Loquendo Speaker Verification: Identity Verification Using Voice Biometrics

A Person’s Voiceprint is Unique!

Loquendo Speaker Verification makes use of the voiceprint as a biometric to verify and demonstrate an individual’s identity.

How Does It Work?

The voice authentication process is made up of two distinct phases:

Enrollment – creates and memorizes the speaker’s voiceprint. Enrollment is carried out just once, in the initial phase, and involves the recording of key words pronounced by the user;
Verification – in this phase, the identity claim is accepted or rejected. Verification is performed each time access to the service is required, and it compares the voice characteristics of the speaker undergoing authentication with the voiceprint previously created for that identity.

Possible Applications

Speaker Verification can be used in a wide variety of application contexts, for example: phone banking and trading, password resetting, accessing customer care services, credit card activation, transactions and payments, phone top-up

Your identity is precious: benefits of Speaker Verification

A biometric authentication system has several benefits:

voice authentication is versatile, simple to use and non-intrusive, which means high user acceptance
compared to other biometric technologies, it is accurate and does not require specialized equipment - just a phone
it is the ideal solution for overcoming security issues around remote telephone access to various types of services.

General Features

it functions in several operating modes, e.g.

text-dependent – verification is performed on a specific text, which is the same in both the verification and authentication phases;
text-prompted – during verification, the system generates a different text each time for the user to repeat. This method reduces the risk of fraudulent access using recordings.

Speaker Verification is built on Loquendo Speech Recognition technology, and therefore:

because it uses the high quality acoustic models from Loquendo ASR (Automatic Speech Recognition), it is able to verify both the biometric characteristics of the voice and the content of the actual utterance, significantly reducing errors;
because it also uses voice recognition, it is able to verify the semantic meaning of the spoken password. If the required information is known only to the user (e.g. How much was your last online payment?), it significantly improves the overall security of the system (knowledge verification);
it is available in the same languages as Loquendo ASR.

System Monitoring and Calibration Tools

Loquendo Speaker Verification is equipped with tools for monitoring and calibrating the system, based on the particular security requirements of an application.
It is versatile and completely personalizable, allowing system integrators to choose:

the content of the phrase used for enrollment, and the number of repetitions required;
the number of repetitions during the verification phase.

Efficiency and Ease Of Integration

Voiceprint size is small (tens of kilobytes), while Loquendo Speaker Verification is compatible with the major database technologies and architectures for maximum ease of integration.