Innovatrics passive liveness biometrics excels in iBeta presentation attack detection test
The passive facial biometric liveness checking technology recently developed by Innovatrics has passed Level 1 iBeta Quality Assurance Presentation Attack Detection (PAD) testing with all 900 spoofing attempts rejected and all genuine user submissions correctly accepted by the system, according to a company announcement.
Innovatrics passive liveness capability was released in August, and assesses the user’s genuine presence from a single frame image, and can be carried out entirely on-device for easier protection of personal data and compliance with regulations like GDPR. The company says the that despite the advanced neural networks included in its Digital Onboarding Toolkit for face biometrics and liveness detection, the software footprint is small enough to be easily incorporated into financial insitutions’ and other customers’ applications.
iBeta tests PAD technologies according to the ISO/IEC 30107-3 standards.
Innovatrics provides its passive liveness technology as part of the company’s facial recognition solution for digital onboarding compliant with know your customer (KYC) regulations. The solution also includes a highly accurate facial recognition algorithm and ID form and document reading for a seamless user experience, the company says.
“In many parts of the world, mobile connectivity is either patchy or expensive,” explains Innovatrics Head of Product Management Daniel Ferak. “That’s why having an optimized algorithm— which is able to run purely on a mobile device with the same accuracy as the server version— is crucial for a good user experience.
The solution can also run on the server side for use cases with additional security requirements.
Innovatrics’ biometric accuracy results in recent NIST testing stood out, particularly among ABIS providers.
Article Topics
biometric liveness detection | biometric testing | biometrics | facial recognition | iBeta | Innovatrics | passive facial liveness | presentation attack detection | spoof detection | standards
Comments