Voice-activated object recognition and mobile-based facial authentication powered by Infineon’s PSOC™ Edge platform
At Embedded World 2025, embedUR systems partnered with Infineon Technologies to unveil two cutting-edge Edge AI demonstrations, highlighting the growing viability of fully local, real-time AI applications in both consumer and industrial environments. Powered entirely by Infineon’s PSOC™ Edge platform, the demos drew attention for their responsiveness, power efficiency, and complete on-device intelligence—no cloud connectivity required.
“Edge AI is no longer a concept—it’s production-ready and cloud-free.” — John Marconi, VP Technology, embedUR systems
Demo 1: Voice-Activated Object Recognition + Recipe Generation
The first live demo used voice commands to trigger object detection via a camera feed, identifying ingredients and instantly generating recipe suggestions—all processed on the PSOC™ Edge hardware. This smart home scenario underlined the future of intuitive, AI-powered interfaces that protect user privacy by operating solely at the edge.
Demo 2: Facial Recognition with Mobile Registration
In the second showcase, embedUR presented mobile-based facial onboarding. Users could register their facial profiles through an app and then be recognized locally by the edge device. The real-time facial authentication, running without any cloud dependency, demonstrated a secure and efficient access control mechanism suitable for homes and industrial sites alike.
“These demos highlight how PSOC™ Edge enables real-time, power-efficient Edge AI,” said Omar Cruz, Senior Manager, Infineon Technologies. “We’re empowering developers to bring intelligent automation to life across industries.”
“At embedUR, we’re focused on making Edge AI real,” added John Marconi, VP of Technology, embedUR systems. “This collaboration shows that advanced AI like object recognition and facial authentication can now run securely, efficiently, and completely offline.”
More Edge AI, Less Complexity
All models used in the demos are available via ModelNova, embedUR’s AI model library for edge developers. With expertise in embedded software, AI model optimization, and system-level integration, embedUR helps product teams rapidly build and deploy intelligent solutions—even on memory-constrained devices.