Schedule of Events

Fall 2006 Colloquium Series

Fred Dretske

Duke University (Philosophy)
What Change Blindness Teaches about Consciousness
Bousfield 160
Friday, September 8, 2006
4:00 PM
Abstract

Change blindness is often described as a failure of subjects with normal vision to see visible, sometimes quite conspicuous, objects and properties. The evidence typically used to support this charge of blindness, this alleged failure to perceive things, is a failure on the part of subjects to notice or detect the visible differences. If, though, one doesn't have to notice or detect something, doesn't have to think one perceives it, to consciously perceive it, the striking deficits manifested in change blindness reveal absolutely nothing about conscious experience. They reveal nothing about what a person is visually conscious of. They tell us something, perhaps, about what a subject knows or doesn't know, what he thin ks or doesn't think, but nothing about what he sees or fails to see. This critical assumption--that noticing or detecting x is necessary for awareness of x-is an assumption that many participants in this research would happily admit to making. More often than not it is simply taken for granted. [.] Despite its widespread acceptance, though, I think this assumption is false. It is false in a way that seriously distorts our understanding of what perceptual experience is like. It represents our conscious experience of the world as a much more barren, a more impoverished, a less textured, affair than it actually is. Ridding oneself of this assumption, then, is an important step in getting a clearer picture of consciousness. This paper is an attempt to take that step.

Karen De Valois

University of California, Berkeley
The Appearance of Images
The Alvin Liberman Room, Psychology Building (BOUS 160)
Friday, October 20, 2006
4:00 PM
Abstract

The eyes may not lie, but they can be stingy with the truth. Although we know a great deal about the factors that determine whether an image will be detectable and when it can be discriminated from another, we know much less about what determines appearance. My colleagues and I ask what factors are particularly important in making an image appear to be a good representation of a natural scene. I will demonstrate some of these. For example, you will see a horse running but be unable to tell whether he is running backwards or forwards until certain kinds of information are added. You will see an eagle appear to fly like a hummingbird. A cheetah will be so blurred that he is hardly recognizable until he begins to run. Based on our work on these and related phenomena, I will demonstrate that:

1. A moving image can paradoxically be made to appear sharper when much of its high spatial frequency information is eliminated.

2. Removing the high temporal frequency information from most pixels has little effect on the appearance of a moving image.

3. Though color alone cannot support good motion vision, the addition of color can dramatically improve the appearance of a moving image.

4. The variations in the amplitude spectrum of a still image are of minimal importance in determining its appearance. I will demonstrate a new method of still image compression based on this observation.