Spring 2001 Seminar Series - April 5, 2001
University of Minnesota
School of Statistics
College of Liberal Arts
Issues in Visualization Research:
Conveying 3D Shape, Multivariate Distributions,
and More, via Images
Victoria Interrante
Department of Computer Science and Engineering
University of Minnesota
Thursday, April 5, 2001
4:00 PM,
211
Vincent Hall
Minneapolis, East Bank Campus
Social at 3:30 PM,
300
Ford Hall
Abstract
Visualization is concerned with the design and implementation of algorithms for
creating images that can facilitate our understanding of a set of data. My
research in visualization draws broadly from work in perceptual psychology, art,
computer graphics and computer vision, and is loosely organized around the
central philosophy that there is a definable science behind the art of effective
visual representation, a fundamental theoretical basis for knowing how to map
data to pictures in a way that allows the important information to be easily and
accurately understood.
In this talk I will give an overview of some of my recent research in
information representation, emphasizing fundamental issues in choosing promising
methods and in evaluating their effectiveness. Two of the specific applications
I will discuss include: using color and texture to simultaneously convey
multiple scalar distributions over a 2D map, and using texture to facilitate the
accurate and intuitive understanding of the 3D shape of a smoothly curving
surface.