Investigations into new multimedia and video modalities

点击翻译按钮获取中文摘要

Since Edison Labs developed the first video camera (the Kinetograph), society has embraced video cameras through home movies, entertainment, education and security. From the use of cameras to create home movies, to our daily use of television, to blockbuster films created purely for entertainment value, video plays an important role in our every day life. Just two decades ago, we can easily remember having only a few television channels, whereas today we have access to hundreds of stations from around the world. This glut of video information introduces unique problems and research opportunities in the manipulation, management, indexing, browsing and searching of the volumes of video data we have at our finger tips.  However, unlike its textual counterpart since the advent of hypertext and the World Wide Web, there is little interactivity in video and it remains a static medium from the viewer standpoint.  This is to say we watch video; we do not interact with video! The core objective of this research program is to develop systems to enable radical changes in the new generations of computer based television (such as HDTV and digital television) so that it becomes an interactive medium.  Through continuing research into computational video techniques and the novel application of hyper-media methodologies to video data it is expected that the research results will help to make video become as ubiquitous and easy to access an information source as text and the World Wide Web is today.  This research represents the beginning of a new modality through which we will experience video. It will facilitate changing users from simply being static observers of video on a television set or monitor to being a dynamic and interactive part of a system where input and feedback is applied to adapt the system to meet current user whims.