ROC.ai's Patented Single-Frame Liveness Solution Completes New Milestone iBeta Level-2 Testing with 100% Success Rate
In fact, while the threshold for compliance is no more than 1% penetration error, ROC.ai achieved 0.0% penetration error during the tests.
DENVER, Aug. 29, 2023 /PRNewswire/ -- Rank One Computing (ROC.ai) has achieved Level-2 testing compliance for facial Presentation Attack Detection (PAD) standards in accordance with ISO/IEC 30107-3 and tested by iBeta.
Level-2 compliance builds on Level-1 with more sophisticated presentation attack methods, such as use of 3D, resin and latex masks to spoof or bypass facial biometric security. In fact, while the threshold for compliance is no more than 1% penetration error, ROC.ai achieved 0.0% penetration error during the tests. The importance of iBeta certifications has grown substantially in the past several years as the digital identity sector has experienced material growth. The need for robust liveness detection has become paramount as more and more facial recognition use cases proliferate across cyberspace. These trends are accompanied by an increase in the sophistication of fraudsters seeking to spoof or defeat identity proofing platforms that leverage facial recognition technologies. The demand for liveness in cyberspace will only increase over time.
According to their website, iBeta has been providing critical software testing services for the world's most trusted brands since 1999. QA and software testing services conform to the requirements of the International Standard ISO/IEC 17025:2005(E), the needs of the clients, and the accrediting organizations for inspection and product certification.
"Building on our previous achievements in the liveness arena, we are extremely proud to have the next level of iBeta's testing compliance under our belt," said Brendan Klare, Chief Scientist. "We continuously look for ways to improve our offerings and better serve our customers who need ID Proofing technologies. Having obtained Level-2 compliance, ROC.ai has a stronger, competitive edge in the evolving global landscape of identity authentication technologies and the future looks brighter now."
Unlike other facial liveness solutions on the market, ROC's patented approach is entirely sensor and system agnostic, works with images and videos from any source, and returns a highly accurate liveness/anti-spoofing score in less than one second. ROC.ai achieved a Bona fide Presentation Classification Error Rate or BPCER of 3.2%. This is the proportion of bona fide presentations incorrectly classified as presentation attacks in a specific scenario. Essentially how many genuine (live human) samples did we incorrectly reject. iBeta allows for this to be 15% so we were well below that threshold to get the certification.
"We are really proud of our team's ability to achieve such significant technical results," said Kayur Patel, Director of Machine Learning. "Advancing our liveness capabilities to this new level of certification demonstrates that we have the technical acumen to tackle some of the hardest problems we see in the spoofing space."
ROC.ai's approach is protected by US Patent US 10,839,251 B2 "Method and system for implementing image authentication for authenticating persons or items" since June 26, 2017. This foundational patent broadly covers any computer vision liveness approach that analyzes image regions to identify spatial relationships amongst pixels or groups of pixels. ROC'ai's method uses a standard camera without the need of additional sensors, such as infrared (IR), which others in the field may leverage. ROC.ai will continue to invest in liveness and will continue to develop improvements to stay ahead of the curb as bad actors think of more sophisticated attacks.
About ROC.ai – Rank One Computing is a world leader in accurate and efficient face recognition (FR), fingerprint, and object recognition built on AI/ML computer vision. We are the most trusted provider of AI/ML biometric matchers to the U.S. Military, Law Enforcement, Fintech, and Commercial organizations. We are employee-owned, ethics driven, and 100% Made in America. Our AI/ML computer vision algorithms lead the industry in security, accuracy, and speed as proven in NIST government testing, tactical military applications, and hundreds of millions of identity proofing transactios. Our company has offices in Denver, Colorado and Morgantown, West Virginia.
Media Contact:
Jessica Sell, Vice President of Congressional Affairs & Community Outreach
949-874-2347
[email protected]
View original content to download multimedia:https://www.prnewswire.com/news-releases/rocais-patented-single-frame-liveness-solution-completes-new-milestone-ibeta-level-2-testing-with-100-success-rate-301912201.html
SOURCE Rank One Computing
