1950 University Ave, Suite 200
Berkeley, California 94704

(Free and open to the public - and followed at 4pm with a Social Jam that includes refreshments - and beer. It would be helpful if you mark attending/watching above).

Paper-Based and Mobile Multimedia Document Browsing Techniques

In contrast to conventional two-dimensional text-based documents that are easily browsed in limited amounts of time, video and audio files are difficult to skim because they are inherently one-dimensional. We've attempted to overcome this limitation with an approach we call Video Paper that summarizes multimedia recordings as documents that show what's in a recording and provide random access to the multimedia data. Several formats for the documents will be discussed including a multimedia newspaper we call Media Times. In addition, our experience using tools for creating Video Paper documents will be presented including a PC desktop application and a digital photocopier. A recently developed bar code-based retrieval interface for Video Paper will be demonstrated on a Windows Mobile Treo.

A method called Multimedia Thumbnails for viewing high resolution multi-page documents on devices with small displays will also be presented. In this technique, we convert a text-based document into an audiovisual clip of fixed duration that can be viewed on a handheld device. The visual channel presents important visual document parts such as titles and figures by zooming and panning. The audio channel presents synthesized audible document information, such as keywords and figure captions.

The work described in this presentation is joint with Jamey Graham and Berna Erol.

References for this presentation include:
J. Graham and J. J. Hull, "The Video Paper Multimedia Playback System," ACM Multimedia, Berkeley, CA, Nov. 2-7, 2003.

B. Erol, K. Berkner, S. Joshi, and J. J. Hull, "Computing a Multimedia Representation for Documents Given Time and Display Constraints,"
IEEE Int. Conf. on Multimedia and Expo (ICME), Toronto, Canada, pp.
2133-2136, July 9-12, 2006. Winner of the Conference's Best Paper Award.

Jonathan J. Hull is the Manager of the Multimedia Document Analysis group at the Ricoh Innovations, Inc. California Research Center in Menlo Park. His research interests include multimedia applications, document analysis, computer vision, and novel uses of paper.
Dr. Hull is a Fellow of the International Association for Pattern Recognition (IAPR), and he's an Associate Editor of Pattern Recognition as well as Computer Vision and Image Understanding. He's a past Associate Editor of the IEEE Transactions on Pattern Analysis and Machine Intelligence. Before joining Ricoh in 1994,
Dr. Hull was a Research Associate Professor in Computer Science at the State University of New York at Buffalo. He is an inventor on over 40 issued and 70 published patent applications. His academic publications include more than 100 journal articles and conference papers.

More information can be found at http://www.crc.ricoh.com/~hull

Yahoo! Research Berkeley's Brain Jam is an "almost weekly" speaker series on topics related to media, social media, mobile media, media annotation, and the leftist media. Well, maybe not the last one. To join our mailing list, please visit:

Added by berkeleybohemian on February 22, 2007