We built a TinyML AI model zoo, featuring optimized versions of popular models like YOLO and Meta's Segment Anything, designed for embedded hardware. Quickly deploy AI on small devices and speed up your time-to-market.
ARM Processor MLPerf Benchmarking for AI with Integrated NPUs
We implemented the MLPerf benchmark suite on 2 ARM processors with NPUs, porting MLPerf Tiny to an MCU with RTOS and MLPerf Edge to an MPU with Linux, running benchmarks to differentiate chip vendors' solutions.
Advancing Edge Intelligence: Precision Object Detection with Low-Power MCUs
We crafted a cutting-edge microcontroller unit (MCU) that epitomizes smart, energy-efficient object detection at the edge of technological innovation.
AI-Enhanced Wi-Fi Sensing: Revolutionizing Space Monitoring and Security
We've developed a Wi-Fi Access Point system that integrates AI neural networks, bestowing APs with sophisticated sensing abilities.
Slashing Home Wi-Fi Support Calls with AI
Our project has mitigated customer support requirements for a major optical network service provider by employing AI to predict service-related calls with 98% accuracy.






