Deictic Image Maps: An Abstraction For Learning Pose Invariant Manipulation Policies

Deictic Image Maps: An Abstraction For Learning Pose Invariant Manipulation Policies
复制标题

指示图像映射:学习姿势不变操作策略的抽象

DOI:
--
复制
发表时间:
2018
期刊:
AAAI Conference on Artificial Intelligence
影响因子:
--
通讯作者:
Marcus Gualtieri
Marcus Gualtieri
中科院分区:
--
文献类型:
--
作者:
Robert W. Platt;Colin Kohler;Marcus Gualtieri

文献摘要

参考文献

被引文献

相似文献

在深度强化学习在机器人技术中的应用中,我们通常希望学习姿势不变策略:对世界中对象的位置和方向的变化保持不变的策略。例如,考虑一个钉孔插入任务。如果智能体学会了将一个钉插入一个洞,我们希望该策略可以推广到不同姿势的洞。不幸的是,这是使用常规方法的挑战。本文提出了一种新的状态和动作抽象,这种抽象对姿势偏移是不变的,称为指示图像映射,可以用于深度强化学习。我们提供了广泛的条件下,最佳的抽象政策是最佳的底层系统。最后,我们表明,该方法可以帮助解决具有挑战性的机器人操作问题。
In applications of deep reinforcement learning to robotics, it is often the case that we want to learn pose invariant policies: policies that are invariant to changes in the position and orientation of objects in the world. For example, consider a pegin-hole insertion task. If the agent learns to insert a peg into one hole, we would like that policy to generalize to holes presented in different poses. Unfortunately, this is a challenge using conventional methods. This paper proposes a novel state and action abstraction that is invariant to pose shifts called deictic image maps that can be used with deep reinforcement learning. We provide broad conditions under which optimal abstract policies are optimal for the underlying system. Finally, we show that the method can help solve challenging robotic manipulation problems.
无需几何对象模型即可拾取和放置
DOI: 10.1109/icra.2018.8460553
发表时间: 2018
期刊: Proceedings of 2018 IEEE International Conference on Robotics and Automation (ICRA
影响因子: --
作者:
Gualtieri, Marcus;Pas, Andreas ten;Platt, Robert
通讯作者: Platt, Robert
DOI: --
发表时间: 2018-06
期刊: --
影响因子: --
作者:
Marcus Gualtieri;Robert W. Platt
通讯作者: Marcus Gualtieri;Robert W. Platt