We have presented several multimodal content processing methods for story segmentation. Some of these algorithms are optimized for content from specific genres, while others are general-purpose and have wider applicability. We have shown how the capability of automatically segmenting multimedia content into semantically meaningful units can enable novel applications. We presented one such application involving multimedia personalization based on user profiles. The techniques presented simplify the user interaction and permit implementations on a wide range of client devices with various user interface and networking capabilities.
Ken Huber and Bob Markowitz conceptualized several new services made possible by this technology and provided support for the prototype development. Gary Zamchick is largely responsible for the user interface design. Michael Zalot and Antoinette Lee contributed to the design realization. Larry Ruedisueli is responsible for the video endpoint.