Impact sound detection device and impact sound detection method

A detection device and technology for impact sound, applied in speech analysis, instruments, etc., can solve the problems of not considering the initial audio frame attribute of the audio segment, high false alarm rate, low impact sound detection performance, etc.

Inactive Publication Date: 2016-10-05
CANON KK
View PDF1 Cites 7 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, in the above-mentioned related technologies such as US Patent Application US2012 / 0185418, the recognition part only uses the features of the audio segments (such as the boundary information of the audio segments) obtained from the segmentation part, without considering the initial audio of each audio segment Frame Properties
Therefore, impact sound detection performance using existing sound detection methods is low; especially due to ambient noise, false alarm rate is high

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
  • Impact sound detection device and impact sound detection method
  • Impact sound detection device and impact sound detection method
  • Impact sound detection device and impact sound detection method

Examples

Experimental program
Comparison scheme
Effect test

no. 1 example

[0036] image 3 is a block diagram illustrating an exemplary functional configuration related to impact sound detection by the impact sound detection device 100 according to the first embodiment of the present invention. In a first embodiment, the present invention optimizes the result of impact sound recognition using the classification result of the onset point, where the onset point is regarded as the point with the largest change in energy in the corresponding audio segment. When the CPU 101 executes programs stored in the ROM 103 and / or the hard disk 104, the following functional units are realized.

[0037] Such as image 3 As shown, the input device 105 may receive, for example, from figure 1 The audio data output by the audio sensor 11 is shown.

[0038] The audio obtaining unit 301 obtains audio input from the input device 105 .

[0039] The preprocessing unit 302 first extracts at least one feature from the obtained audio output from the audio obtaining unit 301 ...

no. 2 example

[0105] Figure 8 is a block diagram illustrating an exemplary functional configuration related to impact sound detection by the impact sound detection device 100 according to the second embodiment of the present invention. In a second embodiment, the present invention uses the classification result of the starting point to verify the result of impact sound recognition. When the CPU 101 executes programs stored in the ROM 103 and / or the hard disk 104, the following functional units are realized.

[0106] Figure 8 and image 3 compared to Figure 8 The following main differences exist in the impact sound detection device 100 shown:

[0107] The start point detection unit 304 detects the start point of the audio segment only within the audio segment recognized as the first shock sound by the impact sound recognition unit 303 . That is to say, the starting point detection unit 304 can detect the starting point of the audio segment in at least a part of the audio segment, for...

no. 3 example

[0110] Figure 9 is a block diagram illustrating an exemplary functional configuration related to impact sound detection by the impact sound detection device 100 according to the third embodiment of the present invention. In the third embodiment, the present invention uses the classification result of the start point to select the audio segment to be recognized in the impact sound recognition operation, and the present invention can also use the detected position of the start point to optimize the audio segment to be recognized in the shock sound recognition operation. The starting position of the audio segment identified in the operation. When the CPU 101 executes programs stored in the ROM 103 and / or the hard disk 104, the following functional units are realized.

[0111] Figure 9 and image 3 compared to Figure 9 There are two main differences in the illustrated impact sound detection device 100:

[0112] First, the impact sound detection device 100 further includes ...

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 provides a shock sound detection device and a shock sound detection method. The impact sound detection device includes: a unit configured to obtain an audio input; a unit configured to extract at least one feature from the obtained audio, and based on the extracted feature, divide the obtained audio into at least one audio segment; A unit configured to identify a first impact sound from an audio segment based on a pre-generated sound model and extracted features; configured to detect a start point of the audio segment within at least a portion of the audio segment, and output a location and a start point of the start point an element of the energy-related characteristics of the onset; at least two types of elements configured to determine the onset based on the pre-generated onset model and the energy-related characteristics of the onset; and configured to determine the onset based on the identified impact sound and the determined onset The type of , which determines the unit of the second impact sound from the audio segment. According to the present invention, impact sound detection performance can be improved by using the information of the starting point.

Description

technical field [0001] The invention relates to sound detection, in particular to a shock sound detection device and a shock sound detection method. Background technique [0002] Now, sound detection technology is widely used in security monitoring, such as detecting audio anomalies in a given environment, detecting audio anomalies generated when a given device such as a multi-function printer (MFP) is working. [0003] Generally, a sound detection system includes two main parts, namely a segmentation part and a recognition part. First, in the segmentation part, the sound detection system segments the audio input into audio segments using any kind of Voice Endpoint Detection (VAD) algorithm familiar to those skilled in the art. Then, in the recognition part, the sound detection system recognizes the audio segments obtained from the segmentation part based on pre-generated sound models generated from the training data using any supervised and / or unsupervised means familiar t...

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): G10L25/78
Inventor 胡伟湘
Owner CANON KK
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