Research And Implementation Of Individual Speech Synthesis

With the rapid development of computer communication and multi-media technology, the technology of new human machine interaction has become a hot spot in the computer science field at present. Speech synthesis and speech recognition technology are the two key technologies to implement human-machine communication by speech, to establish a oral system which has the ability of listening and talking.Now speech synthesis technology is more and more mature, and the various speech synthesis products have already come out. It is a hotspot that how to make the speech synthesis system more natural and more expressive forces. There are some speech synthesis research on different speech style and different emotion. But it is rare that the speech synthesis research on individual characteristics.Individual speech synthesis is based on the studying of speech analysis, speech synthesis and speech recognition technology. At first, this paper analyzes the reason of speech individual characteristics and bring forward the methods of establishing speech signal models according to the process of speech form. It weightily study on the parameters effecting speech characteristics and their extracting methods and adjusting methods. In the end, the author attempts to synthesize speech through MS TTS , then to add controlling acoustic parameters on the synthesized speech , to let speech synthesis have more expressive forces, meanwhile the author test and appraise the synthesized speech . The experiments indicate that we can let synthesized speech has the speaker's individual characteristics in this way, it needn't record the speaker's speech database before we synthesize speech.