Goal Babbling Permits Direct Learning of Inverse Kinematics
- 3 August 2010
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Autonomous Mental Development
- Vol. 2 (3), 216-229
- https://doi.org/10.1109/tamd.2010.2062511
Abstract
We present an approach to learn inverse kinematics of redundant systems without prior- or expert-knowledge. The method allows for an iterative bootstrapping and refinement of the inverse kinematics estimate. The essential novelty lies in a path-based sampling approach: we generate training data along paths, which result from execution of the currently learned estimate along a desired path towards a goal. The information structure thereby induced enables an efficient detection and resolution of inconsistent samples solely from directly observable data. We derive and illustrate the exploration and learning process with a low-dimensional kinematic example that provides direct insight into the bootstrapping process. We further show that the method scales for high dimensional problems, such as the Honda humanoid robot or hyperredundant planar arms with up to 50 degrees of freedom.Keywords
This publication has 32 references indexed in Scilit:
- The robot in the crib: a developmental analysis of imitation skills in infants and robotsInfant and Child Development, 2008
- Recurrent Cerebellar Loops Simplify Adaptive Control of Redundant and Nonlinear Motor SystemsNeural Computation, 2007
- The control system for the Honda humanoid robotAge and Ageing, 2006
- Failure of Motor Learning for Large Initial ErrorsNeural Computation, 2004
- Recurrent cerebellar architecture solves the motor-error problemProceedings Of The Royal Society B-Biological Sciences, 2004
- Robot hand-eye coordination based on stereo visionIEEE Control Systems, 1995
- Neonatal finger and arm movements as determined by a social and an object contextEarly Development and Parenting, 1994
- A Self-Organizing Neural Model of Motor Equivalent Reaching and Tool Use by a Multijoint ArmJournal of Cognitive Neuroscience, 1993
- Forward models: Supervised learning with a distal teacherCognitive Science, 1992
- From Babbling to Speech: A Re-Assessment of the Continuity IssueLanguage, 1985