Paravision Face Recognition Delivers Top Performance in Latest DHS Biometric Test

Company achieves 100% success rate across all demographics on multiple image acquisition devices

SAN FRANCISCO, May 02, 2022 (GLOBE NEWSWIRE) -- Paravision, the U.S.-based leader in mission-critical Vision AI, today announced that it delivered the top-ranked performance among face recognition providers selected to participate in the U.S. Department of Homeland Security’s (DHS) Biometric Technology Rally. Paravision’s face recognition system achieved a 100% True Identification Rate (0% error rate) on a range of test subject demographics on images captured by three different acquisition devices, a feat unmatched by any other face recognition provider.

The DHS Biometrics Technology Rally tests acquisition devices and matching systems in an operational setting, delivering statistically sound results reflecting broad demographic diversity. Published results include breakdowns for gender, race (self-reported, including Asian, Black, and White) and skin tone (measured with a calibrated color meter device). Acquisition and matching system vendors are reported by an alias; for this evaluation, Paravision’s code name was “Salt.”

Paravision’s highlights of the evaluation include:

  • 100% Matching system-focused True Identification Rate (“Matching-TIR”) across three of five submitted image acquisition devices
  • 100% Matching-TIR for both Male and Female self-reported gender demographic groups
  • 100% Matching-TIR for Asian, Black, and White self-reported racial demographic groups
  • 100% Matching-TIR for Darker Skin and Lighter Skin demographic groups, as measured with a calibrated color meter.

“Paravision relentlessly strives for outstanding face recognition performance across age, gender, race, and other characteristics, and it’s very rewarding to see that focus manifest in this level of operational performance,” said Charlie Rice, Paravision CTO. “Our work on accuracy improvements, specifically those inclusive of broad demographics, is never done. But 0% error rate across all groups and a range of capture devices in a test of this magnitude is a great milestone.”

Over the last three years, Paravision has consistently ranked as a top performer in NIST’s rigorous Face Recognition Vendor Test, including recently achieving the global #1 ranking in 1:N Identification and a top-three global ranking in Paperless Travel. The DHS Biometric Technology Rally results offer a complementary look at Paravision’s performance. Notably, this test was performed on Paravision’s Gen 4 face recognition. In February of this year, Paravision released its Gen 5 product suite, which offers further reduction in error rates for the most challenging use cases while offering support for deployments from Cloud to Edge.

“Paravision’s performance in the DHS Biometric Technology Rally is the latest clear sign that face recognition is an outstanding tool to verify identity in travel, borders, and a wide range of other use cases,” said Benji Hutchinson, Paravision President and COO. “The data is solid and the numbers are clear: with the best products available in the market, face recognition can work for everyone.”

About Paravision
Leveraging the latest advances in deep learning, Paravision’s industry-leading Vision AI software is trusted globally by security device manufacturers, solution providers, systems integrators, and financial services firms to deliver frictionless experiences and comprehensive safety and security. From the latest in Edge AI platforms to cloud-ready solutions, Paravision powers computer vision-centric solutions with modular, highly scalable toolsets that deliver breakthrough capabilities in the most challenging conditions.

Paravision has developed deep expertise in face recognition specifically and strives to consistently raise the bar for face recognition technology in accuracy, security, inclusion, and convenience. Paravision has consistently ranked among the top global performers in NIST’s gold standard Face Recognition Vendor Tests, including 1:1 Verification, 1:N Identification, Face Mask Effects, and Paperless Travel, and matches these benchmarks with breakthrough performance in a wide variety of identity-centric applications around the world.

Media Contact
Chris Ulbrich
(415) 878-9175