Vector quantization based Gaussian modeling for speaker verification

Gaussian mixture models (GMMs) have become an established means of modeling feature distributions in speaker recognition systems. It is useful for experimentation and practical implementation purposes to develop and test these models in an efficient manner particularly when computational resources a...

Full description

Saved in:
Bibliographic Details
Main Authors: Pelecanos, J., Myers, S., Sridharan, S., Chandran, V.
Format: Conference Proceeding
Language:English
Subjects:
Australia
NIST
Probability density function
Robustness
Speaker recognition
Speech
System testing
Systems engineering and theory
Vector quantization
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
cited_by
cites
container_end_page 297 vol.3
container_issue
container_start_page 294
container_title
container_volume 3
creator Pelecanos, J.
Myers, S.
Sridharan, S.
Chandran, V.
description Gaussian mixture models (GMMs) have become an established means of modeling feature distributions in speaker recognition systems. It is useful for experimentation and practical implementation purposes to develop and test these models in an efficient manner particularly when computational resources are limited. A method of combining vector quantization (VQ) with single multi-dimensional Gaussians is proposed to rapidly generate a robust model approximation to the Gaussian mixture model. A fast method of testing these systems is also proposed and implemented. Results on the NIST 1996 Speaker Recognition Database suggest comparable and in some cases an improved verification performance to the traditional GMM based analysis scheme. In addition, previous research for the task of speaker identification indicated a similar system perfomance between the VQ Gaussian based technique and GMMs.
doi_str_mv 10.1109/ICPR.2000.903543
format conference_proceeding
fullrecord <record><control><sourceid>ieee_6IE</sourceid><recordid>TN_cdi_ieee_primary_903543</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>903543</ieee_id><sourcerecordid>903543</sourcerecordid><originalsourceid>FETCH-LOGICAL-i87t-43445128941ebd3c91b222cae43019ea12f0ff66b98175780e4b43b3d3d984a93</originalsourceid><addsrcrecordid>eNotj1tLw0AUhBcvYKx9F5_2DySes5fs7qNUWwsFRYqvZTc5kdU2qdlU0F9vsMLAvMw3zDB2jVAggrtdzp5fCgEAhQOplTxhmbASc6OMPmVTZyyY0mkwo85YhqAxV6XGC3aZ0juAGCmbsftXqoau558H3w7xxw-xa3nwiWq-8IeUom_5rqtpG9s33ozBtCf_QT3_oj42sfoDrth547eJpv8-Yev5w3r2mK-eFsvZ3SqP1gy5kkppFNYppFDLymEQQlSelAR05FE00DRlGZxFo8f9pIKSQdaydlZ5Jyfs5lgbiWiz7-PO99-b4335C8B4TCg</addsrcrecordid><sourcetype>Publisher</sourcetype><iscdi>true</iscdi><recordtype>conference_proceeding</recordtype></control><display><type>conference_proceeding</type><title>Vector quantization based Gaussian modeling for speaker verification</title><source>IEEE Electronic Library (IEL) Conference Proceedings</source><creator>Pelecanos, J. ; Myers, S. ; Sridharan, S. ; Chandran, V.</creator><creatorcontrib>Pelecanos, J. ; Myers, S. ; Sridharan, S. ; Chandran, V.</creatorcontrib><description>Gaussian mixture models (GMMs) have become an established means of modeling feature distributions in speaker recognition systems. It is useful for experimentation and practical implementation purposes to develop and test these models in an efficient manner particularly when computational resources are limited. A method of combining vector quantization (VQ) with single multi-dimensional Gaussians is proposed to rapidly generate a robust model approximation to the Gaussian mixture model. A fast method of testing these systems is also proposed and implemented. Results on the NIST 1996 Speaker Recognition Database suggest comparable and in some cases an improved verification performance to the traditional GMM based analysis scheme. In addition, previous research for the task of speaker identification indicated a similar system perfomance between the VQ Gaussian based technique and GMMs.</description><identifier>ISSN: 1051-4651</identifier><identifier>ISBN: 9780769507507</identifier><identifier>ISBN: 0769507506</identifier><identifier>EISSN: 2831-7475</identifier><identifier>DOI: 10.1109/ICPR.2000.903543</identifier><language>eng</language><publisher>IEEE</publisher><subject>Australia ; NIST ; Probability density function ; Robustness ; Speaker recognition ; Speech ; System testing ; Systems engineering and theory ; Vector quantization</subject><ispartof>Proceedings 15th International Conference on Pattern Recognition. ICPR-2000, 2000, Vol.3, p.294-297 vol.3</ispartof><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/903543$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>309,310,776,780,785,786,2052,4036,4037,27902,54530,54895,54907</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/903543$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc></links><search><creatorcontrib>Pelecanos, J.</creatorcontrib><creatorcontrib>Myers, S.</creatorcontrib><creatorcontrib>Sridharan, S.</creatorcontrib><creatorcontrib>Chandran, V.</creatorcontrib><title>Vector quantization based Gaussian modeling for speaker verification</title><title>Proceedings 15th International Conference on Pattern Recognition. ICPR-2000</title><addtitle>ICPR</addtitle><description>Gaussian mixture models (GMMs) have become an established means of modeling feature distributions in speaker recognition systems. It is useful for experimentation and practical implementation purposes to develop and test these models in an efficient manner particularly when computational resources are limited. A method of combining vector quantization (VQ) with single multi-dimensional Gaussians is proposed to rapidly generate a robust model approximation to the Gaussian mixture model. A fast method of testing these systems is also proposed and implemented. Results on the NIST 1996 Speaker Recognition Database suggest comparable and in some cases an improved verification performance to the traditional GMM based analysis scheme. In addition, previous research for the task of speaker identification indicated a similar system perfomance between the VQ Gaussian based technique and GMMs.</description><subject>Australia</subject><subject>NIST</subject><subject>Probability density function</subject><subject>Robustness</subject><subject>Speaker recognition</subject><subject>Speech</subject><subject>System testing</subject><subject>Systems engineering and theory</subject><subject>Vector quantization</subject><issn>1051-4651</issn><issn>2831-7475</issn><isbn>9780769507507</isbn><isbn>0769507506</isbn><fulltext>true</fulltext><rsrctype>conference_proceeding</rsrctype><creationdate>2000</creationdate><recordtype>conference_proceeding</recordtype><sourceid>6IE</sourceid><recordid>eNotj1tLw0AUhBcvYKx9F5_2DySes5fs7qNUWwsFRYqvZTc5kdU2qdlU0F9vsMLAvMw3zDB2jVAggrtdzp5fCgEAhQOplTxhmbASc6OMPmVTZyyY0mkwo85YhqAxV6XGC3aZ0juAGCmbsftXqoau558H3w7xxw-xa3nwiWq-8IeUom_5rqtpG9s33ozBtCf_QT3_oj42sfoDrth547eJpv8-Yev5w3r2mK-eFsvZ3SqP1gy5kkppFNYppFDLymEQQlSelAR05FE00DRlGZxFo8f9pIKSQdaydlZ5Jyfs5lgbiWiz7-PO99-b4335C8B4TCg</recordid><startdate>2000</startdate><enddate>2000</enddate><creator>Pelecanos, J.</creator><creator>Myers, S.</creator><creator>Sridharan, S.</creator><creator>Chandran, V.</creator><general>IEEE</general><scope>6IE</scope><scope>6IL</scope><scope>CBEJK</scope><scope>RIE</scope><scope>RIL</scope></search><sort><creationdate>2000</creationdate><title>Vector quantization based Gaussian modeling for speaker verification</title><author>Pelecanos, J. ; Myers, S. ; Sridharan, S. ; Chandran, V.</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-i87t-43445128941ebd3c91b222cae43019ea12f0ff66b98175780e4b43b3d3d984a93</frbrgroupid><rsrctype>conference_proceedings</rsrctype><prefilter>conference_proceedings</prefilter><language>eng</language><creationdate>2000</creationdate><topic>Australia</topic><topic>NIST</topic><topic>Probability density function</topic><topic>Robustness</topic><topic>Speaker recognition</topic><topic>Speech</topic><topic>System testing</topic><topic>Systems engineering and theory</topic><topic>Vector quantization</topic><toplevel>online_resources</toplevel><creatorcontrib>Pelecanos, J.</creatorcontrib><creatorcontrib>Myers, S.</creatorcontrib><creatorcontrib>Sridharan, S.</creatorcontrib><creatorcontrib>Chandran, V.</creatorcontrib><collection>IEEE Electronic Library (IEL) Conference Proceedings</collection><collection>IEEE Proceedings Order Plan All Online (POP All Online) 1998-present by volume</collection><collection>IEEE Xplore All Conference Proceedings</collection><collection>IEEE Xplore</collection><collection>IEEE Proceedings Order Plans (POP All) 1998-Present</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Pelecanos, J.</au><au>Myers, S.</au><au>Sridharan, S.</au><au>Chandran, V.</au><format>book</format><genre>proceeding</genre><ristype>CONF</ristype><atitle>Vector quantization based Gaussian modeling for speaker verification</atitle><btitle>Proceedings 15th International Conference on Pattern Recognition. ICPR-2000</btitle><stitle>ICPR</stitle><date>2000</date><risdate>2000</risdate><volume>3</volume><spage>294</spage><epage>297 vol.3</epage><pages>294-297 vol.3</pages><issn>1051-4651</issn><eissn>2831-7475</eissn><isbn>9780769507507</isbn><isbn>0769507506</isbn><abstract>Gaussian mixture models (GMMs) have become an established means of modeling feature distributions in speaker recognition systems. It is useful for experimentation and practical implementation purposes to develop and test these models in an efficient manner particularly when computational resources are limited. A method of combining vector quantization (VQ) with single multi-dimensional Gaussians is proposed to rapidly generate a robust model approximation to the Gaussian mixture model. A fast method of testing these systems is also proposed and implemented. Results on the NIST 1996 Speaker Recognition Database suggest comparable and in some cases an improved verification performance to the traditional GMM based analysis scheme. In addition, previous research for the task of speaker identification indicated a similar system perfomance between the VQ Gaussian based technique and GMMs.</abstract><pub>IEEE</pub><doi>10.1109/ICPR.2000.903543</doi></addata></record>
fulltext fulltext_linktorsrc
identifier ISSN: 1051-4651
ispartof Proceedings 15th International Conference on Pattern Recognition. ICPR-2000, 2000, Vol.3, p.294-297 vol.3
issn 1051-4651
2831-7475
language eng
recordid cdi_ieee_primary_903543
source IEEE Electronic Library (IEL) Conference Proceedings
subjects Australia
NIST
Probability density function
Robustness
Speaker recognition
Speech
System testing
Systems engineering and theory
Vector quantization
title Vector quantization based Gaussian modeling for speaker verification
url http://sfxeu10.hosted.exlibrisgroup.com/loughborough?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-02-06T13%3A03%3A16IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-ieee_6IE&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=proceeding&rft.atitle=Vector%20quantization%20based%20Gaussian%20modeling%20for%20speaker%20verification&rft.btitle=Proceedings%2015th%20International%20Conference%20on%20Pattern%20Recognition.%20ICPR-2000&rft.au=Pelecanos,%20J.&rft.date=2000&rft.volume=3&rft.spage=294&rft.epage=297%20vol.3&rft.pages=294-297%20vol.3&rft.issn=1051-4651&rft.eissn=2831-7475&rft.isbn=9780769507507&rft.isbn_list=0769507506&rft_id=info:doi/10.1109/ICPR.2000.903543&rft_dat=%3Cieee_6IE%3E903543%3C/ieee_6IE%3E%3Cgrp_id%3Ecdi_FETCH-LOGICAL-i87t-43445128941ebd3c91b222cae43019ea12f0ff66b98175780e4b43b3d3d984a93%3C/grp_id%3E%3Coa%3E%3C/oa%3E%3Curl%3E%3C/url%3E&rft_id=info:oai/&rft_id=info:pmid/&rft_ieee_id=903543&rfr_iscdi=true