Hierarchical dynamical models of motor function

Authors:
S. M. Stringer;E. T. Rolls
Affiliations:
Department of Experimental Psychology, Centre for Computational Neuroscience, Oxford University, South Parks Road, Oxford OX1 3UD, UK;Department of Experimental Psychology, Centre for Computational Neuroscience, Oxford University, South Parks Road, Oxford OX1 3UD, UK
Venue:
Neurocomputing
Year:
2007

Citing 5
Cited 4

Forward models for physiological motor control

Neural Networks - 1996 Special issue: four major hypotheses in neuroscience
Self-organizing continuous attractor networks and motor function

Neural Networks
Self-organising continuous attractor networks with multiple activity packets, and the representation of space

Neural Networks
2005 Special issue: Spatial view cells in the hippocampus, and their idiothetic update based on place and head direction

Neural Networks - Special issue: Computational theories of the functions of the hippocampus
Learning movement sequences with a delayed reward signal in a hierarchical model of motor function

Neural Networks

Self-organizing continuous attractor networks and motor function

Neural Networks
The evolution of cognition: from first order to second order embodiment

ZiF'06 Proceedings of the Embodied communication in humans and machines, 2nd ZiF research group international conference on Modeling communication with robots and virtual humans
Universally manipulable body models--dual quaternion representations in layered and dynamic MMCs

Autonomous Robots
A dynamic field architecture for the generation of hierarchically organized sequences

ICANN'12 Proceedings of the 22nd international conference on Artificial Neural Networks and Machine Learning - Volume Part I

Quantified Score

Hi-index	0.02

Visualization

Abstract

Hierarchical models of motor function are described in which the motor system encodes a hierarchy of dynamical motor primitives. The models are based on continuous attractor neural networks, in which the packet of neural activity representing the position (in the state space) is moved by path integration using a motor efference copy after training with a traced associative learning rule. Sequences of movements can be learned because the state space modulates the effects of an intermediate level motor command. A high-level motor command selects a set of intermediate level motor commands, and the whole movement sequence evolves with the correct order because the state space representation effectively selects the next possible intermediate level command. Thus a single high-level motor command can select whole movement trajectories composed of a set of motor primitives. It is further shown that the network can perform completely novel motor sequences, and thus provides one solution to Lashley's 'problem of serial order in behaviour'.