论文部分内容阅读
在多方面实际应用目标推动下,受国家自然科学基金资助,开展了语音合成研究.我们的目标是开发一个高质量的汉语文语转换系统.为此建立了以微机为基础的语音分析合成系统,并组成了文语转换系统,进行了初步试验.这个系统充分利用汉语语音系统的特点,选择以无调音节和声调模式为基本合成单元.相应的语音参数从自然语流中提取,并赋予每个音节若干语音特征属性,为利用语音规则控制音节的声学参数提供了可能性.建立了供合成用的语音库和基本合成规则,利用语言学知识调整合成语言的韵律特征,使合成语言具有高可懂度和较好的自然度.语音合成利用硬件实现,文语转换可以实时进行.合成语音的发音速度可以在较大的范围内改变。
Under the motivation of many practical applications, the research of speech synthesis is supported by the National Natural Science Foundation of China and our goal is to develop a high-quality Chinese text-to-speech conversion system.Therefore, a microcomputer-based speech analysis and synthesis system , And made up of the language conversion system, conducted a preliminary test.This system takes full advantage of the characteristics of the Chinese phonetic system, select the non-toning and tone mode as the basic unit of synthesis.The corresponding speech parameters extracted from the natural language stream, and given A number of speech feature attributes of each syllable provide the possibility of using the speech rules to control the syllables’ acoustic parameters. The speech database for synthesis and the basic synthesis rules are established, and the prosodic features of the synthesized language are adjusted by linguistic knowledge so that the synthesized language has High intelligibility and good natural degree.With the hardware realization of the speech synthesis, the text-to-speech conversion can be carried out in real time.The pronunciation speed of the synthesized speech can be changed within a larger scope.