This talk summarizes the research on zoomable video done in National University of Singapore in collaboration with University of Toulouse, France, over the last two years. I will first describe how zoomable video can be represented in a scalable and efficient way, followed by a user study on how users would interact with the zoomable video. I will then describe how our findings from the user study leads to an unexpected application in identifying interesting video regions. Finally, I showed how identifying interesting regions can be used for video retargeting and for improving the zoom/pan user interfaces in zoomable video. The talk concludes with an overview of the current research challenges in zoomable video.
Wei Tsang Ooi (黃瑋璨) is currently an Associate Professor at the Department of Computer Science, National University of Singapore. His research interest falls under the area of multimedia systems and networking and is currently investigating into building high quality, easy-to-use, and scalable interactive multimedia systems.
Wei Tsang is an active member of the multimedia research community. Besides publishing regularly, he also contributes as members of the technical program committee and organizing committee for conferences such as ACM MM, ACM NOSSDAV, and ACM MMSYS, as well as reviewers and co-guest editors for journals such as IEEE TMM and ACM TOMCCAP. Most recently, he co-chaired the organizing committee of NOSSDAV 2009 and the technical program committee of NetGames 2011.