Low Power Dendritic Computation for Wordspotting

Low Power Dendritic Computation for Wordspotting

In this paper, we demonstrate how a network of dendrites can be used to build the state decoding block of a wordspotter similar to a Hidden Markov Model (HMM) classifier structure. We present simulation and experimental data for a single line dendrite and also experimental results for a dendrite-bas...

Full description

Bibliographic Details
Main Authors: Stephen Nease, Jennifer Hasler, Shubha Ramakrishnan, Scott Koziol, Suma George
Format: Article
Language:English
Published: MDPI AG 2013-05-01
Series:Journal of Low Power Electronics and Applications
Subjects:
computational modeling
hidden markov models
neuromorphic
dendrites
Online Access:http://www.mdpi.com/2079-9268/3/2/73
id doaj-1d746c61506848eea874dc98ff6d2292
record_format Article
spelling doaj-1d746c61506848eea874dc98ff6d22922020-11-24T21:09:45ZengMDPI AGJournal of Low Power Electronics and Applications2079-92682013-05-0132739810.3390/jlpea3020073Low Power Dendritic Computation for WordspottingStephen NeaseJennifer HaslerShubha RamakrishnanScott KoziolSuma GeorgeIn this paper, we demonstrate how a network of dendrites can be used to build the state decoding block of a wordspotter similar to a Hidden Markov Model (HMM) classifier structure. We present simulation and experimental data for a single line dendrite and also experimental results for a dendrite-based classifier structure. This work builds on previously demonstrated building blocks of a neural network: the channel, synapses and dendrites using CMOS circuits. These structures can be used for speech and pattern recognition. The computational efficiency of such a system is >10 MMACs/μW as compared to Digital Systems which perform 10 MMACs/mW.http://www.mdpi.com/2079-9268/3/2/73computational modelinghidden markov modelsneuromorphicdendrites
collection DOAJ
language English
format Article
sources DOAJ
author Stephen Nease
Jennifer Hasler
Shubha Ramakrishnan
Scott Koziol
Suma George
spellingShingle Stephen Nease
Jennifer Hasler
Shubha Ramakrishnan
Scott Koziol
Suma George
Low Power Dendritic Computation for Wordspotting
Journal of Low Power Electronics and Applications
computational modeling
hidden markov models
neuromorphic
dendrites
author_facet Stephen Nease
Jennifer Hasler
Shubha Ramakrishnan
Scott Koziol
Suma George
author_sort Stephen Nease
title Low Power Dendritic Computation for Wordspotting
title_short Low Power Dendritic Computation for Wordspotting
title_full Low Power Dendritic Computation for Wordspotting
title_fullStr Low Power Dendritic Computation for Wordspotting
title_full_unstemmed Low Power Dendritic Computation for Wordspotting
title_sort low power dendritic computation for wordspotting
publisher MDPI AG
series Journal of Low Power Electronics and Applications
issn 2079-9268
publishDate 2013-05-01
description In this paper, we demonstrate how a network of dendrites can be used to build the state decoding block of a wordspotter similar to a Hidden Markov Model (HMM) classifier structure. We present simulation and experimental data for a single line dendrite and also experimental results for a dendrite-based classifier structure. This work builds on previously demonstrated building blocks of a neural network: the channel, synapses and dendrites using CMOS circuits. These structures can be used for speech and pattern recognition. The computational efficiency of such a system is >10 MMACs/μW as compared to Digital Systems which perform 10 MMACs/mW.
topic computational modeling
hidden markov models
neuromorphic
dendrites
url http://www.mdpi.com/2079-9268/3/2/73
work_keys_str_mv AT stephennease lowpowerdendriticcomputationforwordspotting
AT jenniferhasler lowpowerdendriticcomputationforwordspotting
AT shubharamakrishnan lowpowerdendriticcomputationforwordspotting
AT scottkoziol lowpowerdendriticcomputationforwordspotting
AT sumageorge lowpowerdendriticcomputationforwordspotting
_version_ 1716757475546365952

Similar Items