Unlock instant, AI-driven research and patent intelligence for your innovation.

Anti-noise speech recognition method and device based on signal-to-noise ratio weighted template feature matching

A technology of feature matching and weighted templates, applied in speech recognition, speech analysis, instruments, etc., can solve the problems of poor recognition performance and high robustness, and achieve the effects of wide adaptability, improved accuracy, and wide application fields

Inactive Publication Date: 2017-02-15
SOUTH CHINA UNIV OF TECH
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] However, whether it is MFCC or LPCC, the existing speech recognition features are not very good in the recognition performance of the low signal-to-noise ratio environment. In order to overcome this weakness, the present invention firstly proposes an A new feature with better robustness in the case of noise ratio, that is, the angle between two time-delayed signal vectors is used as a correlation measure, because the angle is a nonlinear transformation of the traditional autocorrelation coefficient product, and the phase can be used in Enhance the role of the peak in the frequency domain, and the peak is more robust to noise

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
  • Anti-noise speech recognition method and device based on signal-to-noise ratio weighted template feature matching
  • Anti-noise speech recognition method and device based on signal-to-noise ratio weighted template feature matching
  • Anti-noise speech recognition method and device based on signal-to-noise ratio weighted template feature matching

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0066] Such as figure 1 As shown, the test speech first enters the preprocessing module, then enters the feature extraction module, and obtains the corresponding feature vector of the test speech. MFCC and PAC-MFCC are input to the template matching module, and the template in the reference database is matched by calculating the weight distance value (specifically The template matching module process is as follows image 3 As shown), the matching template with the smallest weight distance value is obtained, and finally the result is output to the display module.

[0067] Among them, the preprocessing process and feature extraction process are as follows: figure 2 As shown, pre-emphasis, digitization, framing, and windowing are carried out in the preprocessing process, and then the test speech frame features are extracted in the feature extraction process, and the FFT transformation is performed by calculating the autocorrelation coefficient and phase coefficient, and then pa...

Embodiment 2

[0080] This embodiment is the same as Embodiment 1 except for the following:

[0081] The formula for calculating the corresponding weight is as follows:

[0082]

[0083] Take relevant parameters: β=3, θ=0.5.

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 anti-noise voice identification method and device based on signal-to-noise ratio weighing template characteristic matching. The anti-noise voice identification method based on signal-to-noise ratio weighing template characteristic matching comprises the following steps that (1) input voice signals are preprocessed, and a phase position coefficient is obtained; (2) the characteristics of input voice, namely a phase position MFCC, are calculated; (3) characteristic matching is carried out on a template based on SNR. The invention further discloses a device of the anti-noise voice identification method based on signal-to-noise ratio weighing template characteristic matching. The device comprises a power source module, a display module, a storage module, a DSP / ARM digital processing module, a microphone, an A / D converter and a USB interface. The anti-noise voice identification method and device based on signal-to-noise ratio weighing template characteristic matching have the advantages of being wide in application range, high in accuracy, low in cost, convenient and fast to use, high in adaptability and the like.

Description

technical field [0001] The invention relates to a sound signal processing technology, in particular to an anti-noise speech recognition method and device based on signal-to-noise ratio weighted template feature matching. Background technique [0002] Speech recognition has a wide range of applications, almost involving every aspect of daily life. Such as voice dialing system, booking system, medical service, banking service, dictation machine, computer control, industrial control, voice communication system, etc. Speech recognition technology has profoundly changed human beings' existing daily life in various fields such as industry, home appliances, communications, medical care, and home services. Nowadays, the actual environment has higher and higher requirements for the robustness of acoustic noise in speech recognition. Therefore, it is of great significance to extract feature vectors with robustness and strong distinguishing ability for speech recognition systems. [...

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 Patents(China)
IPC IPC(8): G10L15/02G10L19/03G10L19/04
Inventor 宁更新吴丽菲宁小娟
Owner SOUTH CHINA UNIV OF TECH