Innovatrics has reached another important milestone with its on-device passive liveness check technology. The company has announced that it passed Level 2 of iBeta’s Presentation Attack Detection evaluation.
A widely regarded benchmark for third-party testing for anti-spoofing face biometrics technology, iBeta’s program is designed to evaluate various solutions on their effectiveness at detecting presentation attacks in which fraudulent biometrics are used in an attempt to fool a face-based authentication system.
Level 1 of PAD testing is focused on simpler presentation attacks, such as those leveraging a 2D image of a legitimate end user to fool an authentication system. Innovatrics’ anti-spoofing solution reached that touchstone last autumn, just a handful of months after it was launched in April of 2020. Now, in passing Level 2 of iBeta testing, the Innovatrics technology has demonstrated its ability to detect more sophisticated spoofing attempts based on things like latex masks.
In order to successfully pass each Level of testing, liveness detection algorithms have to withstand every attack that iBeta can muster. In other words, they must maintain a 100 percent success rate in detecting presentation attacks.
“We’re very happy to see that our algorithm is able to tackle ever more challenging scenarios of presentation attacks with 100% accuracy,” commented Innovatrics Head of Product Management Daniel Ferak.
The achievement offers the latest testament to the quality of Innovatrics’ technology after the company delivered a top-three result in the Biometric Technology Rally, a competitive evaluation program run by the Department of Homeland Security’s Science & Technology Directorate, earlier this year. Innovatrics has also recently received high praise from its partner Precise Biometrics, which is using its SmartFace solution in its YOUNiQ biometric access control offering.
Meanwhile, Innovatrics continues to upgrade other aspects of its solutions portfolio, having recently announced improvements to its fingerprint algorithm to make it more accurate in matching the fingerprints of children.
May 10, 2021 – by Alex Perala