ROC.ai Achieves 100% Success in iBeta Level-2 Testing for Patented Single-Frame Liveness Solution
We’re thrilled to announce a major achievement for ROC.ai—our patented Single-Frame Liveness Solution has successfully passed iBeta Level-2 testing, meeting stringent facial Presentation Attack Detection (PAD) standards in accordance with ISO/IEC 30107-3.
Elevating Security Standards
Level-2 testing goes beyond its predecessor by evaluating the system’s resilience against advanced presentation attack methods. This includes attempts to trick or bypass facial biometric security using 3D, resin, and latex masks. Notably, we achieved a 0.0% penetration error during these tests, outperforming the standard compliance threshold of 1% penetration error.
The Importance of Robust Liveness Detection
The digital identity sector has witnessed significant growth over the years, and with it, the necessity for dependable liveness detection has become increasingly paramount. The emergence of sophisticated methods by fraudsters to defeat identity proofing platforms amplifies the demand for reliable facial recognition technologies.
A Trusted Testing Partner
iBeta, our testing partner, has been a renowned name in critical software testing services since 1999. Their quality assurance and testing services align with international standards, ensuring that our solution not only meets but exceeds the required benchmarks.
Our Competitive Edge
Brendan Klare, our Chief Scientist, stated, “Securing Level-2 compliance gives ROC.ai a substantial competitive advantage in the rapidly evolving landscape of identity authentication technologies. The future for us and our customers looks promising.”
Unlike other facial liveness solutions available, our patented approach is entirely sensor and system agnostic. It works across a wide range of image and video sources and delivers highly accurate liveness/anti-spoofing scores in under a second. Our system achieved a Bona fide Presentation Classification Error Rate (BPCER) of just 3.2%, far below iBeta’s 15% threshold.
Ongoing Investment and Innovation
Protected by US Patent US 10,839,251 B2, our technology will continue to see significant investment and development. We are committed to staying ahead of the curve as new, more sophisticated attacks emerge.
Rank One Computing is a leading provider of accurate and efficient facial recognition, fingerprint, and object recognition technologies. Trusted by the U.S. Military, Law Enforcement, Fintech, and commercial organizations, we are 100% Made in America and focused on ethical practices. Our technologies have proven their reliability in NIST government testing, military applications, and countless identity proofing transactions.