An introduction of trajectory model into HMM-based speech synthesis
An introduction of trajectory model into HMM-based speech synthesis
复制标题
DOI：
--
复制
发表时间：
2004
期刊：
影响因子：
--
通讯作者：
H. Zen;K. Tokuda;T. Kitamura
中科院分区：
文献类型：
--
作者：
H. Zen;K. Tokuda;T. Kitamura
在我们提出的基于隐马尔可夫模型(HMM)的语音合成系统的合成部分，通过使用语音参数生成算法，从对应于任意给定文本的句子HMM生成语音参数向量序列。然而，存在一个不一致的问题：虽然语音参数向量序列是在静态和动态特征之间的约束下产生的，但是HMM参数的训练没有任何它们之间的约束，这与标准HMM训练的方式相同。本文在基于HMM的语音合成系统的训练部分，引入了一种在静态和动态特征约束下由HMM得到的轨迹-HMM。实验结果表明，使用轨迹-HMM训练提高了合成语音的质量。
In the synthesis part of a hidden Markov model (HMM) based speech synthesis system which we have proposed, a speech parameter vector sequence is generated from a sentence HMM corresponding to an arbitrarily given text by using a speech parameter generation algorithm. However, there is an inconsistency: although the speech parameter vector sequence is generated under the constraints between static and dynamic features, HMM parameters are trained without any constraints between them in the same way as standard HMM training. In the present paper, we introduce a trajectory-HMM, which has been derived from the HMM under the constraints between static and dynamic features, into the training part of the HMM-based speech synthesis system. Experimental results show that the use of trajectory-HMM training improves the quality of the synthesized speech.