Multonic Markov word models for large vocabulary continuous speech recognition
A new class of hidden Markov models is proposed for the acoustic representation of words in an automatic speech recognition system. The models, built from combinations of acoustically based sub-word units called fenones, are derived automatically from one or more sample utterances of a word. Because...
Gespeichert in:
| Veröffentlicht in: | IEEE transactions on speech and audio processing 1993-07, Vol.1 (3), p.334-344 |
|---|---|
| Hauptverfasser: | , , , , , |
| Format: | Artikel |
| Sprache: | eng |
| Schlagworte: | |
| Online-Zugang: | Volltext bestellen |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | 344 |
|---|---|
| container_issue | 3 |
| container_start_page | 334 |
| container_title | IEEE transactions on speech and audio processing |
| container_volume | 1 |
| creator | Bahl, L.R. Bellegarda, J.R. de Souza, P.V. Gopalakrishnan, P.S. Nahamoo, D. Picheny, M.A. |
| description | A new class of hidden Markov models is proposed for the acoustic representation of words in an automatic speech recognition system. The models, built from combinations of acoustically based sub-word units called fenones, are derived automatically from one or more sample utterances of a word. Because they are more flexible than previously reported fenone-based word models, they lead to an improved capability of modeling variations in pronunciation. They are therefore particularly useful in the recognition of continuous speech. In addition, their construction is relatively simple, because it can be done using the well-known forward-backward algorithm for parameter estimation of hidden Markov models. Appropriate reestimation formulas are derived for this purpose. Experimental results obtained on a 5000-word vocabulary natural language continuous speech recognition task are presented to illustrate the enhanced power of discrimination of the new models.< > |
| doi_str_mv | 10.1109/89.232617 |
| format | Article |
| fullrecord | <record><control><sourceid>proquest_RIE</sourceid><recordid>TN_cdi_ieee_primary_232617</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>232617</ieee_id><sourcerecordid>28393350</sourcerecordid><originalsourceid>FETCH-LOGICAL-c372t-c61a11ff9facd12ad4e1df007ca926ebfa21d4070b23457aa4367c9865c77883</originalsourceid><addsrcrecordid>eNo9kL1PwzAUxC0EEqUwsDJ5QmJI8UfijxFVFJBaWLpbrvNcDGlc7KSI_56gVEzvpPfT6e4QuqZkRinR90rPGGeCyhM0oVWlCsYrfjpoInghhBTn6CLnD0KIorKcoNdV33SxDQ6vbPqMB_wdU413sYYmYx8TbmzaAj5EZzf9oH-wi20X2j72Gec9gHvHCVzctqELsb1EZ942Ga6Od4rWi8f1_LlYvj29zB-WheOSdYUT1FLqvfbW1ZTZugRae0Kks5oJ2HjLaF0SSTaMl5W0tuRCOq1E5aRUik_R7Wi7T_Grh9yZXcgOmsa2MAQzTHHNeUUG8G4EXYo5J_Bmn8JuqGEoMX-DGaXNONjA3oxsAIB_7vj8Bf5kZto</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>28393350</pqid></control><display><type>article</type><title>Multonic Markov word models for large vocabulary continuous speech recognition</title><source>IEEE Electronic Library (IEL)</source><creator>Bahl, L.R. ; Bellegarda, J.R. ; de Souza, P.V. ; Gopalakrishnan, P.S. ; Nahamoo, D. ; Picheny, M.A.</creator><creatorcontrib>Bahl, L.R. ; Bellegarda, J.R. ; de Souza, P.V. ; Gopalakrishnan, P.S. ; Nahamoo, D. ; Picheny, M.A.</creatorcontrib><description>A new class of hidden Markov models is proposed for the acoustic representation of words in an automatic speech recognition system. The models, built from combinations of acoustically based sub-word units called fenones, are derived automatically from one or more sample utterances of a word. Because they are more flexible than previously reported fenone-based word models, they lead to an improved capability of modeling variations in pronunciation. They are therefore particularly useful in the recognition of continuous speech. In addition, their construction is relatively simple, because it can be done using the well-known forward-backward algorithm for parameter estimation of hidden Markov models. Appropriate reestimation formulas are derived for this purpose. Experimental results obtained on a 5000-word vocabulary natural language continuous speech recognition task are presented to illustrate the enhanced power of discrimination of the new models.< ></description><identifier>ISSN: 1063-6676</identifier><identifier>EISSN: 1558-2353</identifier><identifier>DOI: 10.1109/89.232617</identifier><identifier>CODEN: IESPEJ</identifier><language>eng</language><publisher>IEEE</publisher><subject>Automatic speech recognition ; Decoding ; Equations ; Hidden Markov models ; Loudspeakers ; Natural languages ; Parameter estimation ; Power system modeling ; Speech recognition ; Vocabulary</subject><ispartof>IEEE transactions on speech and audio processing, 1993-07, Vol.1 (3), p.334-344</ispartof><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c372t-c61a11ff9facd12ad4e1df007ca926ebfa21d4070b23457aa4367c9865c77883</citedby><cites>FETCH-LOGICAL-c372t-c61a11ff9facd12ad4e1df007ca926ebfa21d4070b23457aa4367c9865c77883</cites></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/232617$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>314,776,780,792,27901,27902,54733</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/232617$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc></links><search><creatorcontrib>Bahl, L.R.</creatorcontrib><creatorcontrib>Bellegarda, J.R.</creatorcontrib><creatorcontrib>de Souza, P.V.</creatorcontrib><creatorcontrib>Gopalakrishnan, P.S.</creatorcontrib><creatorcontrib>Nahamoo, D.</creatorcontrib><creatorcontrib>Picheny, M.A.</creatorcontrib><title>Multonic Markov word models for large vocabulary continuous speech recognition</title><title>IEEE transactions on speech and audio processing</title><addtitle>T-SAP</addtitle><description>A new class of hidden Markov models is proposed for the acoustic representation of words in an automatic speech recognition system. The models, built from combinations of acoustically based sub-word units called fenones, are derived automatically from one or more sample utterances of a word. Because they are more flexible than previously reported fenone-based word models, they lead to an improved capability of modeling variations in pronunciation. They are therefore particularly useful in the recognition of continuous speech. In addition, their construction is relatively simple, because it can be done using the well-known forward-backward algorithm for parameter estimation of hidden Markov models. Appropriate reestimation formulas are derived for this purpose. Experimental results obtained on a 5000-word vocabulary natural language continuous speech recognition task are presented to illustrate the enhanced power of discrimination of the new models.< ></description><subject>Automatic speech recognition</subject><subject>Decoding</subject><subject>Equations</subject><subject>Hidden Markov models</subject><subject>Loudspeakers</subject><subject>Natural languages</subject><subject>Parameter estimation</subject><subject>Power system modeling</subject><subject>Speech recognition</subject><subject>Vocabulary</subject><issn>1063-6676</issn><issn>1558-2353</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>1993</creationdate><recordtype>article</recordtype><recordid>eNo9kL1PwzAUxC0EEqUwsDJ5QmJI8UfijxFVFJBaWLpbrvNcDGlc7KSI_56gVEzvpPfT6e4QuqZkRinR90rPGGeCyhM0oVWlCsYrfjpoInghhBTn6CLnD0KIorKcoNdV33SxDQ6vbPqMB_wdU413sYYmYx8TbmzaAj5EZzf9oH-wi20X2j72Gec9gHvHCVzctqELsb1EZ942Ga6Od4rWi8f1_LlYvj29zB-WheOSdYUT1FLqvfbW1ZTZugRae0Kks5oJ2HjLaF0SSTaMl5W0tuRCOq1E5aRUik_R7Wi7T_Grh9yZXcgOmsa2MAQzTHHNeUUG8G4EXYo5J_Bmn8JuqGEoMX-DGaXNONjA3oxsAIB_7vj8Bf5kZto</recordid><startdate>19930701</startdate><enddate>19930701</enddate><creator>Bahl, L.R.</creator><creator>Bellegarda, J.R.</creator><creator>de Souza, P.V.</creator><creator>Gopalakrishnan, P.S.</creator><creator>Nahamoo, D.</creator><creator>Picheny, M.A.</creator><general>IEEE</general><scope>AAYXX</scope><scope>CITATION</scope><scope>7SC</scope><scope>8FD</scope><scope>JQ2</scope><scope>L7M</scope><scope>L~C</scope><scope>L~D</scope></search><sort><creationdate>19930701</creationdate><title>Multonic Markov word models for large vocabulary continuous speech recognition</title><author>Bahl, L.R. ; Bellegarda, J.R. ; de Souza, P.V. ; Gopalakrishnan, P.S. ; Nahamoo, D. ; Picheny, M.A.</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c372t-c61a11ff9facd12ad4e1df007ca926ebfa21d4070b23457aa4367c9865c77883</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>1993</creationdate><topic>Automatic speech recognition</topic><topic>Decoding</topic><topic>Equations</topic><topic>Hidden Markov models</topic><topic>Loudspeakers</topic><topic>Natural languages</topic><topic>Parameter estimation</topic><topic>Power system modeling</topic><topic>Speech recognition</topic><topic>Vocabulary</topic><toplevel>online_resources</toplevel><creatorcontrib>Bahl, L.R.</creatorcontrib><creatorcontrib>Bellegarda, J.R.</creatorcontrib><creatorcontrib>de Souza, P.V.</creatorcontrib><creatorcontrib>Gopalakrishnan, P.S.</creatorcontrib><creatorcontrib>Nahamoo, D.</creatorcontrib><creatorcontrib>Picheny, M.A.</creatorcontrib><collection>CrossRef</collection><collection>Computer and Information Systems Abstracts</collection><collection>Technology Research Database</collection><collection>ProQuest Computer Science Collection</collection><collection>Advanced Technologies Database with Aerospace</collection><collection>Computer and Information Systems Abstracts Academic</collection><collection>Computer and Information Systems Abstracts Professional</collection><jtitle>IEEE transactions on speech and audio processing</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Bahl, L.R.</au><au>Bellegarda, J.R.</au><au>de Souza, P.V.</au><au>Gopalakrishnan, P.S.</au><au>Nahamoo, D.</au><au>Picheny, M.A.</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Multonic Markov word models for large vocabulary continuous speech recognition</atitle><jtitle>IEEE transactions on speech and audio processing</jtitle><stitle>T-SAP</stitle><date>1993-07-01</date><risdate>1993</risdate><volume>1</volume><issue>3</issue><spage>334</spage><epage>344</epage><pages>334-344</pages><issn>1063-6676</issn><eissn>1558-2353</eissn><coden>IESPEJ</coden><abstract>A new class of hidden Markov models is proposed for the acoustic representation of words in an automatic speech recognition system. The models, built from combinations of acoustically based sub-word units called fenones, are derived automatically from one or more sample utterances of a word. Because they are more flexible than previously reported fenone-based word models, they lead to an improved capability of modeling variations in pronunciation. They are therefore particularly useful in the recognition of continuous speech. In addition, their construction is relatively simple, because it can be done using the well-known forward-backward algorithm for parameter estimation of hidden Markov models. Appropriate reestimation formulas are derived for this purpose. Experimental results obtained on a 5000-word vocabulary natural language continuous speech recognition task are presented to illustrate the enhanced power of discrimination of the new models.< ></abstract><pub>IEEE</pub><doi>10.1109/89.232617</doi><tpages>11</tpages></addata></record> |
| fulltext | fulltext_linktorsrc |
| identifier | ISSN: 1063-6676 |
| ispartof | IEEE transactions on speech and audio processing, 1993-07, Vol.1 (3), p.334-344 |
| issn | 1063-6676 1558-2353 |
| language | eng |
| recordid | cdi_ieee_primary_232617 |
| source | IEEE Electronic Library (IEL) |
| subjects | Automatic speech recognition Decoding Equations Hidden Markov models Loudspeakers Natural languages Parameter estimation Power system modeling Speech recognition Vocabulary |
| title | Multonic Markov word models for large vocabulary continuous speech recognition |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-29T20%3A01%3A26IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_RIE&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Multonic%20Markov%20word%20models%20for%20large%20vocabulary%20continuous%20speech%20recognition&rft.jtitle=IEEE%20transactions%20on%20speech%20and%20audio%20processing&rft.au=Bahl,%20L.R.&rft.date=1993-07-01&rft.volume=1&rft.issue=3&rft.spage=334&rft.epage=344&rft.pages=334-344&rft.issn=1063-6676&rft.eissn=1558-2353&rft.coden=IESPEJ&rft_id=info:doi/10.1109/89.232617&rft_dat=%3Cproquest_RIE%3E28393350%3C/proquest_RIE%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=28393350&rft_id=info:pmid/&rft_ieee_id=232617&rfr_iscdi=true |