Vision-Box says results in NIST facial recognition test show seamless travel potential

Vision-Box has submitted its biometric facial recognition algorithm to the National Institute of Standards and Technology’s (NIST’s) Facial Recognition Vendor Test (FRVT), and the company says it is outperforming some of the field’s top competitors.

The latest NIST FRVT 1:1 and 1:N tests opened June 3, and Vision-Box says that in testing of 109 algorithms from more than 60 providers, the company had a particularly strong result in the live-to-live recognition test-set. The “visionbox-001” algorithm ranked 24th in the mugshot category and 34th in the wild images category in the latest draft report of the FRVT Part 1:Verification. The results demonstrates the technology’s suitability to use in Seamless Travel programs, according to the announcement.

“Receiving this recognition confirms our ambition to provide one of the leading matching services integrated within our Orchestra suite, a technology which is continuously being optimized for passenger handling in large scale scenarios. We are dedicating the same level of attention to matching performance, as we have been successfully doing so for the biometric capture, because both have to go hand-in-hand, if accurate performance is expected in highly scaled implementations,” Vision-Box CEO Miguel Leitmann points out.

“Several ongoing deployments are taking advantage of our technology today, providing conditions where airports, airlines and border police can finally collaborate in a ‘Privacy-by-Design’ certified environment, eliminating the legacy of separated data silos and handling sensitive data challenges related to identity management in collaboration,” Leitmann continues.

The Vision-Box FR facial recognition algorithm is based on artificial intelligence and deep learning techniques, and is part of the company’s flagship Orchestra Seamless Traveler portfolio. Orchestra users can connect stakeholders, including border control authorities, airlines, and airports with its included Identity Management as a Service (IdMaaS) architecture.

The company notes that the most advanced OneID deployments, such as those at Amsterdam-Schiphol, Dubai, Bangalore, and Aruba, include hundreds of passenger process-points from curb to gate. Vision-Box also recently deployed its biometric boarding platform to support JFK Terminal One’s OneID vision.

Article Topics

algorithm  |  biometrics  |  facial recognition  |  NIST  |  testing  |  Vision-Box