Computer Systems Laboratory Colloquium

4:15PM, Wednesday, June 05, 2002
NEC Auditorium, Gates Computer Science Building B03
http://ee380.stanford.edu

Intelligent Image Processing
Eyeglass-based computer vision systems based on the EyeTap principle

Steve Mann
University of Toronto
About the talk:

EyeTap devices are devices that cause the eye itself to, in effect, function as if it were both a camera and display.

Such computer vision systems, built into eyeglasses, can be used to assist the visually impaired, or those suffering from visual memory disability. For example, the wearable face recognizer allows the wearer to see a virtual name tag appear within their field of view.

Eyetaps are also obviously useful in a utilitarian sense, e.g. for refrigeration engineers and other workers seeking advice from a remote expert.

EyeTap systems blur the boundary between cyberspace and the real world, and allow us to see better, and to communicate better.

Like the invention of shoes, clothing, ordinary eyeglasses, and other prosthetics, we will become dependent on these forms of technology as we evolve into becoming "cyborgs", part human, and part computer. Just as we, as a society, have all but forgotten how to survive naked in the wilderness, we may, in the future, also experience difficulties if we are ever "unplugged", perhaps after having worn a computer vision system for more than 20 years.

Issues of discrimination may come into play, until enough of us become one with such technology. However, in the face of such inconveniences as restriction on travel, new ways of remotely delivering lectures and other presentations may make physical travel of the body completely unnecessary.

Like any other new invention, the EyeTap device raises many new questions, from the right to see (e.g. licensing issues, etc.), to our growing dependence on technology, and the need to make these systems function reliably, both technologically, as well as sociologically.

Thus, along with new mathematical theories in the area of Intelligent Image Processing, there are also a whole host of important new questions that have yet to be answered.

About the speaker:

Prof. Steve Mann was born in Hamilton Ontario, Canada where he spent much of his childhood days building electrical systems onto his body, making his own computerized clothing. By the mid 1980s, while still an undergraduate student of physics and electrical engineering, his work began to have an influence on the fashion industry and by 1985 a solo exhibit of his "painting with lightvectors" images (the world as seen through his computerized eyeglasses) resulted in commercial requests for his cybernetic art. He is also the inventor of the chirplet transform, a new mathematical theory that he developed for his BlindVision system (a wearable radar for the blind). In 1991 he took his inventions to the Unites States where he completed his PhD at M.I.T. in 1997. His recent books describe this work in more detail. For more information, see:

Contact information:

Prof. Steve Mann
10 King's College Road, Room 2001
Toronto, Ontario, Canada,
M5S 3G4
(416) 946-3387
(416) 971-2326
mann@eecg.toronto.edu