[vslist] NIPS 2004 Workshop

[Jason Gold]

**NIPS 2004 Workshop Announcement**

"Statistical, computational, and psychophysical techniques for
inferring features from stimulus classification"

Time: Friday December 17, 2004
Place: Whistler, B.C., Canada
NIPS 2004 Web Site (Registration & Accommodations):
http://www.nips.cc/Conferences/2004/
Workshop Web Site (schedule & overview):
http://vislab.psych.indiana.edu/~jgold/jgold/jmg/nips2004/

Organizers:

Richard Shiffrin, Indiana University, Bloomington
Jason M. Gold, Indiana University, Bloomington
Andrew L. Cohen, University of Massachusetts, Amherst
Florin Cutzu, Indiana University, Bloomington

Workshop Description:

The psychophysics and neuroscience communities have been looking at
correlations of noisy stimulus inputs with behavioral decisions and
neural responses in order to infer the stimulus features that mediate
sensory and perceptual processes. The Cognitive Science community has
been developing related techniques to identify stimulus features
extracted by human observers and testing the rules by which these
features are combined to produce categorical decisions. The
computational/statistical/machine learning community has been
developing optimal and other techniques for classifying target
categories in noisy input. This workshop will bring together
researchers from each of these fields to learn about and discuss each
other's approaches to solving the feature induction problem.

One common theme in all these areas is the difficulty of inferring
features that may represent configural properties, otherwise known as
high order interactions,=A0 when the dimensionality of the data is high,
and when the data are very noisy. For example, this is often the case
for data collected in classification image studies, in which one
analyzes the noise that has been added to stimuli to bring performance
halfway between pure guessing and perfectly correct. But this problem
is quite universal, occurring in image classification, categorization
studies, and neural receptive field data, to take just a few examples.

This workshop is therefore designed to bring together researchers from
diverse areas who are facing similar challenges, in each case trying to
develop techniques to infer features that underlie classifications of
high dimensional and noisy data. Because the workshop speakers,
participants, and attendees come from diverse areas with different
languages (e.g. neural computation, visual psychophysics, human
decision making, machine learning, statistical inference) we have asked
speakers to aim their talks at a general mixture of experts rather than
colleagues in their own local research domain.

Scheduled Speakers:

Richard Shiffrin, Indiana University
"Introduction: Overview of workshop goals"

Richard F. Murray, University of Pennsylvania
"An overview of reverse correlation"

Adrienne Fairhall, University of Washington
"Uncovering neural computation with reverse correlation: biophysics to
psychophysics"

Matthias O. Franz and Bernhard Schoelkopf, Max Planck
"Inferring features as high order pixel interactions in images"

Florin Cutzu, Arnab Dhua, Jason M. Gold, Chen Yu, Andrew L. Cohen and
Richard Shiffrin, Indiana University and University of Massachusetts
"Inferring Image Templates from Classification Decisions"

Peter Neri, University of Cambridge
"General methods for estimating nonlinear operators in simple visual
tasks"

Jack Gallant, University of California Berkeley
"Regression-based approaches to intermediate vision"

Bosco S. Tjan, University of Southern California
"Recovering the template in the face of uncertainty"

Antonio Torralba, Kevin P. Murphy and William T. Freeman, MIT
"Feature sharing for multiclass object detection leads to edge and line
detection"

Andrew L. Cohen, David Ross, Jason M. Gold, Florin Cutzu, Chen Yu and
Richard Shiffrin,Indiana University and University of Massachusetts
"Features from noise, or noise from features?"

--=20
Jason M. Gold
Assistant Professor of Psychology, Program in Cognitive Science
Department of Psychology
Indiana University
1101 East 10th Street
Bloomington, IN  47405
Office: (812) 855-4635
Lab: (812) 856-0365
Fax: (812) 855-4691
Web: http://vislab.psych.indiana.edu/