The Research Of Chinese Text-To-Speech Based On PSOLA

The successful applications in many fields of TTS, which aim to change interface between human and machine, have made more and more protenal value.As the effort of researchers, it has been put into effort and realize many things which only be descrilbed in fantansy previsouly.This paper reviews the history of TTS at the beginning, introduce its present suitation and prospect. then, analyze and compare the dis/advantage of present technolgys, choose the waveform concatenation as the guideline, also introduce its procedures. Base on the necessary knowledge of linguistics and the phonetics, the function of every parts to compose the TTS and the technology to accomplish these functions are introduced in this paper.Collect present rhythm regulations and describle them in quantity to control the rhythm parameters in speech synthesize parts. improve the present pitch extraction algorithm, realize the TD-PSOLA algorithm simply but effectively.The discrimination and the clearance of synthetic pronunciation is satisfied except the naturalness of synthetic pronunciation which still need to be improved obviously. The research of this paper paves the way for coming research that aim at a high naturation TTS eveturally.