基于语音数据库的文语转换系统过程分析  

Process Analysis of Text-to-speech System Based on Speech Database

作　　者：周开来[1] 

机构地区：[1]西南林业大学计算机与信息科学系,云南昆明650224

出　　处：《计算机时代》2010年第7期7-9,共3页Computer Era

摘　　要：文语转换是中文信息处理中研究的热点,是实现人机语音通信的一项关键技术。文章对实现中文文语转换的整个过程进行了初步分析和研究,给出了基于语音数据库的文语转换方法和实现过程。具体介绍了语音库的建立,分析了文本录入、文本分词、文本正则化、语音标注、韵律处理和语音合成等各个环节处理的内容及技术难点。Text to speech is a hotspot in the study of Chinese language information processing and a key technology to implement human-machine communication by speech. The whole process of Chinese language text to speech is preliminarily analyzed and studied,and the transformation method and realization process of text to speech based on speech database are put forward. The creation of speech database is concretely introduced,and the content and technical difficult points of some stages,such as text inputting,text words dividing,text normalizing,pronunciation labeling,prosody handling and speech synthesizing are analyzed.

关 键 词：文语转换 语音数据库 过程分析 语音合成 

分 类 号：TN912.33[电子电信—通信与信息系统]