The Research And Realization Of Corpus Based Speech Synthesis System For Uyghur

With the development of information society and computer science, the voice information service systems are widely spreading. All these systems require the support of speech synthesis technology. Multi-lingual speech synthesis is an interesting research subject in speech synthesis research field. This article has researched and implemented a corpus based speech synthesis system for Uyghur. Mostly has solved the following essential technology:1. Solved some key problems in the Uyghur text analysis, such as syllable separation, and summarize stress, pause rules of prosodic based on the features of the Uyghur language and phonetics.2. Implemented a waveform concatenation algorithm. In order to make the synthesized speech more natural, some experiments have been carried out, such as using different concatenation unit such as phone, syllable, word and sentence. The results demonstrated that using a larger unit could achieve higher naturalness and less concatenation.3. Based on the result of speech perception experiment, the effect of coarticulation on concatenation synthesis is studied and a multilayer and low redundancy corpus is designed.4. In order to reduce the distortion in concatenation points, a time domain smoothing method was presented to smoothing the synthesized speech. Experiment proved that this method can reduced some distortion in the concatenation point.5. A corpus based speech synthesis system for Uyghur is developed. This system can synthesize speech with high naturalness and maintain the speaking style of original speaker.