CORAA ASR: a large corpus of spontaneous and prepared speech manually validated for speech recognition in Brazilian Portuguese

Registro completo de metadados
MetadadosDescriçãoIdioma
Autor(es): dc.contributorFederal University of Technology — Paraná (UTFPR)-
Autor(es): dc.contributorUniversidade de São Paulo (USP)-
Autor(es): dc.contributorFederal University of Goias-
Autor(es): dc.contributorUniversidade Estadual Paulista (UNESP)-
Autor(es): dc.creatorCandido Junior, Arnaldo-
Autor(es): dc.creatorCasanova, Edresson-
Autor(es): dc.creatorSoares, Anderson-
Autor(es): dc.creatorde Oliveira, Frederico Santos-
Autor(es): dc.creatorOliveira, Lucas-
Autor(es): dc.creatorJunior, Ricardo Corso Fernandes-
Autor(es): dc.creatorda Silva, Daniel Peixoto Pinto-
Autor(es): dc.creatorFayet, Fernando Gorgulho-
Autor(es): dc.creatorCarlotto, Bruno Baldissera-
Autor(es): dc.creatorGris, Lucas Rafael Stefanel-
Autor(es): dc.creatorAluísio, Sandra Maria-
Data de aceite: dc.date.accessioned2025-08-21T15:47:39Z-
Data de disponibilização: dc.date.available2025-08-21T15:47:39Z-
Data de envio: dc.date.issued2023-07-29-
Data de envio: dc.date.issued2023-07-29-
Data de envio: dc.date.issued2021-12-31-
Fonte completa do material: dc.identifierhttp://dx.doi.org/10.1007/s10579-022-09621-4-
Fonte completa do material: dc.identifierhttp://hdl.handle.net/11449/246341-
Fonte: dc.identifier.urihttp://educapes.capes.gov.br/handle/11449/246341-
Descrição: dc.descriptionAutomatic Speech recognition (ASR) is a complex and challenging task. In recent years, there have been significant advances in the area. In particular, for the Brazilian Portuguese (BP) language, there were around 376 h publicly available for the ASR task until the second half of 2020. With the release of new datasets in early 2021, this number increased to 574 h. The existing resources, however, are composed of audios containing only read and prepared speech. There is a lack of datasets including spontaneous speech, which are essential in several ASR applications. This paper presents CORAA (Corpus of Annotated Audios) ASR with 290 h, a publicly available dataset for ASR in BP containing validated pairs of audio-transcription. CORAA ASR also contains European Portuguese audios (4.6 h). We also present a public ASR model based on Wav2Vec 2.0 XLSR-53, fine-tuned over CORAA ASR. Our model achieved a Word Error Rate (WER) of 24.18% on CORAA ASR test set and 20.08% on Common Voice test set. When measuring the Character Error Rate (CER), we obtained 11.02% and 6.34% for CORAA ASR and Common Voice, respectively. CORAA ASR corpora were assembled to both improve ASR models in BP with phenomena from spontaneous speech and motivate young researchers to start their studies on ASR for Portuguese. All the corpora are publicly available at https://github.com/nilc-nlp/CORAA under the CC BY-NC-ND 4.0 license.-
Descrição: dc.descriptionFederal University of Technology — Paraná (UTFPR)-
Descrição: dc.descriptionInstituto de Ciências Matemáticas e de Computação - University of São Paulo-
Descrição: dc.descriptionFederal University of Goias-
Descrição: dc.descriptionSão Paulo State University-
Descrição: dc.descriptionSão Paulo State University-
Idioma: dc.languageen-
Relação: dc.relationLanguage Resources and Evaluation-
???dc.source???: dc.sourceScopus-
Palavras-chave: dc.subjectAutomatic speech recognition-
Palavras-chave: dc.subjectBrazilian Portuguese-
Palavras-chave: dc.subjectPrepared speech-
Palavras-chave: dc.subjectPublic datasets-
Palavras-chave: dc.subjectPublic speech corpora-
Palavras-chave: dc.subjectSpontaneous speech-
Título: dc.titleCORAA ASR: a large corpus of spontaneous and prepared speech manually validated for speech recognition in Brazilian Portuguese-
Tipo de arquivo: dc.typelivro digital-
Aparece nas coleções:Repositório Institucional - Unesp

Não existem arquivos associados a este item.