Voice Biometrics

Voice Biometrics

Why Voice?

ValidSoft utilises voice biometrics because it provides the unique combination of very strong security and a natural and personal user experience that works across all channels. Unlike synthetic credentials such as PINs, passwords and tokens that are attributed to a user, voice biometrics is naturally generated by the user, wholly unique and always available. This ease of use and availability is a key differentiator from other biometrics as no specialist hardware is required. A user’s voice can always be provided via a mobile, telephone, tablet or PC. What could be more personal than purely having to speak to successfully complete a transaction?

Why not other biometrics?

At ValidSoft we do believe that other biometrics have value, and we see particular contexts where they can provide strong results. As businesses continue to deploy multi-channel modes of operation voice is unique in working consistently across all channels, providing a consistent user experience without the need for specialist hardware. Adding other biometric modalities can add value in certain scenarios and thus we have built our User Authentication platform with flexibility to encompass additional biometric modalities. However, it is the security and flexibility of use across all channels that has driven ValidSoft to focus our biometric research and development resources on building the best voice biometrics solution in the market today.

​A voice signature is not a recording of your voice

A voice signature is not as many people believe a recording of one’s voice, stored in a computer system and which if stolen, can be reused, like a password or a PIN. A voice signature is in reality a mathematical representation of features extracted from the voice. This is created using a specific mathematical algorithm, a process that takes place in fractions of a second, and is typically represented by a series of hashed integers that cannot be reused in any other system. In the unlikely event that a voice signature were compromised it could be immediately revoked and the user could create a new and wholly unique voice signature.

What is the difference between text dependent and text independent?

A text dependent system is one where the user repeats a prompted phrase whenever they are required to authenticate themselves. Text dependent solutions are most typically used within mobile apps, tablets, PC’s and automated telephony systems, where they have the advantage of delivering a great performance with a small amount of speech. In a text-independent mode, no specific text is required, instead the user is simply required to speak naturally, therefore this technology is most typically used while processing conversations such as in call centres.

How can a user enrol?

The system comes with a full and flexible enrolment module that supports multiple enrolment techniques, including app-based, telephony and web based, all of which can be secured. Crucially, the system allows for the creation and deployment of multiple enrolment processes that meet the customer's needs.

What is the difference between passive and active enrolment?

Passive enrolment is the process by which the audio used to build a voice signature for an individual user is captured during a normal speech interaction e.g. a discussion with a call centre agent, therefore the customer does not have to take a specific action to be enrolled. This is in contrast to an active enrolment during which the customer is prompted to repeat some given phrases. ValidSoft supports both passive and active enrolment models.

If you'd like to see how ValidSoft could help your business, get in touch