Detecção automática de fronteiras prosódicas na fala espontânea

Detalhes bibliográficos
Ano de defesa: 2022
Autor(a) principal: Bárbara Helohá Falcão Teixeira
Orientador(a): Não Informado pela instituição
Banca de defesa: Não Informado pela instituição
Tipo de documento: Tese
Tipo de acesso: Acesso aberto
Idioma: por
Instituição de defesa: Universidade Federal de Minas Gerais
Brasil
FALE - FACULDADE DE LETRAS
Programa de Pós-Graduação em Estudos Linguísticos
UFMG
Programa de Pós-Graduação: Não Informado pela instituição
Departamento: Não Informado pela instituição
País: Não Informado pela instituição
Palavras-chave em Português:
Link de acesso: http://hdl.handle.net/1843/47273
https://orcid.org/0000-0002-4484-3590
Resumo: Speech is segmented into intonational units delimited by prosodic boundaries. This segmentation is claimed to have important consequences for syntax, information structure and cognition. This work aims to investigate the phonetic-acoustic parameters involved in the perception of prosodic boundaries, and to develop models for automatic detection of prosodic boundaries in Brazilian Portuguese male monological spontaneous speech. Two samples were segmented into intonational units by two groups of trained annotators. The boundaries perceived by the annotators were tagged as terminal or non-terminal. A script was used to extract phonetic-acoustic parameters along the speech signal in both a rightward and a leftward window around the boundary of each phonological word. The extracted parameters comprise measures of (1) Speech rate and rhythm; (2) Normalized duration; (3) Fundamental frequency; (4) Intensity; (5) Physical pause. The script considers as prosodic boundaries positions at which at least 50% of the annotators indicated a boundary of the same type. The models were developed from a Linear Discriminant Analysis algorithm and different heuristic training strategies were used. The models presented similar problems due to the importance of parameters related to physical pauses. A pause perception test was performed by a group of annotators. The duration of physical pause that can be perceived as a pause, avoiding confusion with the occlusive phase of the segments, was analyzed. In this work, the proposed duration is 100 ms. Additionally, submodels for identifying boundaries with and without a perceived pause were developed. The submodel for identifying boundaries with a perceived pause showed better results than the submodel for identifying boundaries without a perceived pause.