Research On Technology Of Individual Characteristics Speech Synthesis And Simulation

To make computer can "speak" is an important aspect of human-computer interaction. With the rapid development of the computer communication technology and multimedia technology, computers have been able to speak clearly and naturally aided by speech synthesis systems. However, the existing speech synthesis systems often speak only in accordance with one or several specific patterns, the user not only requires that the computer can listen and speak, but also hope the output voice be able more personalized. The individuality characteristic speech synthesis and the simulation is an important part of the speech synthesis research, it helps to broaden the application scope of speech synthesis, riches expressive force of machine speech synthesis.This paper has mainly completed the following task:(1) By studying Chinese and foreign literature, understood present research situation of individual characteristics speech synthesis. (2) From the pronunciation forming process, analyzed the individuality characteristic physiological reason, summarized the pronunciation various acoustics parameter and established the voice signal physical model and the numerical model. (3) Done extraction and adjustment experiments of individual characteristic acoustics parameters. (4) Implemented individual characteristics speech synthesis based on the MS TTS, and evaluated the speech synthesis system. The experiments show that the speech synthesis may have a speaker's individual characteristics in this way, it need not prerecord the speaker's corpus.Based on the research of speech synthesis and speech analysis techniques, this paper studied the individual characteristics parameters that can reflect the speaker's voice, completed extraction adjustment analysis of the parameters, and through secondary development VB programming SAPI 5.1, these parameters were used to control the TTS voice, this method can not add new voice library terms of expressive, rich voice adds personalized color, reduces the complexity of synthetic speech, the practical effect is good. It can be applied to TTS, movie dubbing and voice camouflage. Because the system is small and cheap, it has the higher promoted value.