Edge AI, facial recognition tech come together with Blaize, Innovatrics partnership

SmartFace Embedded is a customized variant of its facial recognition libraries for use in Ambarella CV22 and CV25 system-on-a-chip processors designed for AI-based face biometrics functions like face detection and face template extraction in images and videos. When combined with AI computing on the edge from the Pathfinder P1600 Embedded SoM, SmartFace Embedded can detect facial landmarks within 5 milliseconds, face template extraction within 15 milliseconds, and face detection within 20 milliseconds, according to a joint announcement.

The paired solutions are designed for scalability, minimal compute capability, and a small footprint for data processing on the edge. Blaize says the edge-to-cloud approach it takes will enable any surveillance camera to become “smart” with facial recognition.

Innovatrics has taken steps to proceed in the edge biometrics market, like with cascaded architecture based on edge video processing that can support real-time video analysis and biometric face identification.

“Blaize focuses on delivering low power and low latency AI inference solutions at the edge, and Innovatrics with their SmartFace Embedded enables customers to realize the value of AI in smart city, smart retail, safety and security applications,” says Barrie Mullins, head of marketing at Blaize. “Innovatrics facial recognition technology seamlessly integrates into our cost-effective and flexible edge form factor solutions, making it easy to use.”

Edge Industry Review Editor Jim Davis recently outlined the potential of biometrics processing at the network edge in a white paper.

This article was originally posted on Biometric Update.

Article Topics

Blaize  |  edge AI  |  embedded hardware  |  facial recognition  |  Innovatrics