A video editing support system using users' gazes

Abstract

Digital video cameras have become popular consumer items, and users are now able to create and view a lot of video contents. However, there is not an environment to view and edit them efficiently. That is, the editing environments available today are not easy enough for a general user to use all the tools effectively. Yet it seems logical that if people could edit video easily, they would produce more and more edited video. Thus, an easy-to-use editing environment would have mass appeal. In this paper, we propose an easy method of editing video contents by using the user's gaze when viewing a video. There are three factors with regard to editing video in it. With them we transform raw video contents together with information on multiple users' intentions into a semi-edited video. The semi-edited video is not a completely edited, but is ready to be easily edited by a method that can be extensively varied by repeated viewing. The paper discusses a prototype system that supports viewing such video and describes the method to recognize users' intentions.