Please use this identifier to cite or link to this item:https://hdl.handle.net/20.500.12259/101536
Full metadata record
DC FieldValueLanguage
dc.contributor.authorRaškinis, Gailius-
dc.contributor.authorPaškauskaitė, Gintarė-
dc.contributor.authorSaudargienė, Aušra-
dc.contributor.authorKazlauskienė, Asta-
dc.contributor.authorVaičiūnas, Airenas-
dc.coverage.spatialLT-
dc.date.accessioned2019-10-29T18:26:19Z-
dc.date.available2019-10-29T18:26:19Z-
dc.date.issued2019-
dc.identifier.issn08684952-
dc.identifier.otherVDU02-000061604-
dc.identifier.urihttps://www.mii.lt/informatica/pdf/INFO1233.pdf-
dc.identifier.urihttps://www.mii.lt/informatica/pdf/INFO1233.pdf-
dc.identifier.urihttp://dx.doi.org/10.15388/Informatica.2019.219-
dc.description.abstractConventional large vocabulary automatic speech recognition (ASR) systems require a mapping from words into sub-word units to generalize over the words that were absent in the training data and to enable the robust estimation of acoustic model parameters. This paper surveys the research done during the last 15 years on the topic of word to sub-word mappings for Lithuanian ASR systems. It also compares various phoneme and grapheme based mappings across a broad range of acoustic modelling techniques including monophone and triphone based Hidden Markov models (HMM), speaker adaptively trained HMMs, subspace gaussian mixture models (SGMM), feed-forward time delay neural network (TDNN), and state-of-the-art low frame rate bidirectional long short term memory (LFR BLSTM) recurrent deep neural network. Experimental comparisons are based on a 50-hour speech corpus. This paper shows that the best phone-based mapping significantly outperforms a grapheme-based mapping. It also shows that the lowest phone error rate of an ASR system is achieved by the phoneme-based lexicon that explicitly models syllable stress and represents diphthongs as single phonetic unitsen
dc.description.sponsorshipHumanitarinių mokslų fakultetas-
dc.description.sponsorshipInformatikos fakultetas-
dc.description.sponsorshipKauno technologijos universitetas-
dc.description.sponsorshipLietuvos sveikatos mokslų universitetas-
dc.description.sponsorshipLituanistikos katedra-
dc.description.sponsorshipTaikomosios informatikos katedra-
dc.description.sponsorshipVytauto Didžiojo universitetas-
dc.format.extentp. 573-593-
dc.language.isoen-
dc.relation.ispartofInformatica : an international journal. Vilnius : Vilnius university Institute of mathematics and informatics, 2019, vol. 30, no. 3-
dc.relation.isreferencedbyScience Citation Index Expanded (Web of Science)-
dc.relation.isreferencedbyScopus-
dc.relation.isreferencedbyINSPEC-
dc.relation.isreferencedbyAcademic OneFile (Gale)-
dc.relation.isreferencedbyZentralblatt MATH-
dc.relation.isreferencedbyCentral & Eastern European Academic Source (EBSCO)-
dc.subjectSpeech recognitionen
dc.subjectG2P conversionen
dc.subjectLithuanian languageen
dc.subjectGraphemeen
dc.subjectPhonemeen
dc.subject.otherInformatika / Informatics (N009)-
dc.titleComparison of phonemic and graphemic word to sub-word unit mappings for Lithuanian phone-level speech transcriptionen
dc.typeStraipsnis Clarivate Analytics Web of Science ar/ir Scopus / Article in Clarivate Analytics Web of Science or / and Scopus (S1)-
dc.identifier.doihttps://doi.org/10.15388/Informatica.2019.219-
dc.identifier.isiWOS:000486412200007-
dcterms.bibliographicCitation32-
dc.date.updated2019-10-29T16:10Z-
local.object{"source": {"code": "vdu", "handle": "61604"}, "publisher": {"name": "Vilnius university Institute of mathematics and informatics", "list": false}, "db": {"clarivate": true, "scopus": true, "list": true}, "issn": ["0868-4952"], "doi": "10.15388/Informatica.2019.219", "code": "S1", "subject": ["N009"], "url": ["https://www.mii.lt/informatica/pdf/INFO1233.pdf", "http://dx.doi.org/10.15388/Informatica.2019.219"], "country": "LT", "language": "en", "area": "N", "original": true, "pages": 21, "sheets": 1.5, "timestamp": "20191029161050.0", "account": {"year": 2019, "late": false}, "na": 5, "nip": 0, "affiliation": [{"contribution": 0.2, "aip": 1, "country": ["LT"], "rel": "aut", "org": [{"create": true, "contribution": 0.2, "name": "Vytauto Didžiojo universitetas", "id": "111950396", "level": "0", "type": "uni", "research": "1", "status": "1", "unit": {"name": "Informatikos fakultetas", "id": "04", "level": "2", "type": "fak", "research": "1", "status": "1", "unit": {"name": "Taikomosios informatikos katedra", "id": "0401", "level": "3", "type": "kat", "research": "1", "status": "0"}}}], "id": "BF06463EA15F567F20E41F7E5EB68D2A", "lname": "Raškinis", "fname": "Gailius", "status": "1", "name": "Raškinis, Gailius"}, {"contribution": 0.2, "aip": 1, "country": ["LT"], "rel": "aut", "org": [{"create": false, "contribution": 0.2, "name": "Kauno technologijos universitetas", "id": "111950581"}], "lname": "Paškauskaitė", "fname": "Gintarė", "status": "0", "name": "Paškauskaitė, Gintarė"}, {"contribution": 0.2, "aip": 2, "country": ["LT"], "rel": "aut", "org": [{"create": true, "contribution": 0.1, "name": "Vytauto Didžiojo universitetas", "id": "111950396", "level": "0", "type": "uni", "research": "1", "status": "1", "unit": {"name": "Informatikos fakultetas", "id": "04", "level": "2", "type": "fak", "research": "1", "status": "1", "unit": {"name": "Taikomosios informatikos katedra", "id": "0401", "level": "3", "type": "kat", "research": "1", "status": "1"}}}, {"create": false, "contribution": 0.1, "name": "Lietuvos sveikatos mokslų universitetas", "id": "30253989"}], "id": "CD95057360CE064DF5F4B48A9FD85439", "lname": "Saudargienė", "fname": "Aušra", "status": "1", "orcid": "0000-0003-2768-3334", "name": "Saudargienė, Aušra"}, {"contribution": 0.2, "aip": 1, "country": ["LT"], "rel": "aut", "org": [{"create": true, "contribution": 0.2, "name": "Vytauto Didžiojo universitetas", "id": "111950396", "level": "0", "type": "uni", "research": "1", "status": "1", "unit": {"name": "Humanitarinių mokslų fakultetas", "id": "03", "level": "2", "type": "fak", "research": "1", "status": "1", "unit": {"name": "Lituanistikos katedra", "id": "0309", "level": "3", "type": "kat", "research": "1", "status": "1"}}}], "id": "672EE6DF7E66BF6CDFA781E05CD8345C", "lname": "Kazlauskienė", "fname": "Asta", "status": "1", "orcid": "0000-0002-6135-1635", "name": "Kazlauskienė, Asta"}, {"contribution": 0.2, "aip": 1, "country": ["LT"], "rel": "aut", "org": [{"create": true, "contribution": 0.2, "name": "Vytauto Didžiojo universitetas", "id": "111950396", "level": "0", "type": "uni", "research": "1", "status": "1", "unit": {"name": "Informatikos fakultetas", "id": "04", "level": "2", "type": "fak", "research": "1", "status": "0", "unit": {"name": "Taikomosios informatikos katedra", "id": "0401", "level": "3", "type": "kat", "research": "1", "status": "0"}}}], "id": "BA04EA086704E121186E410A598CB44C", "lname": "Vaičiūnas", "fname": "Airenas", "status": "1", "name": "Vaičiūnas, Airenas"}]}-
local.typeS-
item.grantfulltextopen-
item.fulltextWith Fulltext-
crisitem.author.deptTaikomosios informatikos katedra-
crisitem.author.deptTaikomosios informatikos katedra-
crisitem.author.deptLituanistikos katedra-
crisitem.author.deptTaikomosios informatikos katedra-
Appears in Collections:Universiteto mokslo publikacijos / University Research Publications
Files in This Item:
marc.xml7.78 kBXMLView/Open

MARC21 XML metadata

Show simple item record

Google ScholarTM

Check

Altmetric


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.