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

Cross-channel biological speech recognition verification system and verification method

A technology of speech recognition and verification methods, applied in speech recognition, speech analysis, digital data authentication, etc., can solve problems such as affecting the verification accuracy rate of different channels

Pending Publication Date: 2021-06-08
效生软件科技(上海)有限公司
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to solve the problem that the verification accuracy rate of different channels will be affected by the voiceprint model update in the existing voice biometric system during the verification process of different channels

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
  • Cross-channel biological speech recognition verification system and verification method
  • Cross-channel biological speech recognition verification system and verification method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0033] Refer to attached figure 1 - attached figure 2 , a cross-channel biological voice recognition verification system of the present embodiment, comprising an audio receiving module 300, an audio matching module 500 and a main database 700 connected by mutual communication, the main database 700 is connected with several sub-databases, and the main database 700 The voiceprint model of the user's highest quality broadband data channel audio is stored inside, and several sub-databases respectively store the voiceprint models of the voiceprint model in the main database 700 after passing through different channels. The audio receiving module 300 includes several A one-to-one corresponding audio receiving unit and audio transmission unit, several said audio receiving units receive voice audio data of different channels and deliver to the audio matching module 500 through corresponding audio transmission units, several said audio transmission units are set There is a one-to-on...

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 cross-channel biological voice recognition verification system and a verification method. The system can avoid the influence of the voiceprint model generated in the verification process on the voiceprint model of the broadband data channel audio with the highest quality in the main database during verification; the voiceprint models of the corresponding sub-databases can be updated for the verification voices sent by different channels, so that the influence of the verification voices sent by different channels on the updating of the voiceprint models of the unified database is prevented, and the voice verification accuracy of different channels is improved; moreover, the plurality of sub-databases depend on the same main database, so an independent database does not need to be established for each channel, and the use and storage are more convenient.

Description

technical field [0001] The invention belongs to the technical field of voice recognition, and specifically relates to a cross-channel biological voice recognition verification system and verification method. Background technique [0002] A voice biometric system is a technology that recognizes individuals by voice by comparing a given voiceprint model with registered voiceprint models. The measure of the accuracy of voice biometric systems is usually measured by the equal error rate (EER). Current generation voice biometric systems suffer when a given voiceprint model comes from a different channel or device+. Each data channel or device leaves a unique signature on the audio file, reducing the equal error rate (EER) if the registration is on a different channel or device than the authentication. For example, if the system compares mobile app registrations with IVR authentication requests, the accuracy of IVR authentication requests is affected. [0003] In order to maint...

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): G06F21/32G06F16/61G10L15/22
CPCG06F21/32G06F16/61G10L15/22
Inventor 张寅张翼
Owner 效生软件科技(上海)有限公司