Please use this identifier to cite or link to this item:
Title: Generation of prosody and speech for Mandarin Chinese
Keywords: text-to-speech, prosody, speech synthesis, unit selection, prosodic break, Chinese TTS
Issue Date: 19-Feb-2004
Citation: DONG MINGHUI (2004-02-19). Generation of prosody and speech for Mandarin Chinese. ScholarBank@NUS Repository.
Abstract: This research investigates the use of prosody in unit selection based Chinese text-to-speech system. It studies the problem of correctly representing perceptual effects and implementing them in generated speech. Especially, it is focused on break and tone in Chinese. The work includes the following main tasks: (1) The work studies the prediction of prosodic breaks, especially the prediction of prosodic word break. The factors that affect the performance of prediction are examined. A dependency model for break prediction is developed. (2) The problem of prosody parameters is studied. An approach is given to evaluate and determine a set of prosody parameters for unit selection based synthesis. The relationships between the parameters and the features for prediction are also investigated. (3) The prosody parameters are applied into unit selection to help generate speech. The experiments show that the determined prosody helps to improve the speech quality.
Appears in Collections:Ph.D Theses (Open)

Show full item record
Files in This Item:
File Description SizeFormatAccess SettingsVersion 
DongMH.pdf3.21 MBAdobe PDF



Page view(s)

checked on Dec 30, 2018


checked on Dec 30, 2018

Google ScholarTM


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