Prosody Improvement for HMM-based Mandarin Speech Synthesis Using the Tone Nucleus Model

• Item type: SIG Technical Reports(1)
• 公開日: 2011-07-14
• タイトル: Prosody Improvement for HMM-based Mandarin Speech Synthesis Using the Tone Nucleus Model
• タイトル / 言語: en
• タイトル: Prosody Improvement for HMM-based Mandarin Speech Synthesis Using the Tone Nucleus Model
• 言語: eng
• キーワード / 主題Scheme: Other
• キーワード / 主題: 合成
• 資源タイプ / 資源タイプ識別子: http://purl.org/coar/resource_type/c_18gh
• 資源タイプ: technical report
• 著者所属: Department of Electrical Engineering and Information Systems, the University of Tokyo
• 著者所属: Department of Electrical Engineering and Information Systems, the University of Tokyo
• 著者所属: Department of Information and Communication Engineering, the University of Tokyo
• 著者所属: Department of Information and Communication Engineering, the University of Tokyo
• 著者所属(英): en
• 著者所属(英): Department of Electrical Engineering and Information Systems, the University of Tokyo
• 著者所属(英): en
• 著者所属(英): Department of Electrical Engineering and Information Systems, the University of Tokyo
• 著者所属(英): en
• 著者所属(英): Department of Information and Communication Engineering, the University of Tokyo
• 著者所属(英): en
• 著者所属(英): Department of Information and Communication Engineering, the University of Tokyo
• 著者名: Miaomiao, Wang; Miaomiao, Wen; Keikichi, Hirose; Nobuaki, Minematsu
• 著者名(英): Miaomiao, Wang; Miaomiao, Wen; Keikichi, Hirose; Nobuaki, Minematsu
• 論文抄録 / 内容記述タイプ: Other
• 論文抄録 / 内容記述: The HMM-based Text-to-Speech System has attracted great interest due to its compact and flexible modeling of spectral, F0 and duration parameters. The synthesized speech is highly dependent on the context model. However, the complex F0 variations make it rather difficult to define the tone type of Mandarin continuous speech. Then the F0 and duration trajectories, generated by HMM-based speech synthesis are often excessively smoothed and lack of prosodic variance. Tone nucleus of a syllable is assumed to be the target F0 of the associated lexical tone, and usually conforms more likely to the standard tone pattern. In this paper, by modeling F0 variations at different levels ranging from segmental factors to tone co-articulations, and apply the tone nucleus model to HMM-based Mandarin speech synthesis.
• 論文抄録(英) / 内容記述タイプ: Other
• 論文抄録(英) / 内容記述: The HMM-based Text-to-Speech System has attracted great interest due to its compact and flexible modeling of spectral, F0 and duration parameters. The synthesized speech is highly dependent on the context model. However, the complex F0 variations make it rather difficult to define the tone type of Mandarin continuous speech. Then the F0 and duration trajectories, generated by HMM-based speech synthesis are often excessively smoothed and lack of prosodic variance. Tone nucleus of a syllable is assumed to be the target F0 of the associated lexical tone, and usually conforms more likely to the standard tone pattern. In this paper, by modeling F0 variations at different levels ranging from segmental factors to tone co-articulations, and apply the tone nucleus model to HMM-based Mandarin speech synthesis.
• 書誌レコードID / 収録物識別子タイプ: NCID
• 書誌レコードID / 収録物識別子: AN10442647
• 書誌情報: 研究報告音声言語情報処理（SLP）; 巻 2011-SLP-87, 号 1, p. 1-6, 発行日 2011-07-14
• Notice: SIG Technical Reports are nonrefereed and hence may later appear in any journals, conferences, symposia, etc.
• 出版者 / 言語: ja
• 出版者: 情報処理学会