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

A method and system for converting speech into songs

A technology for converting into songs, applied in speech analysis, electroacoustic musical instruments, instruments, etc., can solve the problems of reducing user experience, song sound distortion, unnaturalness, etc., achieving remarkable efficiency and versatility, and improving the effect of sound quality

Active Publication Date: 2021-11-16
北京中科深智科技有限公司
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] However, the song sound generated by the song synthesis method in the prior art is distorted and unnatural, which greatly reduces the user experience.

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
  • A method and system for converting speech into songs
  • A method and system for converting speech into songs

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0023] The technical solutions of the present invention will be further described below with reference to the accompanying drawings.

[0024] The drawings are for exemplary description, which is merely a schematic diagram, rather than experiencing, is not understood to limit the present invention; in order to better illustrate the embodiments of the invention, there will be omitted, To zoom in, it does not represent the size of the actual product; it will be understood by those skilled in the <RTIgt;

[0025] Voice embodiment of the present invention to provide an embodiment of the method is converted into a song, such as figure 1 , It comprises the following:

[0026] Speech signal processing, and converted to a mel spectrum;

[0027] F0 contours extracted from different sound sources through the acoustic melody extractor;

[0028] The mel spectrum to a time stretching F0 contours of the same length, and of mel spectrum and F0 contours encoded by the two encoders, respectively;

...

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 method for converting speech into songs, which includes the following contents: processing speech signals and converting them into mel spectrograms; extracting F0 contours from different sound sources through a melody extractor; time-drawing the mel spectrograms Extend to the same length as the F0 profile, and encode the mel spectrogram and the F0 profile through two encoders; associate the encoded mel spectrogram with the F0 profile through a decoder, and generate a song spectrogram; through MelGAN The vocoder processes the song spectrogram to improve the sound quality of the output song. The invention also discloses a system for converting speech into songs. The invention can effectively improve the sound quality of songs and improve user experience.

Description

Technical field [0001] The present invention relates to speech signal processing technology, and particularly relates to a speech conversion system and method into a song. Background technique [0002] Currently, there are songs synthetic applications in entertainment, karaoke ok, music production. Songs synthesis under certain conditions, such as: lyrics, label or reference audio pitch, creating natural song. Wherein the reference audio may be a person singing passages, the task is to convert the voice to another person singing passages sound. The audio can also be a reference to a person of a voice, whose task is to convert it into a paragraph with the same singing tone identity and language content, without reference to its base sequence of phonemes. [0003] However, the song synthesis method of the prior art produced songs sound distortion, unnatural, greatly reducing the user experience. Inventive content [0004] Object of the present invention is to provide a method for ...

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 Patents(China)
IPC IPC(8): G10H1/00G10L21/007G10L25/24G10L25/03
CPCG10H1/0025G10H2210/101G10L21/007G10L25/03G10L25/24
Inventor 不公告发明人
Owner 北京中科深智科技有限公司