Brief Bio:
Ph.D. Columbia University
Post-Doctoral Research Fellow, Harvard University
Research Staff Member, IBM T.J. Watson Research Center
Manager, IBM T.J. Watson Research Center
Member of the Board, IS&T, International Imaging Society
Founder and Co-Chair, IS&T/SPIE Conference on Human Vision and Electronic Imaging
Longer Bio:
I grew up in Lexington, Massachusetts. As an undergraduate at Brandeis University, I discovered Experimental Psychology, and have been involved in trying to understand human perception and cognition ever since. My first experiments were in problem solving, but as a graduate student at Columbia University, I became increasingly interested in the foundations of visual perception. My PhD thesis used a metacontrast masking paradigm to study interactions in spatial and temporal information processing. As an NIH post-doctoral fellow in the Laboratory for Psychophysics at Harvard, I studied the perception of complex visual and auditory patterns, looking to extend low-level models of visual processing to the analysis of higher-level phenomena. My first job at the IBM T.J. Watson Research Center, was in a display physics group. We developed a psychophysical method for measuring perceived flicker on video displays and studied spatial sampling effects of the then new thin-film transistor liquid crystal displays. We also did experiments on shape perception, color constancy, and even discovered a new visual illusion involving stroboscopic vision ("Bernice's looming effect"). As a psychophysicist working in a computer company, however, my technical horizon was rather narrow, so I was delighted to be offered the opportunity to join the technical staff of the VP for Computer Science, Abe Peled, where I learned about computer science research, about research in the marketplace and about business. I think the main lesson I learned was that every turn of the crank of Moore's law meant new opportunities to create better, more intuitive tools for manipulating and understanding the data processed by these computer systems. Suddenly, we could realistically build data analysis systems which let people do visual "what-if" experiments, or dynamically explore the results of data mining algorithms, or explore links between multiple slices of multidimensional data. Suddenly, it became possible to build interactive perceptual rules into visualization and visual design systems, and to design image query systems based on experiments measuring human judgments of image similarity. Because we can build systems which provide almost arbitrary capabilities, the problem of building such systems has fundamentally changed. We can now build systems matched to human vision, perception and problem solving capabilities...which means that the problems drive the research.