login

Video scouting demonstration

Published 30 October 2000
Nevenka Dimitrova, Lalitha Agnihotri, Radu Jasinschi, John Zimmerman, George Marmaropoulos, Thomas McGee
Citations7

TL;DR

This Video Scouting demonstration presents a system that allows for the filtering and retrieving of TV sub-programs based on user's content preferences, and personalizes the TV experience in the areas of celebrity and financial information.

Abstract

Smart video content selection and recording is the best selling feature of the current personal TV receivers like TiVo. These devices operate at the TV program level in that they use electronic program guides and user's program personal preferences to help consumers record and watch programs that match their interests. In this Video Scouting demonstration, we present a system that allows for the filtering and retrieving of TV sub-programs based on user's content preferences. The filtering process is realized via real-time video, audio, and transcript analysis. The demonstrator personalizes the TV experience in the areas of celebrity and financial information. The technology can translate into differentiating storage and set-top box product features for finding your favorite actors, most interesting personalized financial news of the day, commercial compaction and enhancement, and content augmentation with other sources of information such as Web pages and encyclopedia. The demonstrator also reflects our active involvement in the MPEG-7 standard (Content Description Interface).

Keywords

Computer Science