Auraya Systems Pty Ltd, the developer of next generation voice biometrics ArmorVox, has been awarded patent recognition in China for its innovation in fraud prevention using voice biometrics systems.
Auraya Systems today announced that the Chinese patent office has awarded patent recognition in China for its innovative technology that protects customers against fraudsters.
The innovation allows for a random challenge question to be used when verifying on telephones or digital channels such as chat, websites and voice assistants. The user is asked to say the words or digits displayed and ArmorVox recognises if it is the authorised account holder that is speaking and if the correct random phrase is being spoken. This technology helps businesses to verify the identity of a customer’s voice and what is being said to increase security with any high-risk transactions.
“This patented innovation provides Auraya with unique capabilities to provide secure voice user interfaces.” said Dr Summerfield, founder, CTO and co-inventor of the technology.
The technology has already been used where a four-digit random challenge increases the security threshold for high value transactions. Customers enrol and verify with randomly generated one-time “spoken tokens” and this prevents recordings of the user being used to thwart the system.
“This is like a voice version of the one-time SMS passwords that we are all familiar with for doing Internet bank transactions, just a lot more secure. Because the voice biometric tokens are randomly generated, a recording of a previous authentication cannot be used. Each voice authentication has to be live and unique!” explained Dr Summerfield.
“The applications for a fusion of speech and voice are endless – we are already seeing consumer demand for the new capability in automated on-line virtual assistants, where the assistant can recognise not only what is being said, but who’s speaking. This technology breakthrough gives consumers total ownership of their devices and privacy with their interactions.”
The technology is fast gaining popularity for secure voice driven financial services especially digital applications where customers of banks are increasingly making transactions with their voice. “But this is not the only area of interest.” explained Paul Magee, Auraya’s CEO. “Consumer devices, smartphones, personal assistants, in car controls are all areas where the new technology has profound applications and uses. Knowing who is speaking eliminates the need for PINs, passwords and access codes, and allows systems to ignore requests from speakers it does not recognise”
“It stops the children from changing the channel on the TV”; explained Mr Magee, “and that has got to be a good thing!”
The two components – speech recognition and speaker recognition are essential and work together to deliver a seamless, user-friendly experience. It has significant potential to streamline mobile payments and eliminate the inconvenience with PINs, passwords or SMS tokens and transform customer experience.
The patent in China is in addition to patents already granted in Australia, USA, Canada, UK and HongKong.
Auraya, a world leader in biometric voice verification technology empowers people and organizations to interact and engage with security and convenience. As a specialist voice biometric technology developer, we have a track record of delivering unparalleled security performance that is simple to deploy, integrate and maintain whilst delivering the most delightful customer experience.
Auraya’s biometric voice verification engine – ArmorVox – is speaker adaptive, which means it constantly learns about the customer’s unique voice and optimises the voice print performance with each use. This means we not only protect the privacy of the individual and the security of the organisation, but also deliver a truly differentiated experience for the customer. ArmorVox is the perfect voice verification engine that includes passive, active and background enrolment and verification options. It also provides eight different text dependent tokens, two text prompted tokens and a text independent token.