Get real-time insight into how people move through your space
Exploring real-time people analytics at the edge? In this demo, we run a YOLOX-S-Lite–based people tracking pipeline on our MitySOM-AM62A edge AI camera platform. The SOM executes detection and tracking on-device, enabling per-frame person counts, occupancy metrics, dwell time, and heatmap generation without a backhaul GPU. Engineers get an ARM-based SOM with integrated accelerators, camera input, and a software stack that can be adapted for custom models or pre/post-processing. Ideal if you’re evaluating embedded vision for retail analytics, building automation, or safety systems and want to stay within tight power and thermal budgets.
Check out the MitySOM-AM62A System on Module
See more MitySOM Demos: www.criticallink.com/demos
Use our SOM Selector