Voice denoising method and system based on L1/2 sparse constraint convolution non-negative matrix decomposition

A non-negative matrix decomposition, speech denoising technology, applied in speech analysis, instruments and other directions, can solve the problems of noise pollution, inaudible content and so on

Active Publication Date: 2016-09-21
ANHUI UNIVERSITY
View PDF4 Cites 26 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0002] Speech is an important carrier of daily communication, but it is often polluted by vari

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
  • Voice denoising method and system based on L1/2 sparse constraint convolution non-negative matrix decomposition
  • Voice denoising method and system based on L1/2 sparse constraint convolution non-negative matrix decomposition
  • Voice denoising method and system based on L1/2 sparse constraint convolution non-negative matrix decomposition

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0064] The present invention will be further described below in combination with specific embodiments and accompanying drawings.

[0065] The present invention is based on L 1 / 2 Sparse Constrained Convolutional Nonnegative Matrix Factorization (hereinafter referred to as "CNMF_L 1 / 2 ”) speech denoising method, figure 1 It is the general flowchart of the denoising of the present invention. The overall input is a certain type of noise and the speech after mixed noise, where the noise can be of different types (such as stationary noise, non-stationary noise, etc.); the output is the speech after denoising.

[0066] figure 2 is the flow chart of the noise training process in step 1.

[0067] Step 1.1, perform short-time Fourier transform (Short-Time Fourier Transform, STFT) transformation on the noise to obtain its amplitude spectrum N.

[0068] Step 1.2, perform CNMF decomposition on the noise amplitude spectrum to obtain the noise basis W n And its corresponding coefficie...

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 voice denoising method and system based on L1/2 sparse constraint convolution non-negative matrix decomposition. In single-channel voice enhancement, it is assumed that noised voice signals v(i) are additively relevant to noise signals n(i) and voice signals s(i), i.e., v(i)=n(i)+s(i), and noise-base information is obtained by training specific noise by use of a CNMF method; and then by taking a noise base as prior information, a voice base is obtained by decomposing noised voice by use of a CNMF_L1/2 method, and finally, voice after denoising is synthesized. According to the method, correlation of voice between frames can be better described; and strong-sparse constraining is performed on a voice-base coefficient matrix by use of L1/2 regular item, and the voice after separation comprises less residual noise. Compared to conventional methods such as a spectral subtraction method, a wiener filtering method and a minimum mean square deviation logarithm domain spectrum estimation method and the like, the voice after enhancement can be understood more easily.

Description

technical field [0001] The invention belongs to the field of acoustic signal processing, in particular to a 1 / 2 Speech denoising method and system for sparsely constrained convolutional nonnegative matrix factorization. Background technique [0002] Speech is an important carrier of daily communication, but it is often polluted by various noises in the real environment, making it difficult for people to hear the content clearly. Speech enhancement extracts as clean speech as possible from the polluted speech signal by suppressing or eliminating these noises, so as to obtain intelligible speech. Speech enhancement technology is often used in the fields of speech recognition, speech coding and intelligent communication. [0003] The speech enhancement method based on Non-negative matrix factorization (NMF) is a part-based expression method. By decomposing the speech signal, the basis vector and coefficient matrix representing speech characteristics can be obtained. At prese...

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
IPC IPC(8): G10L21/0272
Inventor 周健路成
Owner ANHUI UNIVERSITY
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap