Real-time wearable computer vision system for improved museum experience

The goal of this demo is to perform real-time object classification and artwork recognition using a wearable device, to improve user experience during a museum visit by providing contextual information and performing user profiling. We propose the use of a compact CNN network that performs object classification and artwork localization and, using the same CNN features, we perform a robust artwork recognition. Shape based filtering, artwork tracking and temporal filtering further improve recognition accuracy. © 2016 Copyright held by the owner/author(s).