An Image Classification-based Approach to Automate Video Playing Detection at System Level

Abstract

Tech distraction has become a critical issue on people’s work and study productivity, particularly with the growing amount of digital content from the social media site such as Youtube. Although browser-based plug-ins are available to help block and monitor the sites, they do not work for all scenarios. In this paper, we present a system-level video playing detection engine that captures screenshots and analyze the screenshot image using deep learning, in order to predict whether the image has videos in it or not. A mobile app has also been developed to enable parents to control the video playing detection remotely.