Acoustic Model Interpolation for Non-Native Speech Recognition
This paper proposes three interpolation techniques which use the target language and the speaker's native language to improve non-native speech recognition system. These interpolation techniques are manual interpolation, weighted least square and eigenvoices. Each of them can be used under diff...
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 | IV-1012 |
|---|---|
| container_issue | |
| container_start_page | IV-1009 |
| container_title | |
| container_volume | 4 |
| creator | Tien-Ping Tan Besacier, L. |
| description | This paper proposes three interpolation techniques which use the target language and the speaker's native language to improve non-native speech recognition system. These interpolation techniques are manual interpolation, weighted least square and eigenvoices. Each of them can be used under different situation and constraints. In contrast to weighted least square and eigenvoices methods, manual interpolation can be achieved offline without any adaptation data. These methods can also be combined with MLLR to improve the recognition rate. Experiments presented in this paper show that the best non native adaptation method, combined with MLLR can give 10% WER absolute reduction on a French automatic speech recognition system for both Chinese and Vietnamese native speakers. |
| doi_str_mv | 10.1109/ICASSP.2007.367243 |
| format | Conference Proceeding |
| fullrecord | <record><control><sourceid>ieee_6IE</sourceid><recordid>TN_cdi_ieee_primary_4218274</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>4218274</ieee_id><sourcerecordid>4218274</sourcerecordid><originalsourceid>FETCH-LOGICAL-i241t-6da1184aa00bbbec50c823f5de28debb37e95cc032c356eb50751eb6ccd27f43</originalsourceid><addsrcrecordid>eNpVjstOwzAURM1LIpT-AGz8AynX13Zsb5CqqkClUhDpgl0VOzcQFOIoCUj8PUWwYTUandHRMHYhYCYEuKvVYp7njzMEMDOZGVTygE2dsUKhUmDQZocsQWlcKhw8H_1jxh2zRGiENBPKnbKzYXgDAGuUTdj1PMSPYawDv48lNXzVjtR3sSnGOra8ij3fxDbd7Osn8bwjCq_8iUJ8aeufxTk7qYpmoOlfTtj2Zrld3KXrh9v953VaoxJjmpWFEFYVBYD3noKGYFFWuiS0JXkvDTkdAkgMUmfkNRgtyGchlGgqJSfs8ldbE9Gu6-v3ov_aKRQWjZLfyHhOYw</addsrcrecordid><sourcetype>Publisher</sourcetype><iscdi>true</iscdi><recordtype>conference_proceeding</recordtype></control><display><type>conference_proceeding</type><title>Acoustic Model Interpolation for Non-Native Speech Recognition</title><source>IEEE Electronic Library (IEL) Conference Proceedings</source><creator>Tien-Ping Tan ; Besacier, L.</creator><creatorcontrib>Tien-Ping Tan ; Besacier, L.</creatorcontrib><description>This paper proposes three interpolation techniques which use the target language and the speaker's native language to improve non-native speech recognition system. These interpolation techniques are manual interpolation, weighted least square and eigenvoices. Each of them can be used under different situation and constraints. In contrast to weighted least square and eigenvoices methods, manual interpolation can be achieved offline without any adaptation data. These methods can also be combined with MLLR to improve the recognition rate. Experiments presented in this paper show that the best non native adaptation method, combined with MLLR can give 10% WER absolute reduction on a French automatic speech recognition system for both Chinese and Vietnamese native speakers.</description><identifier>ISSN: 1520-6149</identifier><identifier>ISBN: 9781424407279</identifier><identifier>ISBN: 1424407273</identifier><identifier>EISSN: 2379-190X</identifier><identifier>EISBN: 9781424407286</identifier><identifier>EISBN: 1424407281</identifier><identifier>DOI: 10.1109/ICASSP.2007.367243</identifier><language>eng</language><publisher>IEEE</publisher><subject>adaptation ; Adaptation model ; Automatic speech recognition ; Interpolation ; Least squares methods ; Loudspeakers ; Matrices ; Maximum likelihood linear regression ; Natural languages ; non-native ASR ; Speech recognition ; Tongue</subject><ispartof>2007 IEEE International Conference on Acoustics, Speech and Signal Processing - ICASSP '07, 2007, Vol.4, p.IV-1009-IV-1012</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/4218274$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>309,310,780,784,789,790,2058,27925,54920</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/4218274$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc></links><search><creatorcontrib>Tien-Ping Tan</creatorcontrib><creatorcontrib>Besacier, L.</creatorcontrib><title>Acoustic Model Interpolation for Non-Native Speech Recognition</title><title>2007 IEEE International Conference on Acoustics, Speech and Signal Processing - ICASSP '07</title><addtitle>ICASSP</addtitle><description>This paper proposes three interpolation techniques which use the target language and the speaker's native language to improve non-native speech recognition system. These interpolation techniques are manual interpolation, weighted least square and eigenvoices. Each of them can be used under different situation and constraints. In contrast to weighted least square and eigenvoices methods, manual interpolation can be achieved offline without any adaptation data. These methods can also be combined with MLLR to improve the recognition rate. Experiments presented in this paper show that the best non native adaptation method, combined with MLLR can give 10% WER absolute reduction on a French automatic speech recognition system for both Chinese and Vietnamese native speakers.</description><subject>adaptation</subject><subject>Adaptation model</subject><subject>Automatic speech recognition</subject><subject>Interpolation</subject><subject>Least squares methods</subject><subject>Loudspeakers</subject><subject>Matrices</subject><subject>Maximum likelihood linear regression</subject><subject>Natural languages</subject><subject>non-native ASR</subject><subject>Speech recognition</subject><subject>Tongue</subject><issn>1520-6149</issn><issn>2379-190X</issn><isbn>9781424407279</isbn><isbn>1424407273</isbn><isbn>9781424407286</isbn><isbn>1424407281</isbn><fulltext>true</fulltext><rsrctype>conference_proceeding</rsrctype><creationdate>2007</creationdate><recordtype>conference_proceeding</recordtype><sourceid>6IE</sourceid><sourceid>RIE</sourceid><recordid>eNpVjstOwzAURM1LIpT-AGz8AynX13Zsb5CqqkClUhDpgl0VOzcQFOIoCUj8PUWwYTUandHRMHYhYCYEuKvVYp7njzMEMDOZGVTygE2dsUKhUmDQZocsQWlcKhw8H_1jxh2zRGiENBPKnbKzYXgDAGuUTdj1PMSPYawDv48lNXzVjtR3sSnGOra8ij3fxDbd7Osn8bwjCq_8iUJ8aeufxTk7qYpmoOlfTtj2Zrld3KXrh9v953VaoxJjmpWFEFYVBYD3noKGYFFWuiS0JXkvDTkdAkgMUmfkNRgtyGchlGgqJSfs8ldbE9Gu6-v3ov_aKRQWjZLfyHhOYw</recordid><startdate>200704</startdate><enddate>200704</enddate><creator>Tien-Ping Tan</creator><creator>Besacier, L.</creator><general>IEEE</general><scope>6IE</scope><scope>6IH</scope><scope>CBEJK</scope><scope>RIE</scope><scope>RIO</scope></search><sort><creationdate>200704</creationdate><title>Acoustic Model Interpolation for Non-Native Speech Recognition</title><author>Tien-Ping Tan ; Besacier, L.</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-i241t-6da1184aa00bbbec50c823f5de28debb37e95cc032c356eb50751eb6ccd27f43</frbrgroupid><rsrctype>conference_proceedings</rsrctype><prefilter>conference_proceedings</prefilter><language>eng</language><creationdate>2007</creationdate><topic>adaptation</topic><topic>Adaptation model</topic><topic>Automatic speech recognition</topic><topic>Interpolation</topic><topic>Least squares methods</topic><topic>Loudspeakers</topic><topic>Matrices</topic><topic>Maximum likelihood linear regression</topic><topic>Natural languages</topic><topic>non-native ASR</topic><topic>Speech recognition</topic><topic>Tongue</topic><toplevel>online_resources</toplevel><creatorcontrib>Tien-Ping Tan</creatorcontrib><creatorcontrib>Besacier, L.</creatorcontrib><collection>IEEE Electronic Library (IEL) Conference Proceedings</collection><collection>IEEE Proceedings Order Plan (POP) 1998-present by volume</collection><collection>IEEE Xplore All Conference Proceedings</collection><collection>IEEE Electronic Library (IEL)</collection><collection>IEEE Proceedings Order Plans (POP) 1998-present</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Tien-Ping Tan</au><au>Besacier, L.</au><format>book</format><genre>proceeding</genre><ristype>CONF</ristype><atitle>Acoustic Model Interpolation for Non-Native Speech Recognition</atitle><btitle>2007 IEEE International Conference on Acoustics, Speech and Signal Processing - ICASSP '07</btitle><stitle>ICASSP</stitle><date>2007-04</date><risdate>2007</risdate><volume>4</volume><spage>IV-1009</spage><epage>IV-1012</epage><pages>IV-1009-IV-1012</pages><issn>1520-6149</issn><eissn>2379-190X</eissn><isbn>9781424407279</isbn><isbn>1424407273</isbn><eisbn>9781424407286</eisbn><eisbn>1424407281</eisbn><abstract>This paper proposes three interpolation techniques which use the target language and the speaker's native language to improve non-native speech recognition system. These interpolation techniques are manual interpolation, weighted least square and eigenvoices. Each of them can be used under different situation and constraints. In contrast to weighted least square and eigenvoices methods, manual interpolation can be achieved offline without any adaptation data. These methods can also be combined with MLLR to improve the recognition rate. Experiments presented in this paper show that the best non native adaptation method, combined with MLLR can give 10% WER absolute reduction on a French automatic speech recognition system for both Chinese and Vietnamese native speakers.</abstract><pub>IEEE</pub><doi>10.1109/ICASSP.2007.367243</doi></addata></record> |
| fulltext | fulltext_linktorsrc |
| identifier | ISSN: 1520-6149 |
| ispartof | 2007 IEEE International Conference on Acoustics, Speech and Signal Processing - ICASSP '07, 2007, Vol.4, p.IV-1009-IV-1012 |
| issn | 1520-6149 2379-190X |
| language | eng |
| recordid | cdi_ieee_primary_4218274 |
| source | IEEE Electronic Library (IEL) Conference Proceedings |
| subjects | adaptation Adaptation model Automatic speech recognition Interpolation Least squares methods Loudspeakers Matrices Maximum likelihood linear regression Natural languages non-native ASR Speech recognition Tongue |
| title | Acoustic Model Interpolation for Non-Native Speech Recognition |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-26T21%3A17%3A41IST&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=Acoustic%20Model%20Interpolation%20for%20Non-Native%20Speech%20Recognition&rft.btitle=2007%20IEEE%20International%20Conference%20on%20Acoustics,%20Speech%20and%20Signal%20Processing%20-%20ICASSP%20'07&rft.au=Tien-Ping%20Tan&rft.date=2007-04&rft.volume=4&rft.spage=IV-1009&rft.epage=IV-1012&rft.pages=IV-1009-IV-1012&rft.issn=1520-6149&rft.eissn=2379-190X&rft.isbn=9781424407279&rft.isbn_list=1424407273&rft_id=info:doi/10.1109/ICASSP.2007.367243&rft_dat=%3Cieee_6IE%3E4218274%3C/ieee_6IE%3E%3Curl%3E%3C/url%3E&rft.eisbn=9781424407286&rft.eisbn_list=1424407281&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rft_ieee_id=4218274&rfr_iscdi=true |