Until the late sixties very little was known about how the brain coordinates eye movements. Consequently, the models proposed to describe the "processing" that the brain does were based on the so-called "black-box" approaches -- having little scientific value. However since 1968, when the first single-cell neural recording took place, a great deal of information has been accumulated regarding the anatomy and the physiology behind eye movements. The work in this lab is aimed at developing better models to describe the workings of the brain in controlling eye movements and their coordination with head movements based on what we have learned from mother nature.

THE VESTIBULO-OCULAR REFLEX (VOR)

Our body relies on a variety of reflexes in order to thrive in a dynamic world. You may have experienced the muscle stretch reflex that a doctor tests by tapping your knee during a routine medical exam. Our eyes are also subject to reflexive mechanisms that are usually taken for granted. One such reflex is the vestibulo-ocular reflex or VOR as it is called.

The goal of our vision system is to keep the image of the world stable on the retinas so that we do not feel nauseous as we go about our daily business. The VOR uses sensory information from semi-circular canals, located in the inner ear, to compensate for head motion. (These canals are also responsible for our sense of balance.) The VOR, for instance, allows you to walk along a street and keep a clear view of the world as your head bounces up and down.

In the lab, we have developed a model of the VOR which relies on the experimental observations recorded by neurophysiologists. Our model is the first in the world which conscientiously tries to mimic the functionality of the brain.

TRY YOUR VOR

ROBOTIC VISION SYSTEM

People who build robots having artificial vision are faced with the same design problems that mother nature had when she designed vertebrate lifeforms. At the very least, a vision system needs to be able to execute slow and rapid eye movements and be able to compensate for camera shake. To date, all of the robotic vision systems have been built using black-box approaches and have yielded poor performance. One branch of research in this lab is to apply biological eye-control strategies to the field of artificial vision.

Presently, the lab has a rudimentary pair of artificial eyes that are used to evaluate the performance of a biological-based control scheme. These eyes are controlled using a computer. The next stage is to place the eyes in a movable head.

BETTER VIRTUAL REALITY