Biophysically-based appearance models: the bumpy road toward predictability

  • Authors:
  • Gladimir V. G. Baranoski

  • Affiliations:
  • University of Waterloo, Waterloo, Ontario, Canada

  • Venue:
  • ACM SIGGRAPH ASIA 2009 Courses
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

This course addresses practical issues involved in the development of biophysically based appearance models. Because these models are used not only in computer graphics, but also in other scientific applications (for example, noninvasive diagnosis of medical conditions and remote sensing of natural resources), the course also aims to foster cross-fertilization with these fields. The course begins by providing a concise biophysical background and discussing the key concept of predictability. It continues by examining the specific constraints and pitfalls found in each of the key stages of the simulation framework (data collection, modeling, and evaluation) and discussing alternatives that could improve the fidelity of the entire process. Once a model is designed, implemented, and evaluated through a sound methodology, its scope of applications can be expanded to address a wide range of scientific questions. For example, computer simulations are regularly being used by life science researchers to understand and predict material-appearance changes prompted by mechanisms that cannot be fully studied using traditional experimental procedures. The course closes with an examination of recent examples of computer graphics appearance models that can also be employed in such interdisciplinary research efforts.