Rank One improves face biometrics’ speed and accuracy in latest NIST test
Rank One Computing (ROC) has reported a 200 percent improvement in accuracy over the last five months based on the latest NIST Face Recognition Vendor Test (FRVT).
More specifically, Rank One says its new algorithm ROC 2.0 achieved an average of 99.4 percent across the various FRVT test cases in the January 20, 2022 edition of the FRVT 1:N report, with accuracy results within 0.1 percent and 0.4 percent of the world leader in each FRVT category.
“The accuracy of this algorithm is representative of how powerful ROC’s AI/ML face recognition has become on diverse imagery, with accuracy measured on persons from around the world,” explains Scott Swann, ROC’s CEO.
“With these substantial error rate reductions, ROC’s AI/ML SDK 2.0 has effectively reached parity with the highest-ranked FRVT performers, which are all currently of Chinese origin.”
ROC’s face biometrics solution was also recently selected by the U.S. Marshals Service to support its Justice Prisoner Alien and Transportation System (JPATS).
COO talks biometrics’ future development in Liminal podcast
David Ray, COO and general counsel at Rank One, recently participated in an episode of Liminal’s ‘State of Identity’ podcast, hosted by Cameron D’Ambrosi, discussing the role of the biometrics community in public education among a range of topics.
The COO traced the company’s roots back to its founders studying under Anil Jain at Michigan State University, to its current pattern of releasing new face biometrics algorithms every three to six months.
“Historically, the role of the biometrics community has been to push the realm of what’s possible through iterative research,” Ray says.
Commenting on the ROC’s recent FRVT results, Ray said the importance of the company’s recent achievements in face biometrics lies in its applicability to real-world scenarios.
The COO also discussed the company’s liveness detection tools and techniques, the use of its face biometrics to assert individuals’ identity in a public manner (compared with other biometric technologies), and the future of identity proofing, which will possibly increasingly rely on decentralized infrastructure based on blockchain technologies.