摘要
汉语多音字的情况为中文文语转换TTS(Text To Speech)系统的建立带来了很大的困难。针对中文文语转换系统中的多音字问题,通过构建多音字词库和非多音字词库,将多音字以词汇的形式区分,并且通过对多音字词库和非多音字词库的结构形式的构造,减少了词库的冗余信息,提高了词汇语音的查找速率。实验证明该方案可以解决中文TTS中的多音字问题。
The Chinese polyphone makes it rather difficult to build the Chinese Text-To-Speech system. In the light of the polyphone problem in the Chinese TTS system,a solution to the problem is presented. The word libraries are built, and polyphone is distinguished through the words. Through the construction of the structure of the word libraries, the redundant information is reduced, and the searching speed of words is accelerated. The experimental result presents that the solution can settle the polyphone problem in the Chinese TTS system.
出处
《计算机应用与软件》
CSCD
北大核心
2008年第2期143-145,共3页
Computer Applications and Software
关键词
文语转换
多音字
语音合成
Text to speech
Polyphone
Speech synthesis
