Hyperverge impresses in NIST biometric identification, Paperless Travel tests
Silicon Valley and India-based biometric algorithm developer Hyperverge has barged its way among the leaders in facial recognition accuracy testing by the U.S. National Institute of Standards and Technology.
The company notes its strong position relative to other biometrics providers in India and Southeast Asia, and outside of China in general.
Hyperverge made strong showings in both the NIST Face Recognition Vendor Test (FRVT) 1:N test update published on January 20, and in the Paperless Travel update on January 24. The company is among six new developers contributing algorithms to the 1:N benchmark, but the company did not submit its algorithm in time for the previous full report on Paperless Travel, which was published in December.
In the update to the 1:N Identification test published January 24, Hyperverge achieved top-10 biometric accuracy results in the Mugshot Profile and Visa Kiosk categories, and further top-20 results in the Mugshot (N=12 million), Visa Border and Border (10 years later) categories.
In the Paperless Travel test, the company was found to have the fourth-best percentage of travelers not matched (multiple enrollee images in gallery), in a virtual five-way tie for the top result at 0.2 percent. Hyperverge also cracked the top 10 in terms of accuracy for matches against a single enrollee image against databases of both 420 and 42,000 subjects.
Hyperverge has also been confirmed for ISO 30107 biometric presentation attack detection (PAD) standard compliance by iBeta.