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

Subspace speech enhancement method based on constrained low rank and sparse decomposition

A speech enhancement and sparse decomposition technology, applied in speech analysis, instruments, etc., can solve problems such as singleness, inability to automatically determine the effective rank of a low-rank matrix, and inability to apply to an unsteady noise environment, achieving high accuracy and robustness Good results

Pending Publication Date: 2018-11-27
CHINESE AERONAUTICAL RADIO ELECTRONICS RES INST
View PDF4 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The disadvantage of this patent is that it belongs to the frequency domain processing method, and the working mode belongs to the offline working mode. It must wait for all the audio data to be ready to start processing, and cannot process audio data online in real time.
In addition, this method cannot automatically determine the effective rank of the low-rank matrix according to the audio environment, and can only use a single rank to decompose the spectrogram, so it cannot be applied to the non-stationary noise environment

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
  • Subspace speech enhancement method based on constrained low rank and sparse decomposition
  • Subspace speech enhancement method based on constrained low rank and sparse decomposition
  • Subspace speech enhancement method based on constrained low rank and sparse decomposition

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0037] The present invention will be further described in conjunction with accompanying drawing and embodiment now.

[0038] Such as figure 1 As shown, the subspace speech enhancement method based on constrained low-rank and sparse decomposition shown in this embodiment includes the following steps:

[0039] (1) Frame the noisy speech signal y(t).

[0040] (2) For each frame of noisy speech, construct a noisy matrix Y in the time domain:

[0041]

[0042] Wherein: band noise matrix Y is the Toeplitz matrix of ((N-M+1) * M) dimension, y () is the noisy speech, N is the sample number in a frame containing the noisy speech, M is adjustment factor, M= N / 3 is a better choice;

[0043] (3) Carry out SVD decomposition to the noisy matrix Y, that is, Y=U∑V H , U ∈ R (N-M+1)×M ,Σ M×M ,V M×(N-M+1) , according to the SVD decomposition results, the effective rank P of the pure speech is determined by the analysis-by-synthesis method, and the steps are as follows:

[0044] 3.1) ...

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 subspace speech enhancement method based on constrained low rank and sparse decomposition. The subspace speech enhancement method based on constrained low rank and sparse decomposition includes the following steps: (1) enframing noised speech; (2) constructing a noised matrix Y in a time domain for each frame of noised speech; (3) performing SVD (singular value hard) decomposition on Y, and using a synthetic analysis method to determine the effective rank P of pure speech according to the SVD decomposition result; (4) performing low rank and sparse matrix decomposition on Y, to obtain a low rank matrix L; (5) performing average operation on the diagonal line of the obtained L, and extracting an enhanced signal therefrom; and (6) for each frame, iteratively repeating the above steps to output a speech signal after the number of iterations are reached. The subspace speech enhancement method based on constrained low rank and sparse decomposition has the advantages of the subspace method and the low rank and sparse matrix decomposition: the data can be processed on line frame by frame, thus being able to be applied to a speech enhancement task for real-time processing; the effective rank of the low rank matrix can be automatically estimated according to the noise environment; and the subspace speech enhancement method can be applied to arbitrary noise environments, including steady state and unsteady noise.

Description

technical field [0001] The invention relates to the technical field of speech signal processing, and relates to a noise reduction method for noisy speech, in particular to a subspace speech noise reduction method based on constrained low-rank and sparse decomposition. Background technique [0002] Voice is one of the most effective, common and convenient means of communication. In recent years, with the continuous development of information science and technology, voice has become an important means of human-computer interaction. Intelligent voice processing technology has broad application prospects in many fields. But in the actual environment, speech signals are always accompanied by different degrees of noise interference. Even when the interference is severe, the speech signal will be completely submerged in the noise, resulting in a serious decline in auditory quality. Speech enhancement is an effective technique to solve noise pollution. Speech enhancement minimiz...

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): G10L21/0224G10L21/0272
CPCG10L21/0224G10L21/0272
Inventor 庄迁政李晓东原帅王文豪胡墨裘旭益岳新启孙成立
Owner CHINESE AERONAUTICAL RADIO ELECTRONICS RES INST
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