Gary Meyer

Gary Meyer's research involves color synthesis and color reproduction techniques for computer graphics. Meyer takes advantage of what is known about the human color vision system to both improve the efficiency and realism of synthetic image generation techniques and to increase the accuracy and quality of the color reproduction accomplished by computer graphic devices. Most of his work in realistic image synthesis focuses on replacing the explicit simulation of a camera with an imaging technique that incorporates more of what is known about the human visual system. In this way, he hopes to avoid some of the artifacts inherent in photographic techniques and to develop more efficient image rendering algorithms. He is also working to advance the state of the art in synthetic image generation by simulating the mechanisms in nature (such as refraction and interference) that determine color. This work could lead to new ways of modelling the spatial and the spectral distribution of light reflected from color coatings such as pearlescent paints and ink on paper. In all of his research, Meyer is concerned with the proper calibration of the color reproduction equipment that he uses. This has led him to employ computer graphics to develop novel ways of visualizing the gamuts of color reproduction devices.

Paul Schrater

Recent years have seen a rapid increase of computer applications that require information to be processed or presented in a manner that matches or mimics human perception. Content-based web retrieval, data mining, data visualization, virtual reality, human computer interfaces, human-like computer vision, computer graphics, etc., all require the incorporation of knowledge of human perceptual processes in order to be effective. However, very little of this research takes advantage of recent progress made in developing quantitative, predictive models of human perception. Schrater's main research interests involve the application of ideas from statistical pattern recognition and probabilistic AI to human and computer vision and motor control. The broad goals of this research include the development of quantitative models of human perception that can be used both to predict a person's interpretation of a display as well as to design display systems that maximize the information relevant to the user's and designer's goals for the system.

Eric Van Wyk

Programming languages and programming tasks are rarely a perfect fit. Many programs could be significantly clarified by introducing language features specific to the programmer's problem domain. Although domain specific languages do just that, they lock the programmer into a language specialized for a single domain while many problems straddle several domains. A better solution is to allow a programmer to extend a general programming language by importing various domain specific language features into a general language framework. For such a solution to be feasible, programming languages must be implemented in a highly modular manner, and part of Van Wyk's research investigates ways that this modularity can be realized. He also wants to ensure that programs written in extensible languages are as efficient as those written in traditional languages. Thus, language extensions must provide more than just convenient notations but must also provide their own optimizations. Since the designers of such domain specific features may not be experts in programming language implementation, Van Wyk also investigates concise declarative techniques for specifying domain specific optimizations.