Speech synthesis method and device and computer readable storage medium

Abstract

The invention discloses a voice synthesis method and device and a computer readable storage medium, and the method comprises the steps: obtaining text information, inputting the obtained text information into an acoustic model based on an end-to-end neural network for coding, and generating a first content vector for summarizing the text information through coding; receiving voice duration information for each piece of sub-text information in the text information; according to the received voice duration information, adjusting the voice duration for the sub-text information in the first content vector, and generating a second content vector; and generating voice information corresponding to the text information according to the generated second content vector. Thus, in the speech synthesisprocess, the text duration in the first content vector is controlled, so that the generated sound becomes emotional, coherent and smooth on the basis that the tone quality is not reduced.

Description

technical field [0001] The present invention relates to the field of speech synthesis, in particular to a speech synthesis method, device and computer-readable storage medium. Background technique [0002] Speech synthesis refers to the technology that the computer automatically generates corresponding speech based on the text. The traditional speech synthesis technology is divided into parametric speech synthesis and splicing speech synthesis. [0003] The sound quality generated by parametric speech synthesis is good. The disadvantage is that the sound is flat, lacks emotion, and contains some background sounds. [0004] The voice generated by splicing speech synthesis has high sound quality. The disadvantage is that a large number of recorded sounds are required to meet the needs of different scenarios. In addition, the splicing algorithm often has the problem of splicing incoherence. Contents of the invention [0005] Embodiments of the present invention provide a spe...

AI Technical Summary

Problems solved by technology

[0003] The sound generated by parametric speech synthesis has good sound quality, but the disadvantage is that the sound is flat, lacks emotion, and contains some background sounds

[0004] The voice generated by splicing speech synthesis has high sound quality. The disadvantage is that a large number of recorded sounds are required to meet the needs of different scenarios. In addition, the splicing algorithm often has the problem of splicing incoherence.

Images