Learning to Play Mini-Golf from Human Demonstration using Autonomous Dynamical Systems

We present a new formulation for autonomous (i.e. time-independent) Dynamical Systems (DS) to perform discrete robot motions with em non-zero velocity at a given target. The proposed model ensures the convergence of all trajectories to the target, and is inherently robust to perturbations. We evaluated the performance of our proposed method to control right-handed swings in mini-golf.


Presented at:
International Conference on Machine Learning (ICML), Seattle, USA, June 28 - July 2
Year:
2011
Laboratories:




 Record created 2011-12-06, last modified 2018-03-17

n/a:
Download fulltext
PDF

Rate this document:

Rate this document:
1
2
3
 
(Not yet reviewed)