Saved in:
Bibliographic Details
Main Authors: Gao, George Jiayuan, Li, Tianyu, Figueroa, Nadia
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2411.03294
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912486679117824
author Gao, George Jiayuan
Li, Tianyu
Figueroa, Nadia
author_facet Gao, George Jiayuan
Li, Tianyu
Figueroa, Nadia
contents We propose an object-centric recovery (OCR) framework to address the challenges of out-of-distribution (OOD) scenarios in visuomotor policy learning. Previous behavior cloning (BC) methods rely heavily on a large amount of labeled data coverage, failing in unfamiliar spatial states. Without relying on extra data collection, our approach learns a recovery policy constructed by an inverse policy inferred from the object keypoint manifold gradient in the original training data. The recovery policy serves as a simple add-on to any base visuomotor BC policy, agnostic to a specific method, guiding the system back towards the training distribution to ensure task success even in OOD situations. We demonstrate the effectiveness of our object-centric framework in both simulation and real robot experiments, achieving an improvement of 77.7\% over the base policy in OOD. Furthermore, we show OCR's capacity to autonomously collect demonstrations for continual learning. Overall, we believe this framework represents a step toward improving the robustness of visuomotor policies in real-world settings.
format Preprint
id arxiv_https___arxiv_org_abs_2411_03294
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Out-of-Distribution Recovery with Object-Centric Keypoint Inverse Policy for Visuomotor Imitation Learning
Gao, George Jiayuan
Li, Tianyu
Figueroa, Nadia
Robotics
Artificial Intelligence
We propose an object-centric recovery (OCR) framework to address the challenges of out-of-distribution (OOD) scenarios in visuomotor policy learning. Previous behavior cloning (BC) methods rely heavily on a large amount of labeled data coverage, failing in unfamiliar spatial states. Without relying on extra data collection, our approach learns a recovery policy constructed by an inverse policy inferred from the object keypoint manifold gradient in the original training data. The recovery policy serves as a simple add-on to any base visuomotor BC policy, agnostic to a specific method, guiding the system back towards the training distribution to ensure task success even in OOD situations. We demonstrate the effectiveness of our object-centric framework in both simulation and real robot experiments, achieving an improvement of 77.7\% over the base policy in OOD. Furthermore, we show OCR's capacity to autonomously collect demonstrations for continual learning. Overall, we believe this framework represents a step toward improving the robustness of visuomotor policies in real-world settings.
title Out-of-Distribution Recovery with Object-Centric Keypoint Inverse Policy for Visuomotor Imitation Learning
topic Robotics
Artificial Intelligence
url https://arxiv.org/abs/2411.03294