Eigenvoices for HMM-based speech synthesis
Eigenvoices for HMM-based speech synthesis
复制标题
DOI：
10.21437/icslp.2002-390
复制
发表时间：
2002-09
期刊：
影响因子：
--
通讯作者：
K. Shichiri;Atsushi Sawabe;Takayoshi Yoshimura;K. Tokuda;T. Masuko;Takao Kobayashi;T. Kitamura
中科院分区：
文献类型：
--
作者：
K. Shichiri;Atsushi Sawabe;Takayoshi Yoshimura;K. Tokuda;T. Masuko;Takao Kobayashi;T. Kitamura
本文介绍了一种用于基于HMM的语音合成系统的本征语音技术，该系统可以合成各种音质的语音。在本征语音技术中，其已经成功地应用于基于HMM的语音识别中的快速说话人自适应，大量的说话人相关HMM集合通过降维技术由几个参数表示，例如，PCA。本文提出了一种语音合成的本征语音技术，并将其应用于一个基于HMM的语音合成系统中，在该系统中，频谱和F0都由Hertons建模，合成语音由Hertons本身生成。给出了产生的频谱和F0图，并讨论了本征语音的权值与语音质量的关系。
This paper describes an eigenvoice technique for an HMM-based speech synthesis system which can synthesize speech with various voice qualities. In the eigenvoice technique, which has successfully been applied to fast speaker adaptation in an HMM based speech recognition, a large number of speaker dependent HMM sets are represented by a few parameters through a dimensionality reduction technique, e.g., PCA. In this paper, we propose an eigenvoice technique for speech synthesis, and apply it to an HMM-based speech synthesis system in which spectrum and F 0 are modeled by HMMs, and synthetic speech generated from HMMs themselves. The generated spectrum and F 0 pattern are shown, and the relation between weights for eigenvoices and voice quality is discussed.