Stanford PhotoBrowser

As increasing portions of consumer and professional photographs are shot with digital cameras, the tedium of managing these online images ressembles that of traditional paper photos. The Stanford PhotoBrowser explores how a lifetime's worth of digital photographs can best be browsed, searched, and captioned with minimum effort.

This PhotoBrowser video presents a first step in this exploration. The video must be viewed with a Quicktime player. Open Quicktime then File-->Open URL. Paste in this URL.

rtsp://limpet.stanford.edu/photoBrowserV1-3.mov

For this initial step we constrained our designs to require no human organizational effort at all. We challenged ourselves to create photo browser designs that use only information that is automatically available for digital photographs.

We show two designs, our Calendar Browser, and our Hierarchical Browser. Both browsers are intended for individual photographers. The two browsers use a clustering engine as a computational substrate. This engine clusters the images recursively by time. The intuition behind the clustering is that a photographer takes pictures in 'bursts'. One burst might be a birthday, another burst could be a vacation. By recursively isolating these batches of photographs at increasingly fine granularity, we can organize the images by photographic themes.

Publications

Adrian Graham and Hector Garcia-Molina and Andreas Paepcke and Terry Winograd, "Time as Essence for Photo Browsing Through Personal Digital Libraries", Proceedings of the Second ACM/IEEE-CS Joint Conference on Digital Libraries, July 14-18 2002, p.326-335. Also as technical report 2002-34.

Karen D. Grant and Adrian Graham and Tom Nguyen and Andreas Paepcke and Terry Winograd, "Beyond the Shoe Box: Foundations for Flexibly Organizing Photographs on a Computer" January 28, 2003. Technical Report 2003-3