My research interests lie in computer vision and multimedia. I am interested in the design of new algorithms to extract content from imagery and in the construction of high-performance systems that manipulate imagery. While large collections of digital imagery are becoming commonplace, the tools available for accessing image databases are still quite primitive. Content-based access will be essential for large image databases.
My work has addressed a number of methods for extracting content from images. While most multimedia researchers have relied on color histograms to solve these problems, we have explored a variety of more effective schemes. For example, we have developed methods for segmenting a video into scenes by automatically detecting production effects (such as cuts, dissolves and captions). After the video has been segmented, one frame per scene can be inserted into an image database. The database can then be queried for other similar images, using new methods that combine color information with spatial layout.
Another goal of my research is to make it easy and inexpensive for computers to compute with images. Toward that end, I am exploring architectures for high-performance video-processing systems. Thorsten von Eicken's research provides an exciting platform for this research. I am currently directing several students who are designing a high-performance vision system based on a network of commercial workstations. I am hopeful that this project will yield a system that is significantly more efficient, flexible, and inexpensive than current ones.
I have recently been thinking about the economic impact of freely available pricing information on the Web. My essay on this subject appeared in the electronic newsletter The Network Observer in March 1996. I believe that the Web will have a major impact in the next decade on the way in which goods and services are bought and sold.
If you have questions or comments please contact: email@example.com.