Pryathna Sankaranarayanan | November 29, 2017 | 4 minutes
Some industry participants have raised concerns that voice biometrics solutions may be subject to failure due to changes in an individual’s voice caused by ageing. These changes are claimed to have a negative effect on accuracy of voice authentication. After all, the voice you have as a teenager is different from the voice you have as an adult – can the system recognize your voice as you age?
Some claims have been made that the technology cant keep track of vocal changes that occur over time. If the system doesn’t track changes then eventually the systems rejects the true speaker and becomes less secure.
Many voice biometric vendors have attempted to address this concern by using historical voice recordings of high profile public figures over a period; such as President Obama, Arnold Schwarzenegger and Morgan Freeman etc., to show how they can identify a speaker despite changes in voice over time.
This approach, however, fails to consider real-world production deployments where voices and audio systems are changing in different ways. Different microphones, for example, capture your voice differently based on audio received from these devices.
You don’t need to ask your existing customers to re-enrol their voice print to switch to the Next Generation ArmorVox solution and you will never have to get them to re-enrol for any future upgrade. This means your customers do not have to spend time re-enrolling their voice to use a next generation voice verification system, saving you time and costs and allowing you to reap the benefits of an omni-channel solution that delivers real time fraud prevention and detection.
Not only can ArmorVox be deployed without needing customers to reenrol their voice prints, you will never have to go through a manual and expensive re-tuning process to get the best performance from the system. ArmorVox uses proprietary machine learning algorithms to continuously optimise the performance of the system every time a user enrols or verifies. Customers can verify with their existing pass phrases or use digits, phrases or a combination – you can customize verification for customers and provide them with increased choice for the telephony and digital channels.
Switching costs are reduced because the investment in existing applications and system integration with telephony, databases and contact centre technologies can all be re-used. Changing the underlying voice biometric engine provides significant usability and security benefits, lowers recurring maintenance costs and delivers a solution that can be easily extended to digital channels.
Talk to our specialist team to find out how long the process will take. Typical implementations can be upgraded with no system downtime and minimal input from internal IT teams in a matter of weeks. Use your own inhouse team to upgrade and make changes as your business needs change or use your existing application development team or IVR provider to support your next generation voice biometric solution.
ArmorVox provides a biometric solution that sets a defined security setting for every transaction, for every individual. You can set your risk profiles on a per transaction basis, which means for transactions of high value and risk, you can increase security with random digit challenges. The solution can be enhanced to improve user experience with a simple verification passphrase such as a phone number or account number. You can even combine device and location information with business rules to manage risks in high value transactions.
The ArmorVox Advanced Security Suite comes with a host of features designed to help you manage fraud in real-time. With our patented Impostor Mapping, you can identify potential duplicate enrolees within the system. ArmorVox has a cross-matching capability of 25 million voiceprints per minute, helping you detect and track fraud for your entire customer base. You can run hot lists – which are special lists for recent activators, known fraudsters and clients with special needs and verify customers accurately in real-time.
ArmorVox’s proprietary active learning algorithm features a tuning method to automatically update a speaker’s voice print every time they verify. This balances the difference between knowing when a true speaker has a change in the way they say things as opposed to someone impersonating them. ArmorVox detects and adjusts for subtle changes in the speaker’s voice over the natural course of his life or when he changes devices.
Our proprietary Speaker Specific Background Model maps the speaker’s gender, ethnicity and other attributes to identify more granular aspects of a speaker’s voice. This innovation enables ArmorVox to clearly differentiate the true speaker from similar sounding impostors such as twins. ArmorVox distinguishes family and other individuals likely to be a close voice match.
For speakers with a weaker voice print, ArmorVox automatically adjusts the threshold so there are no vulnerabilities. The more a speaker verifies with their voice, the more opportunity there is to create a robust voice print – as ArmorVox is designed with active learning which continuously learns about a speaker’s voice and adjusts the threshold accordingly.
Our technology can recognise the underlying distinctive characteristics of a synthetic voice. If an account is being attacked by a synthetically manufactured voice then security procedures can be invoked to protect he individual and their assets
By upgrading to ArmorVox, you get all the features and functionality required for successful voice biometric solution in a single software license. This includes passive, active and background enrolment and verification options. It also provides eight different text dependent tokens, two text prompted tokens, a digit independent and a text independent token.