Speech synthesis model training method and device, computer equipment and storage medium

Abstract

The invention relates to the technical field of artificial intelligence, and discloses a speech synthesis model training method and device, computer equipment and a storage medium. The method comprises the following steps: obtaining original text data and a speaker identifier, processing the original text data, obtaining an original text vector and a corresponding original phoneme vector, and carrying out feature enhancement processing on the original text vector and the original phoneme vector to obtain a target text vector and a target tone vector which are more significant in vectors; based on the speaker identifier, obtaining a corresponding target voiceprint vector, splicing the target voiceprint vector, the target text vector and the target tone vector, and training a speech synthesis model by using the spliced target hidden vector to obtain a target speech synthesis model corresponding to the speaker identifier, therefore, the similarity between the voice data synthesized by the updated target voice synthesis model and the voiceprint of the speaker is improved, and the personification degree of the target voice synthesis model is increased.

Description

technical field [0001] The present invention relates to the technical field of artificial intelligence, and in particular, to a speech synthesis model training method, device, computer equipment and storage medium. Background technique [0002] Speech synthesis is to convert the input text sequence into the corresponding natural speech pronunciation, which is an important speech processing task in the process of human-computer interaction. In recent years, speech synthesis technology based on deep neural network has achieved remarkable synthesis effect. With the rapid development of artificial intelligence industry, speech synthesis system has also been widely used. In addition to intelligibility, the requirements for the naturalness, rhythm and sound quality of speech synthesis are also getting higher and higher. [0003] Using deep models for speech synthesis requires consideration of text and corresponding speech. Usually, a large amount of training data is required in t...

AI Technical Summary

Problems solved by technology

[0004] The embodiment of the present invention provides a speech synthesis model training method, device, computer equipment and storage medium, which solves the problem that the speech generated by the existing speech synthesis model has a low degree of anthropomorphism

Images