Robot Learning by Demonstration with local Gaussian process regression
Authors:
- 1 October 2010
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 6 (21530858) , 255-260
Abstract
In recent years there was a tremendous progress in robotic systems, and however also increased expectations: A robot should be easy to program and reliable in task execution. Learning from Demonstration (LfD) offers a very promising alternative to classical engineering approaches. LfD is a very natural way for humans to interact with robots and will be an essential part of future service robots. In this work we first review heteroscedastic Gaussian processes and show how these can be used to encode a task. We then introduce a new Gaussian process regression model that clusters the input space into smaller subsets similar to the work in [11]. In the next step we show how these approaches fit into the Learning by Demonstration framework of [2], [3]. At the end we present an experiment on a real robot arm that shows how all these approaches interact.Keywords
This publication has 10 references indexed in Scilit:
- A survey of robot learning from demonstrationRobotics and Autonomous Systems, 2009
- Local Gaussian process regression for real-time model-based robot control2008
- Springer Handbook of Robotics2008
- Most likely heteroscedastic Gaussian process regression2007
- Dogged Learning for RobotsProceedings 2001 ICRA. IEEE International Conference on Robotics and Automation (cat. No.01ch37164), 2007
- On Learning, Representing, and Generalizing a Task in a Humanoid RobotIEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics), 2007
- Dynamic Imitation in a Humanoid Robot through Nonparametric Probabilistic Inference2006
- Robotic imitation for human motion capture using gaussian processes2006
- Sparse On-Line Gaussian ProcessesNeural Computation, 2002
- Numerical Optimization1999