Download Abstract View ANALOG TO DIGITAL CONVERSION USING RECURRENT SPIKING NEURAL NETWORKS ;

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Document related concepts

Neuroinformatics wikipedia, lookup

Molecular neuroscience wikipedia, lookup

Neurotransmitter wikipedia, lookup

Neural oscillation wikipedia, lookup

Neuroanatomy wikipedia, lookup

Caridoid escape reaction wikipedia, lookup

Biological neuron model wikipedia, lookup

Neural engineering wikipedia, lookup

Catastrophic interference wikipedia, lookup

Premovement neuronal activity wikipedia, lookup

Optogenetics wikipedia, lookup

Feature detection (nervous system) wikipedia, lookup

Neuropsychopharmacology wikipedia, lookup

Development of the nervous system wikipedia, lookup

Artificial neural network wikipedia, lookup

Neural coding wikipedia, lookup

Synaptic noise wikipedia, lookup

Pre-B├Âtzinger complex wikipedia, lookup

Metastability in the brain wikipedia, lookup

Chemical synapse wikipedia, lookup

Channelrhodopsin wikipedia, lookup

Convolutional neural network wikipedia, lookup

Synaptic gating wikipedia, lookup

Central pattern generator wikipedia, lookup

Nervous system network models wikipedia, lookup

Types of artificial neural networks wikipedia, lookup

Recurrent neural network wikipedia, lookup

Transcript
Itinerary - View Abstract
1 of 1
http://sfn.scholarone.com/itin2004/main.html?new_page_id=126&abstr...
Abstract View
ANALOG TO DIGITAL CONVERSION USING RECURRENT SPIKING NEURAL NETWORKS
B.C.Watson1; T.K.Fu1; A.R.Houweling2; D.J.Spencer2*; P.K.Das1; T.J.Sejnowski2
1. Dept. of Electrical and Computer Engin., Univ. of California San Diego, La Jolla, CA, USA
2. Computational Neurobiology Lab., Salk Inst. for Biological Studies, La Jolla, CA, USA
Networks of integrate-and-fire neurons with recurrent feedback can perform analog to digital conversion at a
rate that is proportional to the size of the network (E.K.Ressler et al, 2004, Proc. SPIE Int. Soc. Opt. Eng.
5200, 91). The individual neurons are coordinated using feedback in a manner that suppresses noise and
makes the output spike rate proportional to the level of the analog input signal without a predetermined
progression of states or an explicit clock. We explored the possibility that cortical networks with strong
excitatory and inhibitory loops and a range of latencies could also perform oversampling and noise shaping,
which would allow the transmission of information with large bandwidths despite the slow and imprecise
characteristics of individual neurons and synapses. Simulations using integrate-and-fire neurons in simple
networks, first with global inhibitory feedback and then with both inhibitory and excitatory feedback, confirm
the analytically predicted performance in terms of noise shaping and cell-to-cell spike train correlations
previously reported. Input information was also preserved through cascades of single-layer networks. Similar
simulations were performed using a recurrent network with Hodgkin-Huxley-like two-compartment models
and a realistic noise environment. The model had 5,000 pyramidal and 1,250 inhibitory interneurons with
local, sparse synaptic connections between all cell types.
Support Contributed By: Howard Hughes Medical Institute
Citation:B.C. Watson, T.K. Fu, A.R. Houweling, D.J. Spencer, P.K. Das, T.J. Sejnowski. ANALOG TO
DIGITAL CONVERSION USING RECURRENT SPIKING NEURAL NETWORKS Program No. 490.5.
2004 Abstract Viewer/Itinerary Planner. Washington, DC: Society for Neuroscience, 2004. Online.
2004 Copyright by the Society for Neuroscience all rights reserved. Permission to republish any
abstract or part of any abstract in any form must be obtained in writing from the SfN office prior to
publication
Site Design and Programming © ScholarOne, Inc., 2004. All Rights Reserved. Patent Pending.
10/28/2005 2:50 PM