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

Real-time convolutive mixed blind signal separation adaptive step length method based on fuzzy system

A technology of blind signal separation and self-adaptive step size, which is applied in speech analysis and instrumentation, and can solve problems such as slow convergence speed and large amount of calculation

Inactive Publication Date: 2016-08-03
TIANJIN UNIV
View PDF2 Cites 8 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] Since the convolution order of the sound signal is usually thousands of orders, the calculation amount of the time-domain deconvolution mixed busy signal problem usually becomes very large, and eventually the convergence speed is very slow

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
  • Real-time convolutive mixed blind signal separation adaptive step length method based on fuzzy system
  • Real-time convolutive mixed blind signal separation adaptive step length method based on fuzzy system
  • Real-time convolutive mixed blind signal separation adaptive step length method based on fuzzy system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0055] The purpose of the present invention is to provide a fuzzy system-based method for adaptively selecting a step size for a real-time convolution hybrid blind signal separation algorithm. For an independent vector analysis method for real-time convolutional hybrid blind signal separation, the choice of step size represents a trade-off between convergence speed and steady-state performance. A larger learning rate can obtain a relatively faster convergence speed, but there will also be large fluctuations in the steady state, which will affect the performance of the algorithm. Although a smaller learning rate has better performance in the steady state, it requires too much data to train to the steady state. In addition, in practical applications, due to possible changes in the environment, the mixing method of the system may change during the whole process, and the separation filter will also change accordingly. Therefore, it is necessary to ensure that the learning rate ca...

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 relates to voice signal processing, and aims to use a fuzzy system to determine the learning rate, ensure the convergence speed of an algorithm in the early stage and the steady state performance in the late stage and achieve a good real-time separation effect. The technical scheme adopted in the invention is that the real-time convolutive mixed blind signal separation adaptive step length method based on a fuzzy system comprises the following steps: (1) switching input signals from a time domain to a frequency domain through short-time Fourier transformation; (2) inputting signals, and sequentially processing the signals input to the frequency domain points at the current moment; (3) building a fuzzy system; (4) calculating a separation matrix at the current moment according to the current step length; (5) solving a frequency domain separation signal based on the separation matrix; and (6) switching the frequency domain separation signal to the time domain through short-time Fourier inverse transformation to get a final separation signal. The method is mainly applied to the occasion of voice signal processing.

Description

technical field [0001] The present invention relates to speech signal processing. Convolutional hybrid blind signal separation is a separation method for many practical application scenarios, such as the separation of mixed sound signals. The adaptive step size method of real-time convolution hybrid blind signal separation can effectively select the step size, and obtain accurate separated signals as soon as possible under the premise of ensuring system stability, which plays an important role in the processing of speech signals. Specifically, it involves an adaptive step size method based on fuzzy system real-time convolution hybrid blind signal separation. Background technique [0002] Blind signal separation refers to recovering the original signal or source only based on the observed mixed data vector, while the mixing matrix is ​​unknown in this process. It has broad application prospects in the fields of biomedical signal processing, image processing and speech recogn...

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): G10L21/0272
CPCG10L21/0272
Inventor 张立毅王哲陈雷李锵
Owner TIANJIN UNIV
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