Transcribing and annotating speech corpora for speech recognition: A three-step crowdsourcing approach with quality control

Ver registo simples
dc.contributor.authorHämäläinen, A.
dc.contributor.authorMoreira, F. P.
dc.contributor.authorAvelar, J.
dc.contributor.authorBraga, D.
dc.contributor.authorDias, M. S.
dc.contributor.editorHartmann, B., and Horvitz, E.
dc.date.accessioned2023-02-13T15:31:41Z
dc.date.available2023-02-13T15:31:41Z
dc.date.issued2013
dc.date.updated2023-02-13T15:27:37Z
dc.description.abstractLarge speech corpora with word-level transcriptions annotated for noises and disfluent speech are necessary for training automatic speech recognisers. Crowdsourcing is a lower-cost, faster-turnaround, highly scalable alternative for expert transcription and annotation. In this paper, we showcase our three-step crowdsourcing approach motivated by the importance of accurate transcriptions and annotations.eng
dc.description.versioninfo:eu-repo/semantics/publishedVersion
dc.event.date2013
dc.event.locationPalm Springs, California, USAeng
dc.event.title1st AAAI Conference on Human Computation and Crowdsourcing, HCOMP 2013
dc.event.typeConferênciapt
dc.identifier.citationHämäläinen, A., Moreira, F. P., Avelar, J., Braga, D., & Dias, M. S. (2013). Transcribing and annotating speech corpora for speech recognition: A three-step crowdsourcing approach with quality control. In B. Hartmann, & E. Horvitz (Eds.), Proceedings of the 1st AAAI Conference on Human Computation and Crowdsourcing, HCOMP 2013 (vol. WS-13-18, pp. 30-31). AAAI Press. https://doi.org/10.1609/hcomp.v1i1.13102
dc.identifier.doi10.1609/hcomp.v1i1.13102
dc.identifier.isbn978-1-57735-607-3
dc.identifier.urihttp://hdl.handle.net/10071/27869
dc.language.isoeng
dc.pagination30 - 31
dc.peerreviewedyes
dc.publisherAAAI Press
dc.relation.ispartofProceedings of the 1st AAAI Conference on Human Computation and Crowdsourcing, HCOMP 2013
dc.rightsopen access
dc.subjectAutomatic speech recognitioneng
dc.subjectSpeech corporaeng
dc.subjectTranscriptioneng
dc.subjectAnnotationeng
dc.subjectCrowdsourcingeng
dc.subject.fosDomínio/Área Científica::Ciências Naturais::Ciências da Computação e da Informaçãopor
dc.subject.fosDomínio/Área Científica::Humanidades::Línguas e Literaturaspor
dc.titleTranscribing and annotating speech corpora for speech recognition: A three-step crowdsourcing approach with quality controleng
dc.typeconferenceObject
dc.volumeWS-13-18
dspace.entity.typePublicationen
iscte.alternateIdentifiers.scopus2-s2.0-84899503030
iscte.identifier.cienciahttps://ciencia.iscte-iul.pt/id/ci-pub-16895

Ficheiros

Pacote original

A mostrar 1 - 1 de 1
Nome:
conferenceobject_16895.pdf
Tamanho:
482.85 KB
Formato:
Adobe Portable Document Format
Descrição:
Versão Editora
Descarregar

Coleções

ISTAR-CRI - Comunicações a conferências internacionais