汉语语音合成的研究进展

Research Advance of Chinese Speech Synthesis

为解决如何让机器像人那样说话的问题,综述了近年来汉语语音合成技术的进步和发展;描述了汉语语音合成系统的设计方法;对基于汉语语音特点的语音合成过程中的一些关键技术和原理进行了详细的分析和比较;阐述了语音合成在多媒体等方面上的应用以及目前评价语音合成系统性能的一些新方法,并对当前语音合成技术的研究热点和语音合成技术存在的问题进行了分析,为语音合成技术的进一步研究提供了借鉴作用。

In order to solve the problem of making machine speak like human the development of current speech synthesis on Chinese is reviewed. The main methods of the modem design of the Chinese TI＇S （Text-To-Speech） system will be described. Then the key technologies and principles in Chinese pronunciation characteristic-based speech synthesis are discussed and analysed in detail. Lastly the applications of speech synthesis in multimedia are summarized and some new methods which appraise the TTS system at present are introduced. Moreover, the hotspots and the problems in the research on the speech synthesis technologies are to be analysed in order to provide reference for the future research.