My blog has been moved to a new location!

You should be automatically redirected in 6 seconds. If not, visit
and update your bookmarks.

Friday, February 08, 2008

[Gfx-cafe] GFX Cafe Seminar Friday Feb 8, 2008

GFX Café Seminar Friday Feb 8, 2008
12noon, ECE118

Food will be served

Sensitivity Analysis for Classification
by Joe Kniss, Advanced Graphics Lab

The principal goal of visualization is to create a visual representation
of complex information and large datasets in order to gain insight and
understanding. Our current research focuses on methods for handling
uncertainty stemming from data acquisition and algorithmic sources. Most
visualization methods, especially those applied to 3D data, implicitly
use some form of classification or segmentation to eliminate unimportant
regions and illuminate those of interest. The process of classification
is inherently uncertain; in many cases the source data contains error and
noise, data transformations such as filtering can further introduce and
magnify the uncertainty. More advanced classification methods rely on
some sort of model or statistical method to determine what is and is not
a feature of interest. While these classification methods can model
uncertainty or fuzzy probabilistic memberships, they typically only
provide discrete, maximum a-posteriori memberships. It is vital that
visualization methods provide the user access to uncertainty in
classification or image generation if the results of the visualization
are to be trusted.

In this talk, I'll cover some basic definitions of visualization and
highlight some of our on going work here at UNM. Specifically, I'll
discuss a new approach for assessing uncertainty stemming from algorithmic
or computational sources called sensitivity analysis. I'll also take a
few detours to cover topics volume rendering, the method we use to make
pictures of 3D datasets.

Joe is an Assistant Professor in Computer Science an member of the Advanced
Graphics Lab at UNM. His research focuses on interactive visualization of
large datasets, light transport, and computer graphics. His current
research investigates methods for capturing and displaying uncertainty
introduced by the data processing pipeline, and frameworks for representing
non-linear features in band-limited raster image data.

Pradeep Sen
Assistant Professor
Advanced Graphics Lab
Dept. of Electrical & Computer Engineering
University of New Mexico
gfx-cafe mailing list

No comments: