Carter Maslan, CEO of Camio, presents the "Blending Cloud and Edge Machine Learning to Deliver Real-time Video Monitoring" tutorial at the May 2017 Embedded Vision Summit.
Network cameras and other edge devices are collecting ever-more video – far more than can be economically transported to the cloud. This argues for putting intelligence in edge devices. But the cloud offers unique, valuable capabilities, such as aggregating information from multiple cameras, applying state-of-the-art algorithms, and providing users with access to their data anywhere, any time.
Camio uses a combination of machine learning at the edge (in network cameras and network video recorders) and in the cloud to generate alerts, highlight the most significant events captured by a camera, and to let users search for events of interest. In this talk, Maslan explores the trade-offs between edge and cloud processing for systems that extract meaning from video, and explains how the two approaches can be combined to create big opportunities.
