Blind separation method for mobile source signal aliasing

A blind separation and source signal technology, applied in the field of blind signal separation, can solve the problems of general effect, difficult blind separation process of aliased signals, non-moving source signal, etc., to achieve good robustness, significant separation effect, and improve separation performance. Effect

Active Publication Date: 2019-08-09
GUANGDONG UNIV OF TECH
View PDF3 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, most of the research work is limited to the source signal is not moving, that is, the time-invariant aliasing signal, relatively little research on the time-varying aliasing signal
[0003] In the prior art, the document N.Q.K.Duong, E.Vincent, Under-determined reverberant audio source separation using a full-rank spatial covariance model, I...

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
  • Blind separation method for mobile source signal aliasing
  • Blind separation method for mobile source signal aliasing
  • Blind separation method for mobile source signal aliasing

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0097] In Embodiment 1, consider the aliasing situation where two microphones receive three moving source signals, the moving traces of the source signals are as follows image 3 As shown, the distance between the two microphones is 5 cm, and the reverberation time is 150 ms; in order to evaluate the performance of blind separation, short-term object intelligibility measurement (STOI) and frequency-weighted segmental signal-to-noise ratio (fwSegSNR) are selected As an evaluation criterion, the experimental results are as follows Figure 4 with Figure 5 As shown, it can be seen that the STOI and fwSegSNR values ​​obtained by the proposed algorithm are better than those obtained by the Full-rank algorithm, indicating that the proposed algorithm of the present invention has better robustness for blind separation of mobile source signal aliasing.

Embodiment 2

[0099] In the second embodiment, the aliasing situation in which three moving source signals are received by two microphones is also considered, and the moving space of the source signals is as follows: image 3 As shown, the distance between the two microphones is 5 cm, and the reverberation time is selected as 200 milliseconds; the experimental results are as follows Image 6 with Figure 7 As shown, from the obtained STOI value and fwSegSNR value, the separation result of the proposed algorithm is still better than the result obtained by the Full-rank algorithm; in addition, comparing the results of Example 1, it can be found that due to the increase of the reverberation time, the separation The result is a decline. Therefore, for a relatively low reverberation time, the algorithm proposed in the present invention has better separation performance for mobile source signal aliasing blind separation.

[0100] The present invention proposes a blind separation method for mobi...

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 blind separation method for mobile source signal aliasing. In real life, an acquired aliasing signal has certain variability, an aliasing filter is time-varying, and comparedwith the process of a traditional time-invariant aliasing filter, the separation process of a source signal becomes more difficult. Aiming at the aliasing condition of the time variation, a robust blind separation algorithm is provided. The method includes the steps of locating a source signal through a time difference estimation algorithm, and providing strict mathematical theory derivation; reconstructing the aliasing filter, updating model parameters by utilizing an expectation maximization algorithm, and separating a source signal by utilizing a Wiener filtering method; and finally, verifying the effectiveness of the proposed algorithm through a simulation experiment, and comparing with a robust mobile source signal aliasing blind separation algorithm (Full-rank algorithm) to prove that the algorithm provided by the invention has better robustness for mobile source signal aliasing blind separation.

Description

technical field [0001] The invention relates to the technical field of blind signal separation, in particular to a blind separation method for mobile source signal aliasing. Background technique [0002] In the classic cocktail party problem, the acquisition of aliased sounds is complicated by the movement of people at the party. How to recover the original signal only from the received aliased signal, while the aliased channel is unknown, this source signal separation process is called "blind separation". In recent years, blind separation technology has been fully applied in speech signal processing. However, most of the research work is limited to the immobile source signal, that is, the time-invariant aliasing signal, and there are relatively few studies on the time-varying aliasing signal. [0003] In the prior art, the document N.Q.K.Duong, E.Vincent, Under-determined reverberant audio source separation using a full-rank spatial covariance model, IEEETrans.Audio Speec...

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): G10L21/0216G10L21/0232G10L21/0272G10L25/69
CPCG10L21/0216G10L21/0232G10L21/0272G10L25/69G10L2021/02161
Inventor 解元谢胜利谢侃吴宗泽
Owner GUANGDONG UNIV OF TECH
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