Anonymous speaker identity verification traceability system

By introducing anonymous speaker identity verification traceability system in the voiceprint authentication system, combined with the ECAPA-TDNN and MFA-Conformer models, the problems of insufficient fake audio detection capabilities, lack of traceability and adaptability in the existing technology are solved, and the problem of insufficient fake audio detection capabilities and lack of traceability and adaptability are realized. Efficient detection and audio traceability of forged audio.

CN119479656APending Publication Date: 2025-02-18NANJING LONGYUAN INFORMATION TECH CO LTD
View PDF 5 Cites 2 Cited by

Patent Information

Application Number
CN202411649065.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-11-19
Publication Date
2025-02-18

AI Technical Summary

Technical Problem

The existing voiceprint authentication system performs well in normal environments, but lacks sufficient mechanisms to identify and defend against sounds processed by advanced anonymity technology, resulting in insufficient detection capabilities of forged audio, lack of traceability and adaptability and flexibility.

Method used

Adopt an anonymous speaker identity verification traceability system, including audio input module, preprocessing module, forged audio detection module, voiceprint traceability module, identity and anonymous technical confirmation module, normal audio processing module and identity verification module, combined with ECAPA-TDNN The model and one class loss perform fake audio detection, and the MFA-Conformer model is used for voiceprint traceability.

Benefits of technology

Effectively distinguish between real audio and fake audio, optimize model structure and algorithms, reduce dependence on a large amount of training data, improve adaptability and sensitivity to new forgery technologies, and realize accurate detection of fake audio and the real identity and anonymity technology behind the audio tracking.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119479656A_ABST
    Figure CN119479656A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of voiceprint recognition, in particular to an anonymous speaker identity verification traceability system which comprises an audio input module, a preprocessing module, a forged audio detection module, a voiceprint traceability module, an identity and anonymous technology confirmation module, a normal audio processing module and an identity verification module. The forged audio detection module is connected with the preprocessing module, the voiceprint traceability module and the normal audio processing module are both connected with the forged audio detection module, and the identity and anonymity technology confirmation module and the identity verification module are respectively connected with the voiceprint traceability module and the normal audio processing module. In this way, the technical problems that in the prior art, a voiceprint authentication system is good in performance in a normal environment, but lacks enough mechanisms to recognize and defend sounds processed through an advanced anonymous technology, and consequently the detection capacity of forged audios is insufficient, traceability is lacked, and adaptability and flexibility are insufficient are solved.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • System and method for detecting identity impersonation in telephone satisfaction survey

    CN103778917A

  • Credit card security monitoring method based on multiple information verification

    CN108985776A

  • Secure automatic speaker verification system

    CN115280410A

  • Speaker authentication system and method

    US20030182119A1

  • Authenticating a user

    US20210125619A1

Cited By

  • Cross-channel forged voice noise reduction traceability method and system

    CN120126502A

  • Information monitoring and tracing method and system based on voiceprint and semantic analysis

    CN121687013A