Automatic detection of disfluencies in a corpus of university lectures

This dissertation focuses on the identification of disfluent sequences and their distinct structural regions. Reported experiments are based on audio segmentation and prosodic features, calculated from a corpus of university lectures in European Portuguese, containing about 32 hours of speech and ab...

Full description

Bibliographic Details
Main Author: Medeiros, Henrique Rodrigues Barbosa de (author)
Format: masterThesis
Language:por
Published: 2015
Subjects:
Automatic disfluency detection
Spontaneous speech
University lectures
Machine learning
Speech processing
Prosodic features
Filled pauses
Statistical methods
Deteção automática de disfluências
Fala espontânea
Corpus de aulas universitárias
Aprendizagem automática
Processamento de fala
Features prosódicas
Pausas preenchidas
Métodos estatísticos
Online Access:http://hdl.handle.net/10071/8683
Country:Portugal
Oai:oai:repositorio.iscte-iul.pt:10071/8683

Internet

http://hdl.handle.net/10071/8683

Similar Items