EmbedUR systems and Infineon collaborate to showcase advancements in voice, object and facial recognition using edge AI – CRN

At Embedded World 2025, embedUR systems partnered with Infineon Technologies to unveil two innovative Edge AI demonstrations powered entirely by the PSOC™ Edge platform. These live showcases drew significant attention for their real-time responsiveness, ultra-low power operation, and practical use cases across consumer and industrial applications.

Voice-Activated Object Recognition + Recipe Generation
The first demo featured wake word detection and real-time object identification. Attendees activated the Edge AI system with a voice command, then presented an image featuring various ingredients. The AI identified the objects in real time and generated a relevant recipe suggestion—all processed locally on the PSOC™ Edge platform without cloud dependency. This demo illustrated the potential for intuitive, AI-powered user experiences in smart home devices.

Facial Recognition with Mobile-Based Registration
The second demo introduced a facial recognition use case with mobile onboarding. embedUR developed a companion mobile app allowing users to upload and register their facial profiles. Once registered, the PSOC™ Edge platform could detect and recognise these users in real time, running all AI inference locally on the board. This fully edge-based authentication solution demonstrated secure, low-latency identity recognition for consumer and industrial systems.

“Our collaboration with embedUR systems showcases the power of PSOC™ Edge in enabling real-time, power-efficient Edge AI applications,” said Omar Cruz, Senior Manager at Infineon Technologies. “These demos highlight how our platform accelerates intelligent automation for smart home, consumer, and industrial use cases.”

“At embedUR, we’re focused on making Edge AI real—delivering practical, production-ready solutions that run on today’s silicon,” said John Marconi, VP of Technology at embedUR systems. “The work we’ve done on Infineon platforms demonstrates how edge intelligence can be deployed quickly and securely, whether it’s voice-activated control or face-based access—all without relying on the cloud.”

Edge AI Innovation Powered by PSOC™ Edge
Infineon’s PSOC™ Edge family enables intelligent, always-on AI applications with hardware-accelerated neural network processing, a Cortex®-M55 core with Helium DSP support, and robust security features. Designed for consumer and industrial applications, the platform provides a seamless AI development experience, empowering developers to build fast, secure, and power-efficient solutions at the edge.

embedUR systems: Engineering the Edge AI Experience
embedUR systems designed, optimised, and deployed both showcased demos, integrating real-time voice and vision AI inference entirely on the PSOC™ Edge platform. With deep expertise in embedded software, AI model development and quantisation, and real-time systems, embedUR enables companies to bring high-performance, AI-powered products to market faster—without requiring extensive in-house AI engineering. For Infineon customers, embedUR offers full-stack engineering support from mobile app integration and AI model tuning to deployment and system-level optimisation.

More AI at the Edge: embedUR’s Broader Demo Showcase
In addition to the joint demos with Infineon, embedUR showcased a wide range of Edge AI applications at their own exhibit, running on various industry-leading AI silicon platforms such as the PSOC™ Edge from Infineon. These demonstrations reinforced embedUR’s ability to deliver efficient, high-performance AI on memory-constrained, low-power devices.
All demo models are available through ModelNova, embedUR’s AI resource hub. Developers can browse a growing library of pre-trained Edge AI models to build fast PoCs. When a model fits the target use case, embedUR offers services to help refine, optimise, and deploy it—accelerating time-to-market.