Title:
TV Watcher: Distributed Media Analysis and Correlation
TV Watcher: Distributed Media Analysis and Correlation
Author(s)
Hilley, David Byron
El-Helw, Ahmed
Wolenetz, Matthew David
Essa, Irfan
Hutto, Phillip W.
Starner, Thad
Ramachandran, Umakishore
El-Helw, Ahmed
Wolenetz, Matthew David
Essa, Irfan
Hutto, Phillip W.
Starner, Thad
Ramachandran, Umakishore
Advisor(s)
Editor(s)
Collections
Supplementary to
Permanent Link
Abstract
The explosion of available content in broadcast media has created a desperate need for applications and prerequisite system architectures to support automatic capture, filtration, categorization, correlation, and higher level inferencing of streaming data from distributed sources. We present TV Watcher, an archetypical example of such an application. TV Watcher performs user-controlled correlation of live television feeds and allows the user to automatically navigate through the available channels based on content of interest. We introduce the Symphony architecture for distributed real-time media analysis and delivery to meet the system requirements for applications with such needs. TV Watcher is built on top of the Symphony architecture, and currently uses closed-captioning information to correlate television programming. We present the results of a user study that shows the correlation engine is consistently able to pick significantly useful and relevant content.
Sponsor
Date Issued
2004-07-08
Extent
550171 bytes
Resource Type
Text
Resource Subtype
Technical Report