Interview: STC Group on the launch of OnePass
21 September 2016 15:41 GMT

As biometric solutions increasingly go mobile, a spotlight is being thrown on both the modalities that best suit devices and secure storage methods.

In this tussle over modalities, there is also a battle being played out over the issue of hardware.

Will there be a killer sensor that takes over the market, or will be most viable solution be one that makes best use of the existing technology that proliferates in virtually all smartphones, such as cameras, microphones and accelerometers?

A new solution which is targeting the existence of the first two, cameras and microphones, is OnePass developed by STC Group.

Uniquely the app relies on both voice and face recognition, using the cloud as a secure source of authentication data.

Planet Biometrics caught up with the company to learn more.

What makes OnePass unique in comparison to other biometric authentication solutions?

It is the first biometric authentication solution to provide bimodal face and voice verification combined with liveness detection - a unique technology developed by STC that allows determining if the person in front of the camera is live, not photographed or recorded.

Our goal was to keep it simple for the user. OnePass has just one button to press and the verification process takes less than 5 seconds. No more complex passwords to remember or one-time passcodes to receive. Face and voice modalities are combined into a single verification process and happen simultaneously. Face the camera, read the randomized numbers from the screen and that’s it. All you need are your face and voice that cannot be copied or stolen.

Why has the cloud been chosen as for this solution?

STC already has a number of non-cloud products and modifications, and we noticed that many companies are intimidated by the amount of efforts and resources it takes to implement and maintain such solutions.

For the user company a cloud solution is simpler, cheaper and faster to integrate into any app, web service or corporate security infrastructure. Compliance with all kinds of devices is much easier to maintain. What is more, cloud provides opportunities for resource pooling and affordable scaling of the services.

Considering the final user experience, as the client data is not stored on any personal device, the user information safety does not depend on security of a particular phone or computer. The biometric data is stored in the cloud in an encrypted form and cannot be used to re-create voice or visual samples. This way, even if the database gets hacked or stolen, it is impossible to use any of the data to access the system, as OnePass only recognizes live voice and dynamic image from the camera.

Can you explain its liveness detection capabilities?

Liveness detection is an algorhythm developed by STC Group – the leading Russian developer of voice and bimodal biometric systems with one of the largest R&D teams in the industry and more than 25 years of experience.

The technology is automatically activated during the verification process. While OnePass compares the voice and face data to a reference sample already stored in the system, it analyses recorded mimic expressions, lip movements and their correlation to received audio and video data. It determines whether the person speaking in front of the camera is real, not recorded. It is the third level of security that ensures impossibility of using any kind of recording – photo, video or audio.

Did you face any challenges in developing this solution?

When we just started developing OnePass, the verification success was rather dependent on recording conditions. A fair amount of light and absence of loud background noises were a necessity. However, after numerous testing sessions, we managed to overcome this difficulty, and now the system is less susceptible to external noises. It still cannot recognize you in the dark or in a very loud environment, but this is the case for any face and voice identification platform out there.

Do you feel that blending modalities in this way presents a stronger security proposition?

Combining face and voice is a convenient and cost-efficient way of utilizing what is already owned by the customers, unlike with fingerprint identification which requires additional scanners. Not everyone owns an iPhone, but majority of people has a smartphone with built-in camera and microphone.

At the same time, accuracy of two modalities is drastically higher than that of a single biometric. According to our tests, OnePass verifies authorized users 98% of the time and can prevent unauthorized access with 99.99% accuracy. These results significantly exceed corresponding indicators for authorization with a single authentication factor.

What applications will OnePass be best suited to?

STC Group expects OnePass to be successful in sectors like banking, healthcare, government, telecom, e-commerce, gaming, social media and travel.

Related articles

SpeechPro and PenRad to bring voice biometrics to healthcare industry
21/07/15
FaceBanx, SpeechPro partner on multimodal solutions
20/05/15
SpeechPro to partner with CallMiner on voice solution
01/05/15