Method for detecting voice fraud replay attack by voice amplitude information and multiple phases

A technology of phase detection and replay attack, which is applied in speech analysis, instruments, etc., can solve problems such as ignoring phase information, and achieve the effects of easy implementation, improved results, and improved experimental results

Inactive Publication Date: 2019-06-04
慧言科技(天津)有限公司
View PDF2 Cites 8 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] Aiming at the problem of ignoring the phase information in the existing replay attack detection method, the present invention provides a method of fraudulent replay attack using speech amplitude information and various phases to detect speech, combined with a feature extraction method of amplitude and phase information, using Gaussian Mixture Model (GMM) to detect differences between real speech and replayed recordings

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
  • Method for detecting voice fraud replay attack by voice amplitude information and multiple phases
  • Method for detecting voice fraud replay attack by voice amplitude information and multiple phases
  • Method for detecting voice fraud replay attack by voice amplitude information and multiple phases

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0053] The present invention will be further described in detail below in conjunction with specific embodiments, so as to better reflect the advantages of the present invention.

[0054] A method of using voice amplitude information and multiple phase detection voice fraud replay attacks, mainly comprising the following steps:

[0055] S1: Amplitude feature and phase feature extraction:

[0056] The amplitude feature is a CQCC feature, and the phase feature is a Mel-RP feature and an MPVT feature; wherein, the extraction of the amplitude feature is to convert the geometric space of the frequency bin into a linear space by using constant Q transform; the extraction of the phase feature mainly The phase spectrum is obtained through pre-emphasis, framing, windowing, and short-time Fourier transform, and then processed on this basis;

[0057] 1) CQCC feature extraction

[0058] The CQCC feature is a kind of amplitude feature. By using the constant Q transformation, the geometric...

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 a method for detecting voice fraud replay attack by voice amplitude information and multiple phases. The method mainly comprises the following steps of Step 1, extracting different phase features and amplitude features; Step 2, performing discriminant training by a GMM model; and Step 3, performing information fusion in fraction level to obtain a discriminant result. The method has the advantages that the phase information in voice is fully considered; the relevant information exiting in the voice is fully utilized; by fusing different phase information, the complementarity between the phases and the amplitude is analyzed; the function of automatically judging whether the voice is real voice or replay voice is achieved; and the method is very effective for the voicefraud detection.

Description

technical field [0001] The invention relates to the technical field of automatic speaker detection, in particular to a fraudulent replay attack method for detecting voice by using voice amplitude information and various phases. Background technique [0002] Speech technology has become an important medium of human-computer interaction, and many studies have been carried out in various fields of speech processing. With the advancement of science and technology, the security and reliability of computer security systems have received great attention, and automatic speaker verification has been widely used in reality. Automatic speaker verification refers to detecting and identifying whether a given speech belongs to a given speaker, among which replay attack is the easiest spoofing attack to implement, but also one of the most dangerous forms of attack. [0003] Replay attacks in automatic speaker detection are performed using voice recordings of the target speaker that are re...

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): G10L17/00G10L17/02G10L17/04G10L17/06G10L25/51
Inventor 李东播关昊天
Owner 慧言科技(天津)有限公司
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