Tsinghua University|中文版

Speech and Language Technology R&D Center Issued Important Industry Standards in Voiceprint Recognition Technology

On Oct. 9th ,Technical Specification for Security Application of Mobile Finance Based on Voiceprint Identification (JR/T 0164—2018), which was drafted by Speech and Language Technology R&D Center of RIIT, was issued by People’s Bank of China. The financial industry standard specifies the function, performance and security requirements of voice-based security applications in mobile financial services. The release of the standard marks that the biometric recognition technology represented by voiceprint recognition has been approved by the financial supervision department for the first time, which helps solve the standard problem of voiceprint recognition security technology in mobile finance.

Mobile financial applications are changing people’s habits of using financial services. However, how to make identification safely and conveniently has become a key constraint of mobile payment.

Voiceprint is a kind of biological feature which has been rising rapidly in recent years. It has the characteristics of uniqueness, not easy to modify, no involvement of privacy, real intention, traceability of behavior, convenience and cheap. With the maturity of this technology, it is urgent to establish norms to guide its development and application.

 China Construction Bank, Tsinghua University, Beijing d-Ear Technologies Co., Ltd. took three years to initiate and accomplish the draft of the standard. This standard comprehensively stipulates the security application technical requirements based on voiceprint recognition in mobile financial services. It is suitable for the development, detection and application of voice-based identification in mobile financial services such as mobile banking. In terms of function, the basic requirements of voiceprint registration, verification, alteration, cancellation and so on are defined. In terms of performance, the technical requirements such as sampling, anti-noise, anti-attack and anti-time-varying are put forward. In terms of security, the whole cycle of voiceprint information collection, transmission, storage, processing and deletion is standardized. In addition, the anti-attack requirements considering security and the anti-time-varying requirements of user experience are also included in the standard, which is a necessary requirement for voiceprint recognition authentication system.

The promulgation of this standard marks that the biometric recognition technology represented by voiceprint recognition has been approved by the financial supervision department for the first time, which opens up a new “voiceprint plus” identity authentication method without supervision.

38766803481362290.jpg

 

【2018-11-08】【Read:27】