Jefferson Provost and Marc Sommer. A topographic model of shifting visual receptive fields. In Society for Neuroscience Abstracts, November 2007.
As the eyes move, the retinas are presented with a rapid sequence of different images, and yet we perceive a cohesive stable scene. Recently, a long-postulated pathway for corollary discharge of saccadic eye movement commands was found to ascend from superior colliculus through mediodorsal thalamus to the frontal eye field (FEF). This signal causes visually responsive cells in FEF to predictively shift their receptive fields (RFs) prior to the saccade. We are studying several hypotheses as to how these shifting RFs contribute to the percept of visual stability. We are using computational models to investigate these, based on a model neuron that combines a very broadly tuned, retinotopically mapped input field, and topographically mapped modulatory input encoding an impending saccade. The broad input field of the neuron gives it covert information about the entire visual field, even though it only responds overtly (with action potentials) to a narrow region of visual space during fixation. When the saccade signal arrives from SC, the modulatory field temporarily increases the weight of synapses offset from the RF, making the neuron responsive at a new location (the "future field"). Based on these neurons, we are implementing a large-scale computational model of the FEF visual and motor circuits using Topographica, a neural simulator designed for investigating the structure and function of neural systems composed of interconnected topographic maps. The goals of our model are to explain the various dynamic signals seen in the FEF, and to provide useful perceptual stability for a robot with a servo-mounted camera.
@InProceedings{provost-sfn07,
author = {Jefferson Provost and Marc Sommer},
title = {A topographic model of shifting visual receptive fields},
booktitle = {Society for Neuroscience Abstracts},
organization = {Society for Neuroscience},
month = {November},
year = 2007,
abstract = {As the eyes move, the retinas are presented with a rapid
sequence of different images, and yet we perceive a cohesive stable
scene. Recently, a long-postulated pathway for corollary discharge
of saccadic eye movement commands was found to ascend from superior
colliculus through mediodorsal thalamus to the frontal eye field
(FEF). This signal causes visually responsive cells in FEF to
predictively shift their receptive fields (RFs) prior to the
saccade. We are studying several hypotheses as to how these shifting
RFs contribute to the percept of visual stability. We are using
computational models to investigate these, based on a model neuron
that combines a very broadly tuned, retinotopically mapped input
field, and topographically mapped modulatory input encoding an
impending saccade. The broad input field of the neuron gives it
covert information about the entire visual field, even though it only
responds overtly (with action potentials) to a narrow region of
visual space during fixation. When the saccade signal arrives from
SC, the modulatory field temporarily increases the weight of synapses
offset from the RF, making the neuron responsive at a new location
(the "future field"). Based on these neurons, we are implementing a
large-scale computational model of the FEF visual and motor circuits
using Topographica, a neural simulator designed for investigating the
structure and function of neural systems composed of interconnected
topographic maps. The goals of our model are to explain the various
dynamic signals seen in the FEF, and to provide useful perceptual
stability for a robot with a servo-mounted camera.}
}
Generated by bib2html.pl (written by Patrick Riley ) on Fri Nov 02, 2007 19:46:51