Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Method for constructing voice idol

A technology for idols and voices, which is applied in the field of speakers with voice functions, and can solve problems such as the huge number of "fans" and the impossibility of idols

Inactive Publication Date: 2018-06-15
SHANGHAI UNIV
View PDF7 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, because the number of "fans" is often too large, it is impossible for idols to answer "fans" questions online in real time or answer fans' questions on Weibo in real time

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
  • Method for constructing voice idol
  • Method for constructing voice idol
  • Method for constructing voice idol

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0033] see figure 1 , this construction voice idol method is characterized in that the operation steps are as follows:

[0034] (1) Voice idol text extraction

[0035] a) A large collection of textual materials about idols;

[0036] b) Applying the LSTM neural network to the text material of the problem collected in step a to convert a large section of text into a text vector;

[0037] c) using the result of step b as the input of the RNN training model to train the style learning model;

[0038] d) Learn the idol's speaking style through training with a large amount of data.

[0039] (2) Voice idol voice synthesis:

[0040] e) Collect a large number of voice files about idols;

[0041] f) applying the bidirectional long-short-term memory prosodic hierarchy model to the voice file collected in step a to obtain the emotional speech synthesis model;

[0042] g) Using the result of Embodiment 1 as text input for speech synthesis, and using the emotional speech synthesis mod...

Embodiment 2

[0044] This embodiment is basically the same as Embodiment 1, and the features are as follows:

[0045] (1) The step a collects a large number of text materials about idols, and the text materials mainly come from the following ways:

[0046] i. The idol's public interviews or interview-type video materials are obtained through speech recognition technology and used as input for the style training model;

[0047] ii. The text of the idol's Weibo and the text of the reply to the fan's message are used as the input of the training model.

[0048] (2) In the step b, the text is used in the LSMT neural network to obtain a text vector, which is used in subsequent training steps.

[0049] (3) The step c uses the step b result as the input of the RNN training model to train the style learning model

[0050] (4) The step e collects a large number of voice files about idols; wherein the voice files mainly come from the public videos and voice files owned by idols.

[0051] (5) descr...

Embodiment 3

[0054] (1) see figure 2 , the idol text answer is obtained from the idol text question, and the operation steps are as follows:

[0055] a. Collect a large number of textual materials about idols;

[0056] b. For the text material collected in step a, apply the LSTM neural network to convert large sections of text into text vectors;

[0057] c. Using the result of step b as the input of the RNN training model to train the style learning model;

[0058] d. Learn the idol's speaking style through training with a large amount of data.

[0059] (2) see image 3 , the idol voice answer is obtained from the idol text answer:

[0060] e. Collect a large number of voice files about idols;

[0061] f. applying the bidirectional long-short-term memory prosodic hierarchical model to the voice file collected in step a to obtain the emotional speech synthesis model;

[0062] g. Using the result of Example 1 as text input for speech synthesis, and using the emotional speech synthesis...

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 relates to a method for constructing a voice idol. In the method, voice technologies of voice recognition, emotional voice synthesis and the like and a deep learning technology are usedto realize the voice idol so as to make voice responses to the questions of ''fans'' by using the same or similar style and timbre as the idol. The method comprises the following steps of a, collecting a lot of text materials about the idol; b, for the text materials collected in the step a, using a LSTM neural network to convert the large section of a text into a character vector; c, taking a result in the step b as the input of an RNN training model so as to train a style learning model; d, through a lot of data training, learning an idol speaking style; e, collecting a lot of voice files about the idol; f, for the voice files collected in the step e, using a bidirectional short and long time memory rhythm hierarchy model to acquire an emotion voice synthesis model; and g, using the result of an embodiment 3 as text input of voice synthesis and making the emotion voice synthesis model in the step f be used for voice synthesis.

Description

technical field [0001] The invention relates to a method for constructing a voice idol, and specifically relates to voice technologies such as voice recognition and semantic understanding, as well as technologies in deep learning, especially speaker technology with voice functions. Background technique [0002] Modern society has entered a highly developed society, and the material materials are already very sufficient. The material materials in today's society have greatly satisfied human beings' pursuit of material life. The pursuit of spiritual life of modern humans has gradually emerged and has reached very high requirements. [0003] In the process of pursuing spiritual life, many people have a sense of idol worship for film and television actors and stage performers. Out of admiration for idols, they are eager to get autographs and group photos of idols, hope that idols can see their messages and expect idols to reply to their messages. [0004] With the development ...

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): G10L13/02G10L13/04G10L15/06G10L15/16G10L15/18
CPCG10L13/00G10L13/02G10L15/06G10L15/16G10L15/18
Inventor 武星张南
Owner SHANGHAI UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com