Several behavioral experiments suggest that the nervous system uses
an internal model of the dynamics of the body to implement a close
approximation to a Kalman filter. This filter can be used to perform
a variety of tasks nearly optimally, such as predicting the sensory
consequence of motor action, integrating sensory and body posture
signals, and computing motor commands. We propose that the neural
implementation of this Kalman filter involves recurrent basis function
networks with attractor dynamics, a kind of architecture that can
be readily mapped onto cortical circuits. In such networks, the tuning
curves to variables such as arm velocity are remarkably noninvariant
in the sense that the amplitude and width of the tuning curves of
a given neuron can vary greatly depending on other variables such
as the position of the arm or the reliability of the sensory feedback.
This property could explain some puzzling properties of tuning curves
in the motor and premotor cortex, and it leads to several new predictions.
%0 Journal Article
%1 Deneve:2007
%A Denève, Sophie
%A Duhamel, Jean-Ren\´e
%A Pouget, Alexandre
%D 2007
%J The Journal of Neuroscience
%K Kalman attractors; basis code; control; filter; functions integration; line motor population sensorimotor
%P 5744 5756
%T Optimal Sensorimotor Integration in Recurrent Cortical Networks:
A Neural Implementation of Kalman Filters
%V 27
%X Several behavioral experiments suggest that the nervous system uses
an internal model of the dynamics of the body to implement a close
approximation to a Kalman filter. This filter can be used to perform
a variety of tasks nearly optimally, such as predicting the sensory
consequence of motor action, integrating sensory and body posture
signals, and computing motor commands. We propose that the neural
implementation of this Kalman filter involves recurrent basis function
networks with attractor dynamics, a kind of architecture that can
be readily mapped onto cortical circuits. In such networks, the tuning
curves to variables such as arm velocity are remarkably noninvariant
in the sense that the amplitude and width of the tuning curves of
a given neuron can vary greatly depending on other variables such
as the position of the arm or the reliability of the sensory feedback.
This property could explain some puzzling properties of tuning curves
in the motor and premotor cortex, and it leads to several new predictions.
@article{Deneve:2007,
abstract = {Several behavioral experiments suggest that the nervous system uses
an internal model of the dynamics of the body to implement a close
approximation to a Kalman filter. This filter can be used to perform
a variety of tasks nearly optimally, such as predicting the sensory
consequence of motor action, integrating sensory and body posture
signals, and computing motor commands. We propose that the neural
implementation of this Kalman filter involves recurrent basis function
networks with attractor dynamics, a kind of architecture that can
be readily mapped onto cortical circuits. In such networks, the tuning
curves to variables such as arm velocity are remarkably noninvariant
in the sense that the amplitude and width of the tuning curves of
a given neuron can vary greatly depending on other variables such
as the position of the arm or the reliability of the sensory feedback.
This property could explain some puzzling properties of tuning curves
in the motor and premotor cortex, and it leads to several new predictions.},
added-at = {2009-06-26T15:25:19.000+0200},
author = {Den\`eve, Sophie and Duhamel, Jean-Ren\´e and Pouget, Alexandre},
biburl = {https://www.bibsonomy.org/bibtex/2dbbe6fb5544de4e79b4c2e4de3e87bd5/butz},
description = {diverse cognitive systems bib},
interhash = {1a217a9859d297570fb568ccbef8cb86},
intrahash = {dbbe6fb5544de4e79b4c2e4de3e87bd5},
journal = {The Journal of Neuroscience},
keywords = {Kalman attractors; basis code; control; filter; functions integration; line motor population sensorimotor},
owner = {butz},
pages = {5744 5756},
timestamp = {2009-06-26T15:25:27.000+0200},
title = {Optimal Sensorimotor Integration in Recurrent Cortical Networks:
A Neural Implementation of {K}alman Filters},
volume = 27,
year = 2007
}