embedUR Systems, a prominent provider of embedded solutions and Edge AI technologies, has joined the STMicroelectronics Authorized Partner program. Leveraging ST’s hardware and software ecosystem, embedUR has created three advanced Edge AI applications designed for the STM32N6 platform, showcasing the power of AI in delivering efficient, compact solutions across various industries.
Through this collaboration, embedUR has optimized AI models, developing, training, and integrating them to function seamlessly on ST’s STM32N6 microcontrollers, supported by the necessary embedded software stack for productization. The partnership emphasizes embedUR’s expertise in utilizing STMicroelectronics’ platforms to drive groundbreaking Edge AI applications that enable transformative solutions for different sectors.
embedUR has developed a high-speed, accurate Image Segmentation solution for detecting and classifying individuals, ideal for applications such as occupancy counting, crowd management, and person-following. Key features include the YOLACT adaptation for STM32N6’s Neural Processing Unit (NPU), proprietary optimization techniques to reduce model size by 75%, and enhanced image quality through STM32N6’s Image Signal Processor (ISP), ensuring robust performance even under varying lighting conditions.
embedUR’s facial recognition solution enhances security and speed, perfect for applications like biometric boarding and keyless entry. It integrates embedUR’s UReyeD framework for facial detection and recognition, supporting on-device enrollment, which enhances privacy by eliminating the need for external infrastructure. Additionally, the solution supports large-scale databases, offering rapid identification in high-traffic environments such as access control systems.
The audio denoising solution from embedUR, optimized for STM32N6’s NPU, delivers clear speech even in noisy environments. By adapting an award-winning audio denoising model for the NPU, embedUR achieves lower power consumption, increased accuracy, and minimal latency, ensuring reliable performance in both high-demand and low-power scenarios.
Rajesh Subramaniam, CEO of embedUR Systems, highlighted the company’s ability to develop rapid, efficient AI solutions across any Edge platform, emphasizing the STM32N6 microcontroller’s versatility. Miguel Castro from STMicroelectronics praised embedUR’s expertise in optimizing AI technology for the STM32N6 with minimal support, recognizing the immense potential of Edge AI for transforming industries. As part of this collaboration, embedUR invites businesses to explore new opportunities in Edge AI and the next generation of intelligent applications.
embedUR Systems, headquartered in Silicon Valley, is a leading innovator in embedded systems, AI, and Edge Computing. With more than 20 years of experience and a strong history of accelerating time-to-market for telecom, network equipment, and semiconductor companies, embedUR provides advanced embedded solutions that power millions of devices globally. Its ModelNova platform offers pre-trained AI models designed for easy integration into intelligent edge systems, allowing AI enthusiasts and developers with minimal AI modeling experience to quickly create proof of concepts and experiment with AI technologies.
