Online voiceprint recognition system and implementation method thereof

Abstract

The invention discloses an online voiceprint recognition system and an implementation method thereof. The online voiceprint recognition system comprises a user recording module and an online voiceprint recognition server, and the implementation method comprises the following steps of: (1) providing generated password content for a user through the online voiceprint recognition server; (2) recording the voice of the user by utilizing the user recording module and providing the voice for a voice recognition module; (3) performing mode matching analysis on voice content provided by the user and the password content through the voice recognition module; (4) judging whether an audio fingerprint of the voice provided by the user is of the legal voice or not through an audio fingerprint recognition module; and (5) inspecting the matching degree between the voice provided by the user and a voiceprint module established by the registered user through a voiceprint recognition module, and regarding login is the real user login and receiving the login of the user if the degree achieves a set value. According to the online voiceprint recognition system and the implementation method thereof, disclosed by the invention, simple, easy-to-use and high-efficient online identity authentication can be realized.

Description

technical field [0001] The invention relates to a voiceprint identity authentication system and a realization method thereof, in particular to an online voiceprint authentication system and a realization method thereof. Background technique [0002] The current password-based identity authentication system usually needs to set different passwords for different objects for security, which increases the memory burden and ease of use of users. [0003] At the same time, voiceprint-based online authentication systems currently on the market usually generate some fixed or random texts for users to speak to identify their voiceprints. However, if the user's system is installed with some hacking tools that record the user's previous login voice, the user's previous login voice can be segmented and spliced ​​according to the text generated by the voiceprint authentication system, and then used The spliced ​​voice pretends to be the user's voice to log in, thereby harming the user's...

AI Technical Summary

Problems solved by technology

However, if the user's system is installed with some hacking tools that record the user's previous login voice, the user's previous login voice can be segmented and spliced ​​according to the text generated by the voiceprint authentication system, and then used The spliced ​​voice pretends to be the user's voice to log in, thus harming the user's interests

Therefore, the fixed password (text) or text-independent password voiceprint recognition technology adopted by the voiceprint-based identity authentication system, regardless of the fixed password or text-independent password, is vulnerable to pre-recorded voice attacks

[0004] Furthermore, in order to obtain a better user experience, the online identity authentication system based on voiceprint usually has less voice, which makes it difficult to guarantee system performance.

Images