top of page
English Lockstone.png
Water

Why We Need It

Accuracy

Voice Biometrics Identificationis one of the most accurate forms of biometric identification. It has an accuracy rate of up to 98%. This is significantly higher than the accuracy rates of other forms of biometric identification, such as fingerprint scanning (95%) and facial recognition (85%).

Speed

Voice recognition is also one of the fastest forms of biometric identification. It typically takes less than two seconds to identify someone using voice recognition software. This is significantly faster than the time it takes to identify someone using other forms of biometric identification, such as fingerprint scanning (typically five to ten seconds) and facial recognition.

Safety

As the most privacy-friendly forms of biometric identification. Voice Biometrics Identification does not require any physical contact with the person being identified and it does not produce any image or recording that could be used for identity theft or fraud. 

​

​

Reliability

Voice Biometrics Identification has a false acceptance rate (FAR) of 0.00008% and a false rejection rate (FRR) of 2%. By comparison, fingerprint scanning has a FAR of 0.002% and facial recognition has a FAR of 0.01%.

​

​

Convenience

Voice Biometrics Identification does not require any special equipment or training to use. By comparison, fingerprint scanning requires special equipment (e.g., a fingerprint scanner) and facial recognition requires special software (e.g., face detection software).

​

​

​

  • White Facebook Icon
  • White Instagram Icon
  • White Twitter Icon
Wavy Abstract Background

Multi-dimensional

Biometrics

Identificaiton

LockStone Voice Biometrics Identifier (VBI in short) offers a promising alternative to traditional PIN and password methods, allowing banks to verify customer identity quickly before processing transactions via mobile app, ATM, phone, or website.

​

Step 1, Enrollment

Active enrollment: By repeating some given phrases or digits, voice identity for customer (agent) will be made

Passive enrollment: By capturing audio segments during customer interactions, together with customer ID including Social ID, phone number, customer's identity will be made.

​

Step 2, Authentication 

Active authentication: users will be promoted to speak out the digits generated by system randomly

Passive Authentication: after active enrollment, the consumer is authenticated via short utterances.

​

Step 3, Build up Trust List and Blocked List basing on pre-record user data 

​

Abstract Futuristic Background

Features

Stability
  • Access availability not less than 99.95%

  • TPS not less than 20

  • The system restart time is no more than 5 minutes

Security
  • DES is used to encrypt the interactive data with voice biometrics engine

  • Server anti cracking login

Reliability
  • Data backup, keep data no less than 30 days

  • Data recovery, full data recovery is no more than 30 minutes

  • System self check function

  • Early warning mechanism of system failure

Compati-bility
  • Provide standard restful API interface

  • Support data acquisition of mainstream traffic system

  • Support IP V4 / v6 dual stack service

Scalability
  • Function modules can be expanded rapidly

  • The system processing capacity supports horizontal expansion

  • Support self balancing load function

© 2023 by LockStone Technologies Ltd.

bottom of page