Development of a Web Browser-based Character in Video Metadata Generation Tool

Abstract

Until recently, activating in video platforms and streaming services, the need for managing metadata of videos is increasing day by day to offer deep learning services using these videos. The broadcaster manages the metadata of videos using the ‘Digital Asset Management System(DAMS)’, a program that saves and manages the metadata of videos. However, since non-professionals cannot use the program for videos they upload directly, they must manually create metadata for each video. But this method is inefficient. Therefore, this paper proposes detecting faces in a video, then recognizing the faces and automatically generating metadata that indexes the person and summarizes the time that the person appeared in the video through a web browser.