Abnormal sound event identification method based on MFCC+MP fusion characteristic

A recognition method and feature fusion technology, applied in character and pattern recognition, computer parts, speech analysis, etc., can solve the problems of good noise robustness, blind spots in video surveillance, etc., to enhance generalization ability, physical interpretability The effect of improving performance and robustness

Active Publication Date: 2019-05-21
GUILIN UNIV OF ELECTRONIC TECH
View PDF4 Cites 11 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

This method has good robustness to noise, and can effectively detect abnormal sounds in the sound signal in a low signal-to-noise ratio environment, solves the problem of blind spots in video surveillance, and provides favorable assistance for security work

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
  • Abnormal sound event identification method based on MFCC+MP fusion characteristic
  • Abnormal sound event identification method based on MFCC+MP fusion characteristic
  • Abnormal sound event identification method based on MFCC+MP fusion characteristic

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0029] refer to figure 1 , a method for identifying abnormal acoustic events based on MFCC+MP fusion features, comprising the following steps:

[0030] 1) Sound preprocessing for the first time: Carry out a series of digital processing to the sound signal in the sound library, so that the signal distribution is more stable and convenient for the extraction of subsequent sound features. The first sound preprocessing includes normalization processing, analysis Frame processing and windowing, such as figure 2As shown, the normalization processing is to normalize the collected sound signal to -1-1, which is convenient for the subsequent processing of the sound signal and the training of the neural network; the framing processing is to divide a section of sound signal into a group of short And the time frame of equal length, when the sampling frequency of the sound signal is 44.1KHz, take 1024 points as a frame, and there is overlap between two adjacent frames, which is called fr...

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 abnormal sound event identification method based on MFCC+MP fusion characteristics, which is characterized by comprising the following steps of: 1) pre-processing the soundfor the first time; 2) extracting the sound characteristic for the first time; 3) training a classifier; 4) inputting actual measurement sound; 5) pre-processing the sound for the second time; 6) extracting the characteristics for the second time; 7) applying a classifier; and 8) outputting a detection result. The method has good robustness to noise, can effectively detect abnormal sound existingin a sound signal under a low signal-to-noise ratio environment, solves the problem that blind areas exist in video monitoring, and provides beneficial help for security work.

Description

technical field [0001] The invention relates to the technical field of sound signal recognition, especially the detection and recognition of abnormal sounds such as gunshots, screams, and glass breaking, and is used for monitoring abnormal events in public places. Specifically, it is a method based on Mel cepstrum coefficient (Mel -frequency cepstral coefficient, referred to as MFCC) and matching pursuit (Matching pursuit, referred to as MP) fusion features of abnormal acoustic event recognition method. Background technique [0002] Human exploration of sound signal recognition began as early as the 1950s. In 1952, researchers from AT&T's Bell Laboratory realized a speech recognition system for isolated English digits for specific speakers. This system uses analog It is implemented by electronic devices, mainly extracting the formant information of vowels in digital pronunciation, and performing isolated digital recognition of specific people through a simple template matchi...

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): G10L25/30G10L25/24G10L25/51G06N3/04G06K9/62
Inventor 罗丽燕李芳足王玫仇洪冰宋浠瑜周陬覃泓铭韦金泉
Owner GUILIN UNIV OF ELECTRONIC 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