Eureka AIR delivers breakthrough ideas for toughest innovation challenges, trusted by R&D personnel around the world.

Endpoint detection method for reducing noise influence

A technology for endpoint detection and noise impact, used in speech analysis, speech recognition, instrumentation, etc.

Pending Publication Date: 2020-05-01
ZHEJIANG UNIV OF TECH
View PDF2 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] Aiming at the deficiencies of the prior art, the technical problem solved by the present invention is to reduce the impact of noise on the correct rate of endpoint detection, and improve the endpoint detection performance in a noisy 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
  • Endpoint detection method for reducing noise influence
  • Endpoint detection method for reducing noise influence
  • Endpoint detection method for reducing noise influence

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0077] The present invention will be described in detail below in conjunction with the accompanying drawings, so that the advantages and features of the present invention can be more easily understood by those skilled in the art, so as to define the protection scope of the present invention more clearly.

[0078] refer to Figure 1 to Figure 5 , an endpoint detection method that reduces the influence of noise, first uses the improved multi-window spectral estimation spectral subtraction method to reduce the noise of the noisy speech to obtain a higher output signal-to-noise ratio, and then uses the parameter-variable sub-band spectral entropy method to eliminate Each spectral line amplitude is affected by the noise, and the sub-band spectral entropy is used as a parameter to perform double-threshold endpoint detection. The detection method includes the following steps:

[0079] Step 1: Perform windowing and frame division on the noisy speech, perform discrete Fourier transform...

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 an endpoint detection method for reducing noise influence, and relates to front-end processing in the field of voice recognition; an improved multi-window spectral estimation spectral subtraction method is used for noise reduction of noisy voice to obtain a high output signal-to-noise ratio, then a sub-band spectral entropy method with variable parameters is used for eliminating the noise influence on each spectral line amplitude, and sub-band spectral entropy is used as a parameter for double-threshold endpoint detection. The method improves the end point detection misjudgment rate in a low signal-to-noise ratio environment, is quick and efficient, is higher in robustness, and has a certain application value in the fields of smart home, industrial control and medical treatment.

Description

technical field [0001] The invention relates to the field of speech signal analysis, in particular to an endpoint detection method in a speech recognition system. Background technique [0002] In the speech recognition system, endpoint detection has always been the focus of people's research, and high-precision endpoint detection helps to improve the recognition rate of the system. With the development of speech recognition technology, related products are becoming more and more abundant, and the fields of application are becoming wider and wider. In the actual application environment, there are often various noises, which affect the accuracy of endpoint detection to a certain extent, thereby reducing the accuracy of speech recognition. Therefore, the research on speech endpoint detection technology in noisy environment has always been a hot spot in speech signal processing. [0003] In speech signal processing, a speech signal generally includes a silent stage, a noisy st...

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): G10L15/26G10L21/0316G10L25/18G10L25/21G10L25/87
CPCG10L15/26G10L21/0316G10L25/18G10L25/21G10L25/87
Inventor 吴哲夫杭慧陶
Owner ZHEJIANG 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
Eureka Blog
Learn More
PatSnap group products