In this session, you will learn about the challenges of creating a cloud-based video analytics service that can easily scale to process hundreds of thousands of cameras in real-time, by utilizing state-of-the-art AI running on GPUs. Most video analytics services in the cloud work in a "batch" format (offline), whereby a video clip is uploaded to the service, analyzed, and then results are delivered to the user. Performing video analytics in real-time on a large number of continuous video streams, and with low latency, poses a significant engineering challenge. Learn how Agent Video Intelligence has overcome these challenges to create innoVi video analytics service.
Hall: Hall G
Track: Smart Cities