Scott MacKenzie

Associate Professor, York University

Evaluating Eye Tracking Systems for Computer Data Entry
59 minutes, 27.3mb, recorded 2010-01-27
Scott MacKenzie

What if you could type with your eyes? People with limited mobility may have no other choice. But it's slow, currently around 12-18 words per minute. Is that the best we can offer? Scott MacKenzie describes the physiology of the eye and technical limits on eye tracking, his work to evaluate various input methods, and his new approach, the "scanning ambiguous keyboard," that helps some disabled people communicate more freely than ever before.

The focus of the human eye continuously leaps from point to point in a motion called a saccade. Even when the eye fixates on a point, it is impossible to accurately measure the coordinates. Focus is broad, and the eye keeps the point in view by continually rescanning.

Making a definitive selection with the eye means ignoring likely errors. This limits the accuracy of eye-tracking studies and means using the eye to do computer input is slow, error-prone, and fatiguing. But these drawbacks can be reduced by various means under development, such as predicting what the eye-typist wants to say.

Scott MacKenzie describes his studies desgined to evaluate the various methods of aiding eye-typing.  He introduces a new method, the BlinkWrite "scanning ambiguous keyboard," which narrows the selection to a word or character using linguistic prediction.

When Scott presented his new method at a conference, he inspired a member of the audience who uses voluntary muscle contractions to operate a scanning keyboard. By freely sharing his work and his code, Scott gave at least one disabled person a way to communicate more freely than ever before.

Scott MacKenzie is associate professor of Computer Science and Engineering at York University, Toronto, Canada. His research is in human-computer interaction with an emphasis on human performance measurement and modeling, experimental methods and evaluation, interaction devices and techniques, alphanumeric entry, language modeling, and mobile computing. He has more than 100 peer-reviewed publications in the field of Human-Computer Interaction, including more than 30 from the ACM's annual SIGCHI conference. He has given numerous invited talks over the past 20 years.


This free podcast is from our BayCHI series.

For The Conversations Network: