Paravision reports 70 percent facial recognition accuracy boost in latest NIST report

Paravision has achieved new biometric accuracy levels for facial recognition applications in NIST’s latest Face Recognition Vendor Test report.

Specifically, the company’s upcoming 5th Generation face recognition algorithm analyzed in the report showed a 70 percent reduction in error rates compared to Paravision’s previous submission.

“Accurate face recognition is vital for security, efficiency, and enhanced user experiences across so many industries and applications,” says Paravision CTO Charlie Rice.

“It comes down to trust — trust that the technology works for everyone, is consistent, reliable, and accurate. That’s why the NIST FRVT reports are so important. They are the independent gold standard when it comes to transparency in performance,” Rice adds.

In addition, the firm’s biometric algorithms ranked first in the 1:N Identification benchmark for both Visa and Border test categories, where NIST evaluates face recognition accuracy using high-quality enrollment images compared against a large dataset of images of people in challenging travel environments.

In these categories, Paravision ranked first globally for accuracy, delivering a False Negative Identification Rate of 0.22 percent at a False Positive Identification Rate of 0.3 percent across a dataset of 1,600,000 images.

The company has also ranked first in NIST’s test of facial recognition at self-service immigration kiosks, a 1:N identification test comparing high-quality enrollment images against a database of 1,600,000 kiosk photos characterized by challenging pose and light conditions.

“Given the scale of legacy providers and the massive funding and resources going specifically to Chinese AI firms, the ranking is a huge testament to our approach,” explains Paravision CEO Doug Aley.

“It’s an inflection point for our industry-trusted, ethical and accurate face recognition from a truly partner-focused company,” he concludes.

When released in Spring 2022, Paravision’s 5th Generation facial recognition technology will reportedly feature further improvements in speed and accuracy, as well as enhanced support for mobile devices, Edge AI, and other toolkits to accelerate the deployment of face biometrics applications.

Paravision has also recently partnered with Intel on a case study focusing on biometric access control applications.

Article Topics

accuracy  |  biometric identification  |  biometric testing  |  biometrics  |  face biometrics  |  Face Recognition Vendor Test (FRVT)  |  facial recognition  |  NIST  |  Paravision  |  research and development