AI algorithms take the Xovis 3D Sensors’ counting accuracy even closer to perfection. “Based on the powerful Xilinx Zynq processing platform, we were able to integrate a Deep Net with a classification accuracy of more than 99.5% on the edge device. Compared with the technologies available today, this process significantly enhances people detection,” comments Xovis co-founder and CTO Markus Herrli.
Moreover, AI-powered Xovis 3D Sensors open up future applications for footfall counting, people flow monitoring and in-store analytics. The retail industry, in particular, will benefit from new possibilities such as returning customer detection, full-path customer analytics and the installation of cashier-free payment systems. At airports, anonymous passenger tracking over long distances will lead to new insights and to a seamless travel experience.
What is AI and Deep Learning?
Artificial intelligence is an algorithmic system that mimics the human brain and its reaction to inputs from the environment. Unlike traditional software, which judges a situation from a limited set of if-then rules, AI concentrates on patterns that are recognizable in a wide variety of settings. Deep Learning is the process by which AI is taught how to interpret inputs from a specific situation and attribute them to an output. The data gathered with a high precision AI-powered people counting and tracking technology lays the foundation for comprehensive in-store analytics.