My research interests span a broad spectrum of the Image Understanding field. The issues of representation and matching lie at the core of it, leading to the following questions:
How can we infer descriptions of objects (world-centered) from images (viewer-centered)? How can we "recognize" objects we have never seen before?
In addition, I am interested in the design and implementation of very reliable vision systems to accomplish difficult tasks, using imperfect modules.
I am also interested in establishing bridges between Computer Vision and Computer Graphics.
Ph.D. in Computer Science, 1983, USC, Los Angeles, CA.