Cognitive Science Research Talk: New Approach to Stereo Vision
College of Science Cognitive Science Research Talk
New Approach to Stereo Vision
Dr. Paul Linton
Presidential Scholar in Society and Neuroscience, Columbia University
Abstract:
Stereo vision (depth perception based on the different views of the two eyes) has a key role to play in the perception of 3D scenes in virtual and augmented reality, as well as our perception of the everyday world. However, the argument of this talk is that the mechanisms underpinning stereo vision are poorly understood, and a new approach is required. First, our ability to use stereo vision to estimate the size and distance of objects is often attributed to ‘vergence’: the neural signal that registers the fact that the closer an object is, the more the eyes have to rotate in order to fixate on it. However, in a series of experiments, I demonstrate that vergence does not appear to affect either our size or distance perception, and I instead advance a new theory of size perception in stereo vision. Second, stereo vision is thought to be engaged in ‘triangulation’: mapping from the position of points on the retina to the position of points in the world. This requires converting retinal ‘disparities’ (the differences in the two retinal images) into world coordinates. However, in pilot experiments I find no evidence for this remapping, challenging some of our most fundamental assumptions about stereo vision.
Intended Audience:
All are welcome.
To request an interpreter, please visit myaccess.rit.edu
Event Snapshot
When and Where
Who
This is an RIT Only Event
Interpreter Requested?
No