This CAREER award provides support for the construction of a research and education program in visual information processing. This field is inherently interdisciplinary, and includes portions of computer science, electrical engineering, applied mathematics, psychology, and neuroscience. Yet, educational curricula often do not cross these disciplinary boundaries. One objective of this award is the construction of an educational curriculum, starting next year with the introduction of two innovative foundational courses. The first, an undergraduate lecture course entitled "Foundations of Vision: Perception and Computation" will cover a broad range of material in vision science, from perceptual psychology, to single-cell physiology, to computational theories of visual processing. The second, a graduate lecture/laboratory course entitled "Fundamentals of Image Processing and Computer Vision", will cover basic tools, both theoretical and applied, that are common to these two fields. The second objective of this award is the study of a set of basic and applied research issues in visual image processing. The research will be based on the family of "steerable pyramid" image representations, and will include the use of this representation in an adaptive context, rotation-invariant pattern matching, noise removal and enhancement, representation of texture and orientation content, and differential recovery of three-dimensional scene structure.
