|
21: Seeing is Believing: Reality Perception in Modeling, Rendering,
and Animation
Sunday, Half Day, 1:30 - 5 pm
West Hall A
Advances
in image synthesis techniques allow very precise simulation
of light-energy distribution in a scene. Unfortunately,
this does not ensure that the displayed image will have
a high-fidelity visual appearance. This course addresses
techniques to compare real and synthetic images, and identify
important visual system characteristics. The ultimate
result: significantly reduced rendering times. Case studies
involving both static and dynamic images are presented,
and their different perception-metric requirements are
compared and contrasted.
Prerequisites
Appreciation of the need for perception evaluation. No
prior knowledge of metrics is assumed, although knowledge
of human vision and image-quality metrics may be an advantage.
Topics
The nature of images; relevant issues in human visual
perception and their investigation using psychophysical
methods; computational models of perception including
spatial-frequency and orientation channels and visual
masking; computational metrics including visual difference
predictors, the Sarnoff model, and animation quality metrics;
and psychophysical evaluation of image quality.
Organizers
Alan Chalmers
University of Bristol
Ann McNamara
The University of Dublin
Lecturers
Alan Chalmers
University of Bristol
Scott Daly
Sharp Laboratories of America, Inc.
Ann McNamara
The University of Dublin
Karol Myszkowski
Max-Planck-Institut für Informatik
Holly Rushmeier
IBM T.J. Watson Research Center
Tom Troscianko
University of Sussex |
|
