Paravision Converge introduces new face biometrics features, optimizations, tools
Paravision says its comprehensive face recognition product suite now provides all of the tools developers need to develop and deploy biometric identification and authentication solutions for a broad range of solutions.
The latest set of product and feature releases from Paravision is part of a refresh of the company’s suite of offerings based on internal development to meet customer needs and an ecosystem approach to providing flexibility, as President and COO Benji Hutchison and CPO Joey Pritikin told Biometric Update in an interview.
New features included in the set of announcements the company calls Paravision Converge 2021 are Person Detection, Edge AI reference designs for on-device face biometrics, Face Recognition SDK for Android and Streaming Container.
Paravision has also introduced Scaled Vector Search (SVS), which enables a search of 10 million records in 50 milliseconds on a single compute instance.
“The ability to deliver that up in various modern ways through flexible APIs and so forth, it really is a game changer,” Paravision President and COO Benji Hutchison told Biometric Update in an interview on the Converge announcements.
“What we’ve done is embraced some of the latest AI technologies around search, and apply them to face recognition,” CPO Joey Pritikin adds. “And so what that means is faster, more responsive systems and also lower operating costs and the ability to hit scale.”
In addition to the new features, the update also includes optimization for leading chipsets like Microsoft CPUs, Nvidia GPUs and Ambarella SoCs, according to a company product sheet. Paravision’s passive real-time Presentation Attack Detection (PAD) is also optimized for biometric liveness and anti-spoofing on camera solutions based on Ambarella and Intel RealSense.
The combined tools and features make Paravision’s computer vision technology easier to deploy, whether on mobile devices or at the edge, in the cloud or an internal data center, according to the executives.
The tools the company provides take in reference applications and reference hardware designs, requirements definitions, architecture workshops, and roadmap workshops for partners.
Paravision has topped other U.S.-based providers in the National Institute of Standards and Technology’s (NIST’s) Face Recognition Vendor Test for biometric accuracy, and ranked among the leaders in each category. Hutchison also teased new models that he says will be smaller, more lightweight, and more accurate.
biometrics | computer vision | edge AI | face biometrics | facial recognition | mobile biometrics | Paravision | person detection | presentation attack detection