Human stereopsis contributes an important cue to the depth of visible objects. It depends on a simple geometric property: our eyes are separated laterally by about 6 cm, so each of them sees a slightly different view of the scene in front of us. The difference (or disparity) between the two retinal images of the same object provides information about its distance, relative to other visible objects. The brain can use disparity information with great precision. From a distance of 10 ft, well-trained observers can detect if one object is 1/3 of an inch farther away than an adjacent object. By experimental manipulation of the images seen by each eye in a computer-driven stereoscope, we have been studying the neural processes used by our brains to translate the binocular disparities into a 3D map of the visible world.

Not everyone can utilize stereoscopic information to judge depth. Some people, who, as children, had crossed eyes, or a large difference in the optical focusing power of their eyes, have lost the ability to respond to binocular disparity. This loss is not crucial because there are many other cues to depth (relative size of familiar objects, motion parallax, texture gradients, to name a few). However, many of these individuals also suffer from an acuity loss in the weaker of their two eyes that cannot be corrected by glasses--an abnormality called 'amblyopia' that affects about 3% of the population. Our research group at Smith-Kettlewell has made many different measurements on a large number of these amblyopic individuals. This extensive study has improved the scientific understanding of the amblyopic deficit, and could lead to better diagnosis and treatment in the future.

When you hike in the hills, the images of the trees, bushes, and other features in your surroundings move across your retina continuously due to your own self-motion. Despite this continuous flow of movement across the retina, you can easily detect the sudden movement of a bird or small animal darting across the periphery of your visual field. Why is it so easy to detect this animal movement amidst all the other retinal motion? For one thing, the animal moves in a consistent direction that is different from the other self-induced motions, and for another, the abrupt change seems to alert the visual system to the possibility that an interesting, or possibly dangerous, target is in motion in one part of the visual field. Using highly simplified laboratory displays, we are studying how this alerting mechanism enhances the detection of moving features amidst noisy moving surroundings.

For more information, visit Suzanne McKee's lab web pages.

Collaborators: Doug Taylor, Anna Ma-Wyatt. Preeti Verghese, Andrew Glennerster, Julie Harris, Dennis Levi, Tony Movshon, Scott Watamaniuk, Bart Farell, Anthony Norcia, Richard Harrad, Julia Hale, Laurie Wilcox