Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
SPEECH SEPARATION METHOD AND DEVICE, MEDIUM AND ELECTRONIC DEVICE
Document Type and Number:
WIPO Patent Application WO/2021/000385
Kind Code:
A1
Abstract:
Provided are a speech separation method and a device, a medium and an electronic device. The method comprises: S300, acquiring a time-frequency domain mixed signal to be processed; S301, obtaining a separation matrix of the time-frequency domain mixed signal according to the time-frequency domain mixed signal and a cost function of independent vector analysis of a Gaussian mixture model comprising an amplitude adjustment factor; and S302, obtaining a time-frequency domain signal of each sound source from the time-frequency domain mixed signal according to the separation matrix. The method is beneficial to improving the implementability of speech separation technology, thereby broadening the scope of application of the speech separation technology.

Inventors:
GU, Zhaoyi (CN)
LU, Jing (CN)
Application Number:
CN2019/100986
Publication Date:
January 07, 2021
Filing Date:
August 16, 2019
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
NANJING INSTITUTE OF ADVANCED ARTIFICIAL INTELLIGENCE, LTD. (CN)
International Classes:
G10L21/0272
Attorney, Agent or Firm:
NANJING XINHUIHENGCHENG INTELLECTUAL PROPERTY AGENT CO., LTD (CN)
Download PDF: