Chennai (Tamil Nadu) [India], January 4: STMicroelectronics is pleased to welcome embedUR systems, a pioneer in Edge AI and embedded solutions, as an Authorized Partner.

Three cutting-edge Edge AI applications for the STM32N6 platform have been created by embedUR Systems using ST’s hardware and software ecosystem. These examples show how AI may revolutionize a variety of sectors on little, effective technology.

As a result of this partnership, embedUR systems built, trained, and integrated AI models to operate on ST’s most recent STM32N6 microcontrollers, which come with the embedded software stack required for productization. Through embedUR’s extensive knowledge of STMicroelectronics’ platforms, cutting-edge Edge AI applications are being made possible, opening the door for game-changing business solutions.

Solutions for Edge AI

Image Segmentation for People Detection

A cutting-edge image segmentation solution for quick person identification and classification is provided by embedUR. Applications like occupancy counting, presence recognition, crowd control, person tracking, and more are well suited for this capability.

This skill combines:

* YOLACT Adaptation for Neural-ART AcceleratorTM, the STM32N6’s Neural Processing Unit (NPU): YOLACT has been tuned for the STM32N6 hardware, guaranteeing low latency and stable performance, and it supports real-time instance segmentation.

* Exclusive optimization methods: embedUR’s sophisticated quantization algorithms maximize speed and accuracy while reducing model size by around 75%, making it perfect for devices with limited resources.

* Improved picture quality thanks to the STM32N6’s picture Signal Processor (ISP): this integration allows the solution to operate at 71 frames per second and guarantees dependable performance in a range of lighting situations.

Using Enrollment and Facial Recognition

Keyless access systems and biometric boarding at airports, which demand strict security, privacy, and performance criteria, are perfect for embedUR’s on-device face recognition solution since it combines security and fast processing.

Features consist of:

* embedUR’s UReyeD framework: Provides rapid, accurate identification via precise face detection, keypoint extraction, and embedding creation.

* Direct on-device enrollment: This feature improves privacy and streamlines deployment by doing away with the requirement for external infrastructure.

* Support for fast identification in huge databases: built for high-user-throughput settings like public safety and access management.

Denoising of Audio

Additionally, embedUR Systems has developed an audio denoising solution that utilizes the Neural-ART Accelerator (NPU) of the STM32N6. It combines a cutting-edge audio denoising model with the superb audio capabilities of the STM32N6 platform to provide crystal-clear speech even in loud settings. Denoising will be used more and more when systems include voice into text to feed LLMs in order to guarantee that the most correct transcription is provided to LLMs.

Notable traits:

* An award-winning audio denoising model created for GPUs was selected by embedUR, which then modified and enhanced it for the STN32N6’s NPU.

* By executing the model on the STM32N6 NPU, the audio solution may function in a range of high demand and low power conditions by lowering power consumption, improving accuracy, and eliminating delay.

Advancing Edge AI Innovation

“This collaboration with STMicroelectronics is a testament to our engineering excellence and ability to bring up AI solutions rapidly and independently on literally any Edge platform (CPU/GPU/NPU),” stated Rajesh Subramaniam, CEO of embedUR systems “We’re proud to demonstrate how efficient, high-performance AI solutions can be realized on the STM32N6 microcontroller, and the myriad applications they can enable.”

“With little assistance from our technical staff, embedUR Systems has shown its ability to develop AI technology and optimize it for our STM32N6 microcontroller. They are really experienced, creative, and efficient. The remarkable potential of edge AI on STMicroelectronics’ components to propel revolutionary AI solutions across sectors has been shown in a matter of weeks. remarked Miguel Castro, STMicroelectronics’ director of AI strategy and commercial development for the Americas. “We are pleased to announce that embedUR is a new ST Authorized Partner.”

Businesses are encouraged to investigate how embedUR Systems’ knowledge and products might support the next Edge AI applications. For further information, go to http://www.embedur.ai.

A pioneer in edge computing, artificial intelligence, and embedded systems, embedUR Systems is headquartered in Silicon Valley. With more than 20 years of experience and a track record of reducing product time-to-market for semiconductor, telecom, and network equipment suppliers, embedUR provides innovative embedded solutions that power millions of devices globally. With the help of its ModelNova platform, which provides pre-trained AI models for smooth integration into intelligent edge systems, both AI enthusiasts and afficionados may quickly and simply construct proof of concepts with little to no experience with AI modeling. Visit modelnova.ai to see pre-trained Edge AI models right now.