Rank One facial recognition algorithm scores high accuracy in DHS Biometric Technology Rally

Rank One Computing has announced its biometric facial recognition algorithm scored among the companies with the highest overall accuracy among 15 entered in the second annual U.S. Department of Homeland Security Biometric Technology Rally.

DHS tested 15 different matching algorithms on 14 different acquisition systems, with the latter collecting images in a customer-friendly process, while the former authenticated the people moving through the simulated high-traffic entry environment.

The algorithm submitted by Rank One achieved 100 percent accuracy in identifying subjects on the ‘Teton’ acquisition system, which Tascent revealed recently was its InSight Face solution. Rank One’s entry also successfully detected and enrolled passenger’s faces for 99 percent of images overall, and delivered the highest individual algorithm accuracy for 6 of the 10 acquisition systems using facial biometrics, tied or alone, and was within 1 percent of the leader on 3 of the other 4, according to the announcement.

Rank One’s entry was ROC SDK v1.18, listed as ‘WABASH.’ ROC SDK v1.19 has been released since the submission was made. The company says the SDK is easy to use for airports, borders, critical infrastructure, visitor management and watch-listing, and that its Denver, Colorado-based team is known for providing seamless licensing and technical support.

This article was edited at 4:12 ET on August 29, 2019 to clarify the algorithm rankings both for individual acquisition systems and overall.

Article Topics

accuracy  |  biometrics  |  DHS  |  facial recognition  |  Rank One  |  testing