Unlock instant, AI-driven research and patent intelligence for your innovation.

Speech conversion model training method and device, speech conversion method and device and related equipment

A technology of speech conversion and training method, applied in speech analysis, speech recognition, instruments, etc., can solve the problems of low matching degree of text content and incomplete removal of speaker timbre.

Pending Publication Date: 2022-07-15
PING AN TECH (SHENZHEN) CO LTD
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] The embodiment of the present invention provides a speech conversion model training and speech conversion method, device, computer equipment and storage medium to solve the problems of incomplete removal of the original speaker's timbre and low matching degree of text content in the converted speech

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Speech conversion model training method and device, speech conversion method and device and related equipment
  • Speech conversion model training method and device, speech conversion method and device and related equipment
  • Speech conversion model training method and device, speech conversion method and device and related equipment

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0043] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0044] The training of the speech conversion model and the speech conversion method provided by this application can be applied in such as figure 1 In an application environment where the computer device can communicate with an external device, such as a server, through a network. Wherein, the computer equipment can be but not limited to various personal computers, notebook computers, smart phones, tablet computers and portable wearable devi...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a training method of a voice conversion model, which is applied to the field of artificial intelligence. A voice conversion model provided by the invention comprises a first neural network, a third neural network and a fourth neural network. The method provided by the invention comprises the following steps: calculating first loss of sample voice data passing through a gradient overturning layer after vectorization; calculating a second loss after the vectorization of the sample voice data and the vectorization of the corresponding text content data; after vectorizing the sample voice data and the text content data, splicing the vectorized sample voice data and the vectorized text content data into a voice implicit vector, and converting the voice implicit vector into target voice data; calculating a third loss of the target voice data and the sample voice data; adjusting parameters of a neural network in the voice conversion model according to the first loss, the second loss and the third loss; and repeating the above steps until the first loss is maximum and the second loss and the third loss are minimum, and obtaining a trained voice conversion model.

Description

technical field [0001] The present invention relates to the field of artificial intelligence, in particular to a method, device and related equipment for training a speech conversion model and speech conversion. Background technique [0002] Voice conversion is to decouple the timbre and content of the original speaker's voice and convert it to the timbre of the target speaker while keeping the content the same. The application scenarios include dubbing of film and television dramas, tone conversion in e-book reading to automatically match different story characters, etc. However, the converted voice obtained by the traditional technology, on the one hand, the converted voice still contains the timbre information of the original speaker, and on the other hand, the content of the converted voice does not match well with the text content of the original speaker's voice. SUMMARY OF THE INVENTION [0003] Embodiments of the present invention provide a voice conversion model t...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G10L21/007G10L21/013G10L15/02G10L15/06G10L25/03G10L25/30
CPCG10L21/007G10L21/013G10L15/063G10L15/02G10L25/30G10L25/03G10L2021/0135
Inventor 张旭龙王健宗程宁
Owner PING AN TECH (SHENZHEN) CO LTD