Colloquia occur: 4:00 pm - 5:00 pm - Room PY 101.
Colloquia titles will be posted as they become available.
Also see: vislab.psych.indiana.edu/~jgold/q733
Invited Colloquia:
- January 24 - Marty Banks, UC - Berkeley
- March 7 - Deb Roy, MIT
- March 28 - Gary Cottrell, UC - San Diego
- April 04 - Lera Boroditsky, Stanford University
- April 18 - David Meyer, University of Michigan
- Host: Jerry Busemeyer
- Title: Contemporary Symbolic Computational Models As A Basis For Fundamental Insights About Human Cognition and Action
- Homepage: www.umich.edu/~bcalab
- abstract
- April 25 - Michael Kearns, University of Pennsylvania
Abstracts
01/24/05: Marty Banks
Why pictures look good when viewed from the wrong place
Pictures have widespread usage because in the convenient format of a 2D surface they allow viewers to perceive the 3D layout and shapes of objects in a scene. Pictures would not be very useful if the observer’s eye had to be positioned at the center of projection, the one place where the image delivered to the eye is geometrically correct. Imagine, for example, there being only one seat in the theater that allowed an acceptable impression of a motion picture. When a picture is viewed from locations other than the center of projection, the retinal image specifies a different layout, but we normally do not notice the change; we achieve some degree of perceptual invariance. Another interesting effect in picture perception is that shape distortions are sometimes perceived even when the retinal image specifies the correct shapes. This is most common with wide-angle formats. Painters, photographers, cinematographers, computer scientists, and vision scientists have for centuries wondered how perceptual invariance across viewing position is achieved with pictures; they have also developed rules of thumb for minimizing apparent distortions. Our work examines the means for achieving invariance and the causes of perceived distortions. We determined the mechanism that underlies invariance and developed a simple quantitative model that is quite consistent with human behavior.
03/07/05: Deb Roy
Meaning Machines
The meaning of words in everyday language use depends on two very different kinds of relations. On one hand, language has referential meaning - words are about objects, events, and relations in the physical world. On the other hand, language has functional meaning -- agents use speech acts to pursue goals in social contexts. Humans effortlessly weave together reference and function to communicate. A grand challenge for the cognitive sciences is to develop a coherent computational framework that simultaneously models both aspects of meaning. I will report on my research group's ongoing efforts to address various facets of this challenge. We have built a series of grounded, situated language processing systems in the form of conversational robots, sensor-grounded cognitive models, and multiplayer language-mediated games. The design and evaluation of these systems is guided by observing situated communication between people. I will highlight some of our implementations, comment on underlying computational pri nciples that are emerging from these activities, and their relevance to understanding human cognitive processes.
03/28/05: Gary Cottrell
What can computational models tell us about face processing?
My lab has been involved in building neurocomputational models of face processing for fifteen years now. Computational models can provide insights into the possible mechanisms of face processing that behavioral, imaging, electrophysiological and single cell recording technologies cannot. In neurocomputational models, one can "hold a stick" into a model cell from "birth" on; one can follow possible alternative life experiences for a cell, and one can explore the representations developed in a whole population of cells. In this talk, I will review some of the results we have obtained over this period concerning holistic processing, the development and role of the Fusiform Face Area and the surrounding controversy, and time permitting, describe some of our future directions in modeling active sampling of images in the form of eye movements.
04/04/05: Lera Boroditsky
How the languages we speak shape the way we think
Do people who speak different languages think differently about the world? Does learning new languages change the way you think? Do polyglots think differently when speaking different languages? I will present several lines of cross-linguistic experiments illustrating how the languages we speak shape the way we attend to, represent, and remember our experiences in the world. The results suggest that the private mental lives of people who speak different languages differ much more than previously thought.
04/18/05: David Meyer
Contemporary Symbolic Computational Models As A Basis For Fundamental Insights About Human Cognition and Action
Numerous insights about human attention, working memory, and executive mental processes have been achieved recently through computational models of multitasking based on the Executive-Process Interactive Control (EPIC) cognitive architecture (Kieras & Meyer, 1997, 2001 ; Meyer & Kieras, 1999). For example, such models have revealed flexible task-scheduling strategies and surprising capacities in parallel information processing. These discoveries are especially relevant for practical applications and interpretations of individual differences in human performance and cognitive aging.
04/25/05: Michael Kearns
Computer Science, Economics, and the Effects of Network Structure
The increasing instrumentation of human and organizational behaviors, via technologies such as the Internet and instant messaging, has led to revolutions of measurement and theory in both computer science and the social sciences. It has also led to an accelerated convergence of interests between the two communities. Areas such as social network theory contain a healthy mix of researchers with computational and sociological concerns, and have resulted in mathematical models of network formation equally applicable to the growth of technological, biological, and human systems.
Many of the most compelling pieces of this convergence lie at the intersection of computer science and economics. The flow of ideas here is notably bi-directional, with computer science contributing new modeling and algorithmic tools to the study of complex game-theoretic and economic systems, and economics providing powerful new ways of thinking about computational issues such as spam and resource management in the Internet. At the core of such efforts is the realization that many of our most beloved and reviled technological systems are, at heart, economic systems, and show all of the associated behaviors: mixtures of competition and cooperation, adaptivity, free riding and tragedies of the commons, and many others.
In this talk I will survey the main trends in this line of thought, and illustrate with several examples, including modeling the effects of network structure on economic equality, and viewing Internet routing and airline baggage security as large-population games.
|
