On-demand webinar

Plumerai Builds World's Most Precise Person Detection Model Deployed on an Alif Semiconductor Microcontroller

This webinar was broadcast on December 15, 2023

This field is for validation purposes and should be left unchanged.

Learn how to improve inference speed for AI-Powered Person Detection

The market interest in computer vision applications for smart home, smart retail, and smart city is rapidly increasing.  These applications are now advancing to include sophisticated functions such as identifying familiar faces, presence detection, people counting, pet detection, and more. As capabilities grow, so does the need for fast, efficient, and low power inferencing.

In this session, you will hear from Plumerai, a creator of edge AI solutions designed for smart home and IOT applications, as they walk you through the steps to design & train an AI person detection model, and then integrate & port it to an Alif MCU. The combined solution provides accurate computer vision AI in a small form factor, which can run on battery for years.

You will also see the performance improvements that can be achieved when porting Plumerai’s person detection model to the Alif Ensemble microcontroller, which has the Ethos-U55 on board. Inferencing time and latency is significantly improved than what has been measured on the Cortex-M7, Cortex-M85, or Cortex-A53. Faster inferencing enables larger and more accurate AI models, the ability to run multiple AI models, and allows the device to return to sleep mode quicker, saving energy.

Meet Your Experts

Marco Jacobs

Head of Product Marketing, Plumerai

Henrik Flodell

Senior Marketing Director, Alif Semiconductor