Grant: Collaborative Research and Development Grants

With the increasing popularity of internet video content from video sharing sites such as Youtube.com and from many content providers such as Disney.com, huge amounts of digital visual media data are available online. The popularity of online video content generates a large market demand for video analysis/management applications. Developing and implementing these video analysis/management applications poses urgent technical challenges. This research is a collaborative effort between Ryerson University and Kidobi to develop advanced theories and innovative intelligent technologies in visual media management and services, and tackle specific application needs in commercial environments. The R&D on content-based intelligent video categorization, summarization, retrieval and management for children's visual media is carried out in three tasks: 1) content-based video analysis and feature extraction; 2) intelligent annotation, indexing and categorization; and 3) intelligent on-demand summarization and navigation. This proposed research builds on and extends our existing work in content-based image/video analysis and processing based on statistical modeling and computer vision. Taking advantage of Kidobi's expertise in web-based media and audio-visual psychology, we will formulate a set of innovative technical problems according to specific application requirements and constraints in children visual media management and services. Existing algorithms will be extended and new theories will be developed for specific application requirements. The developed algorithms will be incorporated into Kidobi products.