Reinforcement Learning of Perceptual Coupling for Motor Primitives

Kober, J; Peters, J

Local TagsRelease HistoryDetailsSummary

Reinforcement Learning of Perceptual Coupling for Motor Primitives

Kober, J., & Peters, J. (2008). Reinforcement Learning of Perceptual Coupling for Motor Primitives. Poster presented at 8th European Workshop on Reinforcement Learning for Robotics (EWRL 2008), Villeneuve d’Ascq, France.

Item is Released

show all hide all

Basic

show hide

Item Permalink: https://hdl.handle.net/11858/00-001M-0000-0013-C885-A Version Permalink: https://hdl.handle.net/21.11116/0000-0003-90FF-B

Genre: Poster

Files

show Files

Locators

show

hide

Locator:
https://ewrl.wordpress.com/past-ewrl/ewrl8-2008/ (Table of contents) Open Access status unknown

Description:
-

OA-Status:

Creators

show

hide

Creators:
Kober, J^{1, 2}, Author
Peters, J^{1, 2}, Author

Affiliations:
1Department Empirical Inference, Max Planck Institute for Biological Cybernetics, Max Planck Society, ou_1497795
2Max Planck Institute for Biological Cybernetics, Max Planck Society, Spemannstrasse 38, 72076 Tübingen, DE, ou_1497794

Content

show

hide

Free keywords: -

Abstract: Reinforcement learning is a natural choice for the learning of complex motor tasks by reward-related self-improvement. As the space of movements is high-dimensional and
continuous, a policy parametrization is needed which can be used in this context. Traditional motor primitive approaches deal largely with open-loop policies which can
only deal with small perturbations. In this paper, we present a new type of motor primitive policies which serve as closed-loop policies together with an appropriate learning
algorithm. Our new motor primitives are an augmented version version of the dynamic systems motor primitives that incorporates perceptual coupling to external
variables. We show that these motor primitives can perform complex tasks such a Ball-in-a-Cup or Kendama task even with large variances in the initial conditions where a
human would hardly be able to learn this task. We initialize the open-loop policies by imitation learning and the perceptual coupling with a handcrafted solution. We first
improve the open-loop policies and subsequently the perceptual coupling using a novel reinforcement learning method which is particularly well-suited for motor primitives.

Details

show

hide

Language(s):

Dates: Published Online: 2008-07

Publication Status: Published online

Pages: -

Publishing info: -

Table of Contents: -

Rev. Type: -

Identifiers: BibTex Citekey: 5676

Degree: -

Event

show

hide

Title: 8th European Workshop on Reinforcement Learning for Robotics (EWRL 2008)

Place of Event: Villeneuve d’Ascq, France

Start-/End Date: 2008-06-30 - 2008-07-03

Legal Case

show

Project information

show

Source

show