Dynamical Movement Primitives: Learning Attractor Models for Motor Behaviors
- A. Ijspeert, J. Nakanishi, Heiko Hoffmann, P. Pastor, S. Schaal
- PsychologyNeural Computation
- 1 February 2013
Dynamical movement primitives is presented, a line of research for modeling attractor behaviors of autonomous nonlinear dynamical systems with the help of statistical learning techniques, and its properties are evaluated in motor control and robotics.
Learning Attractor Landscapes for Learning Motor Primitives
By nonlinearly transforming the canonical attractor dynamics using techniques from nonparametric regression, almost arbitrary new nonlinear policies can be generated without losing the stability properties of the canonical system.
Central pattern generators for locomotion control in animals and robots: A review
- A. Ijspeert
- BiologyNeural Networks
- 1 May 2008
Movement imitation with nonlinear dynamical systems in humanoid robots
- A. Ijspeert, J. Nakanishi, S. Schaal
- Computer ScienceProceedings IEEE International Conference on…
- 7 August 2002
The results demonstrate that multi-joint human movements can be encoded successfully by the CPs, that a learned movement policy can readily be reused to produce robust trajectories towards different targets, and that the parameter space which encodes a policy is suitable for measuring to which extent two trajectories are qualitatively similar.
Computational approaches to motor learning by imitation.
- S. Schaal, A. Ijspeert, A. Billard
- PsychologyPhilosophical transactions of the Royal Society…
- 29 March 2003
This paper will primarily emphasize the motor side of imitation, assuming that a perceptual system has already identified important features of a demonstrated movement and created their corresponding spatial information.
Dynamics systems vs. optimal control--a unifying view.
From Swimming to Walking with a Salamander Robot Driven by a Spinal Cord Model
A spinal cord model and its implementation in an amphibious salamander robot is presented that demonstrates how a primitive neural circuit for swimming can be extended by phylogenetically more recent limb oscillatory centers to explain the ability of salamanders to switch between swimming and walking.
Programmable central pattern generators: an application to biped locomotion control
- L. Righetti, A. Ijspeert
- Computer Science, EngineeringProceedings IEEE International Conference on…
- 15 May 2006
A novel system composed of coupled adaptive nonlinear oscillators that can learn arbitrary rhythmic signals in a supervised learning framework that can modulate the speed of locomotion, and even allow the reversal of direction.
Pattern generators with sensory feedback for the control of quadruped locomotion
A way of designing CPGs with coupled oscillators in which the controller can independently control the ascending and descending phases of the oscillations, and a systematic way of adding sensory feedback from touch sensors in the CPG such that the controller is strongly coupled with the mechanical system it controls.
Learning Movement Primitives
- S. Schaal, Jan Peters, J. Nakanishi, A. Ijspeert
- Computer ScienceInternational Symposium of Robotics Research
- 1 August 2005
A novel reinforcement learning technique based on natural stochastic policy gradients allows a general approach of improving DMPs by trial and error learning with respect to almost arbitrary optimization criteria, and demonstrates the different ingredients of the DMP approach in various examples.