Fully Automated Approach to Broadcast News Transcription in Czech Language
In the paper we propose a complete scheme for automatic transcription of Czech TV news. The scheme first removes the music and noisy parts, then makes segmentation of the speech signal into speaker turns and consequently tries to decode and transcribe single utterances. We employ our own recognizer...
Gespeichert in:
| Hauptverfasser: | , , |
|---|---|
| Format: | Tagungsbericht |
| Sprache: | eng |
| Schlagworte: | |
| Online-Zugang: | Volltext |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | 408 |
|---|---|
| container_issue | |
| container_start_page | 401 |
| container_title | |
| container_volume | |
| creator | Nouza, Jan Žďánský, Jindřich David, Petr |
| description | In the paper we propose a complete scheme for automatic transcription of Czech TV news. The scheme first removes the music and noisy parts, then makes segmentation of the speech signal into speaker turns and consequently tries to decode and transcribe single utterances. We employ our own recognizer recently operating with a 200K-word lexicon and with a bigram language model. The overall recognition rate achieved on all the test data was 71.53%, that obtained on the read parts was 82.72%. The most serious recognition errors occur mainly in the segments that contain background music or extremely loud noise. |
| doi_str_mv | 10.1007/978-3-540-30120-2_51 |
| format | Conference Proceeding |
| fullrecord | <record><control><sourceid>pascalfrancis_sprin</sourceid><recordid>TN_cdi_pascalfrancis_primary_16194695</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>16194695</sourcerecordid><originalsourceid>FETCH-LOGICAL-p228t-bcb12db7e2a6fbfae3c3283078d07175730bce2fb31bf2f781ea2e7b86e0e76a3</originalsourceid><addsrcrecordid>eNo9kE1PwzAMhsOXxDT2DzjkwjGQ2G3THsfE-FAFl3GOkjQZha6tmk5o_HqyDeGLrdePLOsh5FrwW8G5vCtkzpClCWfIBXAGKhUnZBZjjOEhg1MyEZkQDDEpzv53gDwpsnMyiRSwQiZ4SWYhfPJYkMo0xQl5WW6bZkfn27Hb6NFVdN73Q6ftBx07eh-nyuow0lf3Hehq0G2wQ92PddfSuqWLHxfBUrfrrV67K3LhdRPc7K9PyfvyYbV4YuXb4_NiXrIeIB-ZsUZAZaQDnXnjtUOLkCOXecWlkKlEbqwDb1AYD17mwmlw0uSZ405mGqfk5ni318HqxsevbB1UP9QbPexUFFEkWZFGDo5ciKt27QZluu4rKMHVXqyKlhSq6EkdJKq9WPwF6vxmUA</addsrcrecordid><sourcetype>Index Database</sourcetype><iscdi>true</iscdi><recordtype>conference_proceeding</recordtype></control><display><type>conference_proceeding</type><title>Fully Automated Approach to Broadcast News Transcription in Czech Language</title><source>Springer Books</source><creator>Nouza, Jan ; Žďánský, Jindřich ; David, Petr</creator><contributor>Pala, Karel ; Sojka, Petr ; Kopeček, Ivan</contributor><creatorcontrib>Nouza, Jan ; Žďánský, Jindřich ; David, Petr ; Pala, Karel ; Sojka, Petr ; Kopeček, Ivan</creatorcontrib><description>In the paper we propose a complete scheme for automatic transcription of Czech TV news. The scheme first removes the music and noisy parts, then makes segmentation of the speech signal into speaker turns and consequently tries to decode and transcribe single utterances. We employ our own recognizer recently operating with a 200K-word lexicon and with a bigram language model. The overall recognition rate achieved on all the test data was 71.53%, that obtained on the read parts was 82.72%. The most serious recognition errors occur mainly in the segments that contain background music or extremely loud noise.</description><identifier>ISSN: 0302-9743</identifier><identifier>ISBN: 9783540230496</identifier><identifier>ISBN: 3540230491</identifier><identifier>EISSN: 1611-3349</identifier><identifier>EISBN: 9783540301202</identifier><identifier>EISBN: 3540301208</identifier><identifier>DOI: 10.1007/978-3-540-30120-2_51</identifier><language>eng</language><publisher>Berlin, Heidelberg: Springer Berlin Heidelberg</publisher><subject>Acoustic Model ; Applied sciences ; Artificial intelligence ; Automatic Speech Recognition ; Bayesian Information Criterion ; Computer science; control theory; systems ; Exact sciences and technology ; Speech and sound recognition and synthesis. Linguistics ; Speech Recognition ; Speech Signal</subject><ispartof>Lecture notes in computer science, 2004, p.401-408</ispartof><rights>Springer-Verlag Berlin Heidelberg 2004</rights><rights>2004 INIST-CNRS</rights><lds50>peer_reviewed</lds50><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktopdf>$$Uhttps://link.springer.com/content/pdf/10.1007/978-3-540-30120-2_51$$EPDF$$P50$$Gspringer$$H</linktopdf><linktohtml>$$Uhttps://link.springer.com/10.1007/978-3-540-30120-2_51$$EHTML$$P50$$Gspringer$$H</linktohtml><link.rule.ids>309,310,776,777,781,786,787,790,4036,4037,27906,38236,41423,42492</link.rule.ids><backlink>$$Uhttp://pascal-francis.inist.fr/vibad/index.php?action=getRecordDetail&idt=16194695$$DView record in Pascal Francis$$Hfree_for_read</backlink></links><search><contributor>Pala, Karel</contributor><contributor>Sojka, Petr</contributor><contributor>Kopeček, Ivan</contributor><creatorcontrib>Nouza, Jan</creatorcontrib><creatorcontrib>Žďánský, Jindřich</creatorcontrib><creatorcontrib>David, Petr</creatorcontrib><title>Fully Automated Approach to Broadcast News Transcription in Czech Language</title><title>Lecture notes in computer science</title><description>In the paper we propose a complete scheme for automatic transcription of Czech TV news. The scheme first removes the music and noisy parts, then makes segmentation of the speech signal into speaker turns and consequently tries to decode and transcribe single utterances. We employ our own recognizer recently operating with a 200K-word lexicon and with a bigram language model. The overall recognition rate achieved on all the test data was 71.53%, that obtained on the read parts was 82.72%. The most serious recognition errors occur mainly in the segments that contain background music or extremely loud noise.</description><subject>Acoustic Model</subject><subject>Applied sciences</subject><subject>Artificial intelligence</subject><subject>Automatic Speech Recognition</subject><subject>Bayesian Information Criterion</subject><subject>Computer science; control theory; systems</subject><subject>Exact sciences and technology</subject><subject>Speech and sound recognition and synthesis. Linguistics</subject><subject>Speech Recognition</subject><subject>Speech Signal</subject><issn>0302-9743</issn><issn>1611-3349</issn><isbn>9783540230496</isbn><isbn>3540230491</isbn><isbn>9783540301202</isbn><isbn>3540301208</isbn><fulltext>true</fulltext><rsrctype>conference_proceeding</rsrctype><creationdate>2004</creationdate><recordtype>conference_proceeding</recordtype><recordid>eNo9kE1PwzAMhsOXxDT2DzjkwjGQ2G3THsfE-FAFl3GOkjQZha6tmk5o_HqyDeGLrdePLOsh5FrwW8G5vCtkzpClCWfIBXAGKhUnZBZjjOEhg1MyEZkQDDEpzv53gDwpsnMyiRSwQiZ4SWYhfPJYkMo0xQl5WW6bZkfn27Hb6NFVdN73Q6ftBx07eh-nyuow0lf3Hehq0G2wQ92PddfSuqWLHxfBUrfrrV67K3LhdRPc7K9PyfvyYbV4YuXb4_NiXrIeIB-ZsUZAZaQDnXnjtUOLkCOXecWlkKlEbqwDb1AYD17mwmlw0uSZ405mGqfk5ni318HqxsevbB1UP9QbPexUFFEkWZFGDo5ciKt27QZluu4rKMHVXqyKlhSq6EkdJKq9WPwF6vxmUA</recordid><startdate>2004</startdate><enddate>2004</enddate><creator>Nouza, Jan</creator><creator>Žďánský, Jindřich</creator><creator>David, Petr</creator><general>Springer Berlin Heidelberg</general><general>Springer</general><scope>IQODW</scope></search><sort><creationdate>2004</creationdate><title>Fully Automated Approach to Broadcast News Transcription in Czech Language</title><author>Nouza, Jan ; Žďánský, Jindřich ; David, Petr</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-p228t-bcb12db7e2a6fbfae3c3283078d07175730bce2fb31bf2f781ea2e7b86e0e76a3</frbrgroupid><rsrctype>conference_proceedings</rsrctype><prefilter>conference_proceedings</prefilter><language>eng</language><creationdate>2004</creationdate><topic>Acoustic Model</topic><topic>Applied sciences</topic><topic>Artificial intelligence</topic><topic>Automatic Speech Recognition</topic><topic>Bayesian Information Criterion</topic><topic>Computer science; control theory; systems</topic><topic>Exact sciences and technology</topic><topic>Speech and sound recognition and synthesis. Linguistics</topic><topic>Speech Recognition</topic><topic>Speech Signal</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Nouza, Jan</creatorcontrib><creatorcontrib>Žďánský, Jindřich</creatorcontrib><creatorcontrib>David, Petr</creatorcontrib><collection>Pascal-Francis</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Nouza, Jan</au><au>Žďánský, Jindřich</au><au>David, Petr</au><au>Pala, Karel</au><au>Sojka, Petr</au><au>Kopeček, Ivan</au><format>book</format><genre>proceeding</genre><ristype>CONF</ristype><atitle>Fully Automated Approach to Broadcast News Transcription in Czech Language</atitle><btitle>Lecture notes in computer science</btitle><date>2004</date><risdate>2004</risdate><spage>401</spage><epage>408</epage><pages>401-408</pages><issn>0302-9743</issn><eissn>1611-3349</eissn><isbn>9783540230496</isbn><isbn>3540230491</isbn><eisbn>9783540301202</eisbn><eisbn>3540301208</eisbn><abstract>In the paper we propose a complete scheme for automatic transcription of Czech TV news. The scheme first removes the music and noisy parts, then makes segmentation of the speech signal into speaker turns and consequently tries to decode and transcribe single utterances. We employ our own recognizer recently operating with a 200K-word lexicon and with a bigram language model. The overall recognition rate achieved on all the test data was 71.53%, that obtained on the read parts was 82.72%. The most serious recognition errors occur mainly in the segments that contain background music or extremely loud noise.</abstract><cop>Berlin, Heidelberg</cop><pub>Springer Berlin Heidelberg</pub><doi>10.1007/978-3-540-30120-2_51</doi><tpages>8</tpages></addata></record> |
| fulltext | fulltext |
| identifier | ISSN: 0302-9743 |
| ispartof | Lecture notes in computer science, 2004, p.401-408 |
| issn | 0302-9743 1611-3349 |
| language | eng |
| recordid | cdi_pascalfrancis_primary_16194695 |
| source | Springer Books |
| subjects | Acoustic Model Applied sciences Artificial intelligence Automatic Speech Recognition Bayesian Information Criterion Computer science control theory systems Exact sciences and technology Speech and sound recognition and synthesis. Linguistics Speech Recognition Speech Signal |
| title | Fully Automated Approach to Broadcast News Transcription in Czech Language |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-19T23%3A46%3A00IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-pascalfrancis_sprin&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=proceeding&rft.atitle=Fully%20Automated%20Approach%20to%20Broadcast%20News%20Transcription%20in%20Czech%20Language&rft.btitle=Lecture%20notes%20in%20computer%20science&rft.au=Nouza,%20Jan&rft.date=2004&rft.spage=401&rft.epage=408&rft.pages=401-408&rft.issn=0302-9743&rft.eissn=1611-3349&rft.isbn=9783540230496&rft.isbn_list=3540230491&rft_id=info:doi/10.1007/978-3-540-30120-2_51&rft_dat=%3Cpascalfrancis_sprin%3E16194695%3C/pascalfrancis_sprin%3E%3Curl%3E%3C/url%3E&rft.eisbn=9783540301202&rft.eisbn_list=3540301208&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true |