arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1901.00943cs.LGcs.AIcs.NEcs.RO
收藏

连续控制图像嵌入的自监督学习

Self-supervised Learning of Image Embedding for Continuous Control

University of California, Berkeley(加州大学伯克利分校);DeepMind(深度思维)

Carlos Florensa, Jonas Degrave, Nicolas Heess, Jost Tobias Springenberg, Martin Riedmiller

摘要:

直接从原始的高维感官输入(如图像)进行操作仍然是机器人控制的一个挑战。最近,人们提出了强化学习方法来解决端到端的特定任务,从像素到扭矩。然而,这些方法假设访问指定的奖励,这可能需要专门的环境仪器。此外,所获得的策略和表示往往是特定于任务的,并且可能不能很好地传递。在这项工作中,我们调查完全自监督学习的一般图像嵌入和控制原语,找到最短的时间达到任何状态的基础上。我们还引入了一个新的结构的状态-动作价值函数,建立了无模型和基于模型的方法之间的连接,并提高了性能的学习算法。我们的实验证明这些发现在三个模拟机器人任务。

英文摘要:

Operating directly from raw high dimensional sensory inputs like images is still a challenge for robotic control. Recently, Reinforcement Learning methods have been proposed to solve specific tasks end-to-end, from pixels to torques. However, these approaches assume the access to a specified reward which may require specialized instrumentation of the environment. Furthermore, the obtained policy and representations tend to be task specific and may not transfer well. In this work we investigate completely self-supervised learning of a general image embedding and control primitives, based on finding the shortest time to reach any state. We also introduce a new structure for the state-action value function that builds a connection between model-free and model-based methods, and improves the performance of the learning algorithm. We experimentally demonstrate these findings in three simulated robotic tasks.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑