基于HMM的藏语拉萨话语音合成技术研究

Technology Research on Tibetan Lhasa Speech Synthesis Based on HMM

研究基于HMM的合成系统,将其原理应用到藏语拉萨话中,完成了基于HMM的藏语拉萨话的语音合成系统,文中就合成语料库的建设,上下文语境信息标注,问题集设计等数据准备工作进行了描述.此外,在HTS2.0的开源平台下,对藏语语音合成的训练模块和语音合成模块进行了阐述,由此构建藏语HMM语音合成系统,实现少数民族语言语音信号处理技术的多样化.

Based on HMM, the synthetic system was researched and the Tibetan Lhasa speech synthetic system has been completed. It has described detailly about synthesized speech corpus construction, context features labeling and question set designing, etc. for database preparation first. In addition, with the HTS. 2.0 Open - Source platform, it made comments on the training module and the speech synthetic module for Tibetan speech synthesis in order to build a Tibetan speech synthetic system based on HMM and to realize the diversification of the speech signal processing techniques for minority languages.