SAM-E: Leveraging Visual Foundation Model with Sequence Imitation for Embodied Manipulation

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Zhang, Junjie, Bai, Chenjia, He, Haoran, Xia, Wenke, Wang, Zhigang, Zhao, Bin, Li, Xiu, Li, Xuelong
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866909213215686656
author Zhang, Junjie
Bai, Chenjia
He, Haoran
Xia, Wenke
Wang, Zhigang
Zhao, Bin
Li, Xiu
Li, Xuelong
author_facet Zhang, Junjie
Bai, Chenjia
He, Haoran
Xia, Wenke
Wang, Zhigang
Zhao, Bin
Li, Xiu
Li, Xuelong
contents Acquiring a multi-task imitation policy in 3D manipulation poses challenges in terms of scene understanding and action prediction. Current methods employ both 3D representation and multi-view 2D representation to predict the poses of the robot's end-effector. However, they still require a considerable amount of high-quality robot trajectories, and suffer from limited generalization in unseen tasks and inefficient execution in long-horizon reasoning. In this paper, we propose SAM-E, a novel architecture for robot manipulation by leveraging a vision-foundation model for generalizable scene understanding and sequence imitation for long-term action reasoning. Specifically, we adopt Segment Anything (SAM) pre-trained on a huge number of images and promptable masks as the foundation model for extracting task-relevant features, and employ parameter-efficient fine-tuning on robot data for a better understanding of embodied scenarios. To address long-horizon reasoning, we develop a novel multi-channel heatmap that enables the prediction of the action sequence in a single pass, notably enhancing execution efficiency. Experimental results from various instruction-following tasks demonstrate that SAM-E achieves superior performance with higher execution efficiency compared to the baselines, and also significantly improves generalization in few-shot adaptation to new tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2405_19586
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle SAM-E: Leveraging Visual Foundation Model with Sequence Imitation for Embodied Manipulation
Zhang, Junjie
Bai, Chenjia
He, Haoran
Xia, Wenke
Wang, Zhigang
Zhao, Bin
Li, Xiu
Li, Xuelong
Computer Vision and Pattern Recognition
Machine Learning
Robotics
Acquiring a multi-task imitation policy in 3D manipulation poses challenges in terms of scene understanding and action prediction. Current methods employ both 3D representation and multi-view 2D representation to predict the poses of the robot's end-effector. However, they still require a considerable amount of high-quality robot trajectories, and suffer from limited generalization in unseen tasks and inefficient execution in long-horizon reasoning. In this paper, we propose SAM-E, a novel architecture for robot manipulation by leveraging a vision-foundation model for generalizable scene understanding and sequence imitation for long-term action reasoning. Specifically, we adopt Segment Anything (SAM) pre-trained on a huge number of images and promptable masks as the foundation model for extracting task-relevant features, and employ parameter-efficient fine-tuning on robot data for a better understanding of embodied scenarios. To address long-horizon reasoning, we develop a novel multi-channel heatmap that enables the prediction of the action sequence in a single pass, notably enhancing execution efficiency. Experimental results from various instruction-following tasks demonstrate that SAM-E achieves superior performance with higher execution efficiency compared to the baselines, and also significantly improves generalization in few-shot adaptation to new tasks.
title SAM-E: Leveraging Visual Foundation Model with Sequence Imitation for Embodied Manipulation
topic Computer Vision and Pattern Recognition
Machine Learning
Robotics
url https://arxiv.org/abs/2405.19586