A data processing method, device and electronic device

Abstract

Embodiments of the present invention provide a data processing method, device, and electronic device, wherein the method includes: determining a text to be broadcast and a broadcast user who broadcasts the text to be broadcast; a text feature, a first voice expressive feature, and a first voiceprint feature of the broadcasting user; synthesizing the to-be-broadcast text according to the first text feature, the first voice expressive feature, and the first voiceprint feature The corresponding voice to be broadcast; compared with the training model to extract and broadcast the user's acoustic features (such as including voice expressive features, voiceprint features, text features, etc.), the data used to determine the voiceprint characteristics of the broadcast user is less, Therefore, the embodiment of the present invention can improve the efficiency of speech synthesis.

Description

technical field [0001] The present invention relates to the technical field of data processing, and in particular, to a data processing method, device and electronic device. Background technique [0002] With the continuous development of terminal technology, the functions provided in the terminal are also various, such as making calls, viewing / sending short messages, watching videos, and voice broadcasting. [0003] Usually, the terminal provides a variety of voices for voice broadcast, such as male voice, female voice, child voice, etc., to improve the interest of voice broadcast. Among them, in order to better meet the user's needs, the terminal can also customize the voice for the user, and then use the customized voice to broadcast the voice. [0004] However, in the process of customizing the user's voice, it is necessary to collect a large amount of the user's voice data, and then use these voice data to train the model, so that the trained model can extract the user...

AI Technical Summary

Problems solved by technology

[0004] However, in the process of customizing the voice for a user, it is necessary to collect a large amount of voice data of the user, and then use these voice data to train the model, so that the trained model can extract the user's acoustic features (such as voice expressiveness features, acoustic features, etc.). pattern features, etc.); which in turn leads to low efficiency of synthesizing speech based on the acoustic features extracted by the trained model

Images