trajectory Diffusion for Object-Centric Manipulation
topic Robotics
url https://arxiv.org/abs/2411.00965