Presentation 2013-01-30
Improvement of context label in HMM-based speech synthesis for Japanese
Hiroya HASHIMOTO, Keikichi HIROSE, Nobuaki MINEMATSU,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) An improved set of context labels was proposed for HMM-based speech synthesis of Japanese. The conventional labels include those related to sentence length, such as number of mora and order of breath group. When handling sentences with various lengths, label numbers increase, which causes the "explosion" of label combinations. Furthermore, in Japanese, labels related to prosody are mostly designed based on the unit "accent phrase, " whose definition is somewhat unclear; it is not uniquely defined for a given sentence, but also is affected by other factors such as speaker identity, speaking rate, and utterance style. Therefore, reliable prediction of accent phrase boundaries for sentences included in the ti ainmg speech corpus comes difficult, leading to occasional mismatches in the predicted boundaries and prosodic features. In the proposed labels, "bunsetsu" is used instead as the basic unit for prosody. Also, we only view its relations with preceding and following "bunsetsu's." By doing so, the labels not related to the sentence lengths are obtained, with easier automatic prediction of labels only from sentence representations. Validity of the proposed labels was shown through a listening experiment of synthetic speech.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) HMM-based speech synthesis / context labels
Paper # SP2012-103
Date of Issue

Conference Information
Committee SP
Conference Date 2013/1/23(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Speech (SP)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Improvement of context label in HMM-based speech synthesis for Japanese
Sub Title (in English)
Keyword(1) HMM-based speech synthesis
Keyword(2) context labels
1st Author's Name Hiroya HASHIMOTO
1st Author's Affiliation The University of Tokyo()
2nd Author's Name Keikichi HIROSE
2nd Author's Affiliation The University of Tokyo
3rd Author's Name Nobuaki MINEMATSU
3rd Author's Affiliation The University of Tokyo
Date 2013-01-30
Paper # SP2012-103
Volume (vol) vol.112
Number (no) 422
Page pp.pp.-
#Pages 6
Date of Issue