The testing program is designed to assess a subject technology’s capability to detect ‘spoofing’ attacks that are designed to mimic legitimate biometric credentials. Level 2 entails the use of sophisticated attack methods including latex and silicone masks and three-dimensional, curved photos – attacks that go beyond the 2D images that might fool less sophisticated facial recognition systems.
Success in attaining Level 2 compliance requires a perfect score; and Oz Forensics’ facial recognition technology thwarted all of the one thousand attempted attacks thrown at it by the iBeta team.
In a statement announcing the achievement, Oz Forensics explained that its solution is based on machine learning, and that it has trained itself based on hundreds of thousands of attacks. The company also explained that it works closesly with 3D mask manufacturers to further augment its biometric solution’s attack detection capability.
The firm’s success in iBeta testing reflects its biometric solution’s success in real world deployments, suggested Oz Forensics CEO Artem Gerasimov. “Based on our customers’ feedback, Oz Forensics software hasn’t missed any real biometric attack,” he said.
Gerasimov also highlighted the speed of Oz Forensics’ technology, noting that it “allows end users to verify their identity within one second and without any unnecessary actions,” adding that customers can use it “to onboard up to 200 percent more consumers remotely.”
Indeed, the use of facial recognition for remote onboarding is a booming trend, especially in the wake of a pandemic that has accelerated various organizations’ shift into digital channels. Fraudsters are following suit, and as their attacks increase, biometric solutions’ demonstrated abilities to fight presentation attacks will only increase in value.
In a recent FindBiometrics webinar, Oz Forensics COO Svetlana Efimova demonstrated the company’s anti-spoofing technology. That presentation, “Entering the New Era of Biometric FinTech,” is available on-demand.
January 11, 2022 – by Alex Perala