Two-stage prosody prediction for emotional text-to-speech synthesis

Hao Tang; Xi Zhou; Matthias Odisio; Mark Hasegawa-Johnson; Thomas S. Huang

Two-stage prosody prediction for emotional text-to-speech synthesis

Hao Tang, Xi Zhou, Matthias Odisio, Mark Hasegawa-Johnson, Thomas S. Huang

Research output: Contribution to journal › Conference article › peer-review

Abstract

In this paper, we adopt a difference approach to prosody prediction for emotional text-to-speech synthesis, where the prosodic variations between emotional and neutral speech are decomposed into the global and local prosodic variations and predicted using a two-stage model. The global prosodic variations are modeled by the means and standard deviations of the prosodic parameters, while the local prosodic variations are modeled by the classification and regression tree (CART) and dynamic programming. The proposed two-stage prosody prediction model has been successfully implemented as a prosodic module in a Festival-MBROLA architecture based emotional text-to-speech synthesis system, which is able to synthesize highly intelligible, natural and expressive speech.

Original language	English (US)
Pages (from-to)	2138-2141
Number of pages	4
Journal	Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH
State	Published - 2008
Event	INTERSPEECH 2008 - 9th Annual Conference of the International Speech Communication Association - Brisbane, QLD, Australia Duration: Sep 22 2008 → Sep 26 2008

Keywords

CART
Dynamic programming
Prosody prediction
Speech synthesis
TTS

ASJC Scopus subject areas

Human-Computer Interaction
Signal Processing
Software
Sensory Systems

Library availability

Discover UIUC Full Text

Cite this

@article{33f207cd87c14c69bb2183d3e39e91fd,

title = "Two-stage prosody prediction for emotional text-to-speech synthesis",

abstract = "In this paper, we adopt a difference approach to prosody prediction for emotional text-to-speech synthesis, where the prosodic variations between emotional and neutral speech are decomposed into the global and local prosodic variations and predicted using a two-stage model. The global prosodic variations are modeled by the means and standard deviations of the prosodic parameters, while the local prosodic variations are modeled by the classification and regression tree (CART) and dynamic programming. The proposed two-stage prosody prediction model has been successfully implemented as a prosodic module in a Festival-MBROLA architecture based emotional text-to-speech synthesis system, which is able to synthesize highly intelligible, natural and expressive speech.",

keywords = "CART, Dynamic programming, Prosody prediction, Speech synthesis, TTS",

author = "Hao Tang and Xi Zhou and Matthias Odisio and Mark Hasegawa-Johnson and Huang, {Thomas S.}",

year = "2008",

language = "English (US)",

pages = "2138--2141",

journal = "Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH",

issn = "2308-457X",

note = "INTERSPEECH 2008 - 9th Annual Conference of the International Speech Communication Association ; Conference date: 22-09-2008 Through 26-09-2008",

}

TY - JOUR

T1 - Two-stage prosody prediction for emotional text-to-speech synthesis

AU - Tang, Hao

AU - Zhou, Xi

AU - Odisio, Matthias

AU - Hasegawa-Johnson, Mark

AU - Huang, Thomas S.

PY - 2008

Y1 - 2008

N2 - In this paper, we adopt a difference approach to prosody prediction for emotional text-to-speech synthesis, where the prosodic variations between emotional and neutral speech are decomposed into the global and local prosodic variations and predicted using a two-stage model. The global prosodic variations are modeled by the means and standard deviations of the prosodic parameters, while the local prosodic variations are modeled by the classification and regression tree (CART) and dynamic programming. The proposed two-stage prosody prediction model has been successfully implemented as a prosodic module in a Festival-MBROLA architecture based emotional text-to-speech synthesis system, which is able to synthesize highly intelligible, natural and expressive speech.

AB - In this paper, we adopt a difference approach to prosody prediction for emotional text-to-speech synthesis, where the prosodic variations between emotional and neutral speech are decomposed into the global and local prosodic variations and predicted using a two-stage model. The global prosodic variations are modeled by the means and standard deviations of the prosodic parameters, while the local prosodic variations are modeled by the classification and regression tree (CART) and dynamic programming. The proposed two-stage prosody prediction model has been successfully implemented as a prosodic module in a Festival-MBROLA architecture based emotional text-to-speech synthesis system, which is able to synthesize highly intelligible, natural and expressive speech.

KW - CART

KW - Dynamic programming

KW - Prosody prediction

KW - Speech synthesis

KW - TTS

UR - http://www.scopus.com/inward/record.url?scp=84867192290&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84867192290&partnerID=8YFLogxK

M3 - Conference article

AN - SCOPUS:84867192290

SN - 2308-457X

SP - 2138

EP - 2141

JO - Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH

JF - Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH

T2 - INTERSPEECH 2008 - 9th Annual Conference of the International Speech Communication Association

Y2 - 22 September 2008 through 26 September 2008

ER -

Two-stage prosody prediction for emotional text-to-speech synthesis

Abstract

Keywords

ASJC Scopus subject areas

Library availability

Related links

Fingerprint

Cite this