A model distance measure for talker clustering and identification
This paper describes methods of talker clustering and identification based on a "distance" metric between discrete HMM output probabilities. Output probabilities are derived on a tree-based MMI partition of the feature space, rather than the usual vector quantization. The information diver...
Gespeichert in:
| Hauptverfasser: | , |
|---|---|
| Format: | Tagungsbericht |
| Sprache: | eng |
| Schlagworte: | |
| Online-Zugang: | Volltext bestellen |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | I/320 vol.1 |
|---|---|
| container_issue | |
| container_start_page | I/317 |
| container_title | |
| container_volume | i |
| creator | Foote, J.T. Silverman, H.F. |
| description | This paper describes methods of talker clustering and identification based on a "distance" metric between discrete HMM output probabilities. Output probabilities are derived on a tree-based MMI partition of the feature space, rather than the usual vector quantization. The information divergence (relative entropy) between speaker-dependent models is used as a quantitative measure of how much a given talker differs from another talker. An immediate application is talker identification: an unknown speaker may be identified by finding the closest speaker-dependent reference model to a model trained on the unknown speaker's data. Another application is to cluster similar talkers into a group; these may be used to train a HMM model that represents that talker better than a more general model. It is shown that using the model "nearest" a novel talker enhances the performance of a talker-independent speech recognition system.< > |
| doi_str_mv | 10.1109/ICASSP.1994.389292 |
| format | Conference Proceeding |
| fullrecord | <record><control><sourceid>ieee_6IE</sourceid><recordid>TN_cdi_ieee_primary_389292</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>389292</ieee_id><sourcerecordid>389292</sourcerecordid><originalsourceid>FETCH-LOGICAL-i87t-f925ca6f0f0e034ee3660e9c04ab76a0bc874b7590f45c713c93cca5a25480053</originalsourceid><addsrcrecordid>eNotj8lOwzAURS0GiVD6A135B1KepzhvGVVAkSqB1C7YVY7zjAwZkOMu-Hsqlbs5u6NzGVsJWAsB-Pi6afb797VA1GtVo0R5xQqpLJYC4eOa3YOtQQlrDdywQhgJZSU03rHlPH_BedoYLXTBmoYPU0c97-Kc3eiJD-TmUyIepsSz678pcd-f5kwpjp_cjR2PHY05huhdjtP4wG6D62da_nPBDs9Ph8223L29nDN3ZaxtLgNK410VIACB0kSqqoDQg3atrRy0vra6tQYhaOOtUB6V9844aXQNYNSCrS7aSETHnxQHl36Pl-_qD1NgS9M</addsrcrecordid><sourcetype>Publisher</sourcetype><iscdi>true</iscdi><recordtype>conference_proceeding</recordtype></control><display><type>conference_proceeding</type><title>A model distance measure for talker clustering and identification</title><source>IEEE Electronic Library (IEL) Conference Proceedings</source><creator>Foote, J.T. ; Silverman, H.F.</creator><creatorcontrib>Foote, J.T. ; Silverman, H.F.</creatorcontrib><description>This paper describes methods of talker clustering and identification based on a "distance" metric between discrete HMM output probabilities. Output probabilities are derived on a tree-based MMI partition of the feature space, rather than the usual vector quantization. The information divergence (relative entropy) between speaker-dependent models is used as a quantitative measure of how much a given talker differs from another talker. An immediate application is talker identification: an unknown speaker may be identified by finding the closest speaker-dependent reference model to a model trained on the unknown speaker's data. Another application is to cluster similar talkers into a group; these may be used to train a HMM model that represents that talker better than a more general model. It is shown that using the model "nearest" a novel talker enhances the performance of a talker-independent speech recognition system.< ></description><identifier>ISSN: 1520-6149</identifier><identifier>ISBN: 0780317750</identifier><identifier>ISBN: 9780780317758</identifier><identifier>EISSN: 2379-190X</identifier><identifier>DOI: 10.1109/ICASSP.1994.389292</identifier><language>eng</language><publisher>IEEE</publisher><subject>Acoustic emission ; Decision trees ; Entropy ; Hidden Markov models ; Parameter estimation ; Probability distribution ; Q measurement ; Speech recognition ; Vector quantization ; Viterbi algorithm</subject><ispartof>Proceedings of ICASSP '94. IEEE International Conference on Acoustics, Speech and Signal Processing, 1994, Vol.i, p.I/317-I/320 vol.1</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/389292$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>309,310,780,784,789,790,2058,4050,4051,27925,54920</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/389292$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc></links><search><creatorcontrib>Foote, J.T.</creatorcontrib><creatorcontrib>Silverman, H.F.</creatorcontrib><title>A model distance measure for talker clustering and identification</title><title>Proceedings of ICASSP '94. IEEE International Conference on Acoustics, Speech and Signal Processing</title><addtitle>ICASSP</addtitle><description>This paper describes methods of talker clustering and identification based on a "distance" metric between discrete HMM output probabilities. Output probabilities are derived on a tree-based MMI partition of the feature space, rather than the usual vector quantization. The information divergence (relative entropy) between speaker-dependent models is used as a quantitative measure of how much a given talker differs from another talker. An immediate application is talker identification: an unknown speaker may be identified by finding the closest speaker-dependent reference model to a model trained on the unknown speaker's data. Another application is to cluster similar talkers into a group; these may be used to train a HMM model that represents that talker better than a more general model. It is shown that using the model "nearest" a novel talker enhances the performance of a talker-independent speech recognition system.< ></description><subject>Acoustic emission</subject><subject>Decision trees</subject><subject>Entropy</subject><subject>Hidden Markov models</subject><subject>Parameter estimation</subject><subject>Probability distribution</subject><subject>Q measurement</subject><subject>Speech recognition</subject><subject>Vector quantization</subject><subject>Viterbi algorithm</subject><issn>1520-6149</issn><issn>2379-190X</issn><isbn>0780317750</isbn><isbn>9780780317758</isbn><fulltext>true</fulltext><rsrctype>conference_proceeding</rsrctype><creationdate>1994</creationdate><recordtype>conference_proceeding</recordtype><sourceid>6IE</sourceid><sourceid>RIE</sourceid><recordid>eNotj8lOwzAURS0GiVD6A135B1KepzhvGVVAkSqB1C7YVY7zjAwZkOMu-Hsqlbs5u6NzGVsJWAsB-Pi6afb797VA1GtVo0R5xQqpLJYC4eOa3YOtQQlrDdywQhgJZSU03rHlPH_BedoYLXTBmoYPU0c97-Kc3eiJD-TmUyIepsSz678pcd-f5kwpjp_cjR2PHY05huhdjtP4wG6D62da_nPBDs9Ph8223L29nDN3ZaxtLgNK410VIACB0kSqqoDQg3atrRy0vra6tQYhaOOtUB6V9844aXQNYNSCrS7aSETHnxQHl36Pl-_qD1NgS9M</recordid><startdate>1994</startdate><enddate>1994</enddate><creator>Foote, J.T.</creator><creator>Silverman, H.F.</creator><general>IEEE</general><scope>6IE</scope><scope>6IL</scope><scope>CBEJK</scope><scope>RIE</scope><scope>RIL</scope></search><sort><creationdate>1994</creationdate><title>A model distance measure for talker clustering and identification</title><author>Foote, J.T. ; Silverman, H.F.</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-i87t-f925ca6f0f0e034ee3660e9c04ab76a0bc874b7590f45c713c93cca5a25480053</frbrgroupid><rsrctype>conference_proceedings</rsrctype><prefilter>conference_proceedings</prefilter><language>eng</language><creationdate>1994</creationdate><topic>Acoustic emission</topic><topic>Decision trees</topic><topic>Entropy</topic><topic>Hidden Markov models</topic><topic>Parameter estimation</topic><topic>Probability distribution</topic><topic>Q measurement</topic><topic>Speech recognition</topic><topic>Vector quantization</topic><topic>Viterbi algorithm</topic><toplevel>online_resources</toplevel><creatorcontrib>Foote, J.T.</creatorcontrib><creatorcontrib>Silverman, H.F.</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 Electronic Library (IEL)</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>Foote, J.T.</au><au>Silverman, H.F.</au><format>book</format><genre>proceeding</genre><ristype>CONF</ristype><atitle>A model distance measure for talker clustering and identification</atitle><btitle>Proceedings of ICASSP '94. IEEE International Conference on Acoustics, Speech and Signal Processing</btitle><stitle>ICASSP</stitle><date>1994</date><risdate>1994</risdate><volume>i</volume><spage>I/317</spage><epage>I/320 vol.1</epage><pages>I/317-I/320 vol.1</pages><issn>1520-6149</issn><eissn>2379-190X</eissn><isbn>0780317750</isbn><isbn>9780780317758</isbn><abstract>This paper describes methods of talker clustering and identification based on a "distance" metric between discrete HMM output probabilities. Output probabilities are derived on a tree-based MMI partition of the feature space, rather than the usual vector quantization. The information divergence (relative entropy) between speaker-dependent models is used as a quantitative measure of how much a given talker differs from another talker. An immediate application is talker identification: an unknown speaker may be identified by finding the closest speaker-dependent reference model to a model trained on the unknown speaker's data. Another application is to cluster similar talkers into a group; these may be used to train a HMM model that represents that talker better than a more general model. It is shown that using the model "nearest" a novel talker enhances the performance of a talker-independent speech recognition system.< ></abstract><pub>IEEE</pub><doi>10.1109/ICASSP.1994.389292</doi></addata></record> |
| fulltext | fulltext_linktorsrc |
| identifier | ISSN: 1520-6149 |
| ispartof | Proceedings of ICASSP '94. IEEE International Conference on Acoustics, Speech and Signal Processing, 1994, Vol.i, p.I/317-I/320 vol.1 |
| issn | 1520-6149 2379-190X |
| language | eng |
| recordid | cdi_ieee_primary_389292 |
| source | IEEE Electronic Library (IEL) Conference Proceedings |
| subjects | Acoustic emission Decision trees Entropy Hidden Markov models Parameter estimation Probability distribution Q measurement Speech recognition Vector quantization Viterbi algorithm |
| title | A model distance measure for talker clustering and identification |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-21T19%3A11%3A26IST&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=A%20model%20distance%20measure%20for%20talker%20clustering%20and%20identification&rft.btitle=Proceedings%20of%20ICASSP%20'94.%20IEEE%20International%20Conference%20on%20Acoustics,%20Speech%20and%20Signal%20Processing&rft.au=Foote,%20J.T.&rft.date=1994&rft.volume=i&rft.spage=I/317&rft.epage=I/320%20vol.1&rft.pages=I/317-I/320%20vol.1&rft.issn=1520-6149&rft.eissn=2379-190X&rft.isbn=0780317750&rft.isbn_list=9780780317758&rft_id=info:doi/10.1109/ICASSP.1994.389292&rft_dat=%3Cieee_6IE%3E389292%3C/ieee_6IE%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rft_ieee_id=389292&rfr_iscdi=true |