Browsing by Author "Hung, Yu-Ping"

Now showing 1 - 2 of 2

A New Model-based Prosody Coder for Mandarin Speech
(2013)
Chiang, Chen-Yu
;
Hung, Yu-Ping
;
Chen, Sin-Horng
;
Wang, Yih-Ru
;
電機工程學系
In this paper, a novel parametric prosody coding approach for Mandarin speech is proposed. It employs a hierarchical prosodic model (HPM) as a prosody generating model in the encoder to analyze the speech prosody of the input utterance to obtain a parametric representation of four prosodic-acoustic features of syllable pitch contour, syllable duration, syllable energy level, and syllable-juncture pause duration for encoding. In the decoder, the four prosodic-acoustic features are reconstructed by a synthesis operation using the decoded HPM parameters. The reconstructed prosodic features are lastly used in an HMM-based speech synthesizer to help to generate the reconstructed speech. Experimental results show that the reconstructed speech has good quality at low data rates of 114.9 bits/s for a speaker-dependent task. An informal listening test confirmed decoded speeches sounded very fluently.
Improvements on Punctuation Generation Inspired Linguistic Features for Mandarin Prosody Generation
(2016-01-01)
Chiang, Chen-Yu
;
Hung, Yu-Ping
;
Liou, Guan-Ting
;
Wang, Yih-Ru
;
電機工程學系
This paper proposes two types of machine-extracted linguistic features from unlimited text input for Mandarin prosody generation. One is the improved punctuation confidence (iPC) which is a modified version of the previously proposed punctuation confidence that represents likelihood of inserting major punctuation marks (PMs) at word boundaries. Another is the quotation confidence (QC) which measures likelihood of a word string to be quoted as a meaningful or emphasized unit. Since major PMs are highly correlated with prosodic breaks, and a quoted Chinese word string plays an important role in human language understanding, the two features potentially could provide useful information for prosody generation. The idea is realized by employing conditional random field-based models to predict major PMs, quoted word string structures, and their associated confidences, i.e. iPC and QC. Then the predicted confidences are combined with traditional linguistic features to predict prosodic-acoustic features. Both objective and subjective tests showed that the prosody generation with the proposed linguistic features performed better than the ones without the proposed features.