Speech synthesis method and system

Abstract

The invention discloses a voice synthesis method and system. The method includes steps: obtaining a recorded voice correspondingly recorded by a speaker according to a specific text; extracting fundamental frequency information f01 from the recorded voice, performing analysis processing on the recorded voice, and obtaining phoneme duration information corresponding to the recorded voice; generating fundamental frequency information f00 and frequency spectrum information cep0 by employing a preset parameter synthesis model according to the specific text and the obtained phoneme duration information; performing domain adjustment on the fundamental frequency information f01 of the recorded voice by employing the fundamental frequency information f01 of the recorded voice and the fundamental frequency information f00 generated by the preset parameter synthesis model; and obtaining final fundamental frequency information. According to the method and system, the corresponding synthesized voice is obtained through reconstruction of a vocoder by employing the obtained final fundamental frequency information and the frequency spectrum information cep0, the beneficial effect of reducing thevoice recording cost is achieved, and the natural degree of the synthesized voice is further improved.

Description

technical field [0001] The invention relates to the technical field of data processing, in particular to a speech synthesis method and system. Background technique [0002] The naturalness of the existing speech synthesis is not very ideal. In order to obtain a higher synthetic naturalness, some scenarios use the method of combining natural speech and synthetic speech for speech synthesis. For texts with high frequency, pre-recorded speech , for other text, use synthesized speech. Since the text with high frequency of occurrence will often find changes, it is usually necessary to find a speaker to re-record, which consumes a lot of manpower, material and financial resources, and the cost is relatively high; and the voice synthesized by this technology is not ideal for voice consistency. Contents of the invention [0003] The present invention provides a method and system for speech synthesis, aiming to directly use speakers of other pronunciation standards to record speec...

AI Technical Summary

Problems solved by technology

Because the text with high frequency of occurrence will often find changes, it is usually necessary to find a speaker to re-record, which consumes a lot of manpower, material and financial resources, and the cost is relatively high; and the voice synthesized by this technology has a poor voice consistency.

Images