Automatic speech synthesis apparatus and method

The present invention discloses an automatic speech synthesizing apparatus comprising: extracting means for automatically extracting data from the speech voice call; enhanced detection module for automatically distinguishing a plurality of persons voice of the voice data, after distinguishing classify speech data, and voice data dynamically updated classification; integrated tagging module for automatically labeling speech data after the classification, and to generate callouts; verification screening module for screening the speech meet a predetermined criteria data, and the synthesis model generated by the model corresponding to the transformation; and a speech synthesis means for synthesizing the respective model output according to the user's voice. Automatic speech synthesis apparatus according to the present invention includes provides short cycle, low cost and can automatically accumulate.