An End-to-End Scalable Copyright Detection System for Online Video Sharing Platforms
Daniel Zhang, Jose Badilla, Herman Tong, Dong Wang · 2018
Combating copyright infringing multimedia content has arisen as a critical undertaking in online video sharing platforms, such as YouTube and Twitch. In contrast to the traditional copyright detection problem that studies the static content (e.g., music, films, digital documents), the proposed system focuses on a much more challenging problem: detecting copyright infringements in live video streams. This is motivated by the observation that a large amount of copyright-infringing videos bypass the detector while many legal videos are taken down by mistake. In this paper, we present an end-to-end system that is dedicated to combating the copyright infringements in live video streams. The system to be demonstrated consists of 1) a web front-end for user interaction and customized video query, 2) a scalable and real-time video crawling system that can collect video metadata, live chat messages, and visual content of the live video streams on video sharing platforms, and 3) a novel supervised copyright detection engine that leverages the live chat messages of the audience to detect the copyright infringement of live videos.