A "TALKING GLOVE" : THE "ACCELEGLOVE"

by Stan Griffin

Since 1982, scientists have been creating models or prototypes for gloves that can translate the rapid hand movements of American Sign Language into spoken and/or written speech. Complex, ever-changing technology and prohibitive research costs have kept them from becoming a success.

A year ago one was developed that could translate hand movements for all 26 letters of the alphabet; but spelling words out one letter at a time is " ... a painfully slow process for communication ..." In practice, a single gesture in A. S. L. can represent complete words or phrases.

According to its developer, the newest glove can translate close to 200 words, even a few short phrases, " ... both dynamic gestures and the alphabet ... and more complex arm and body motions ..." It is correct in translating "easy" words 95% of the time and "hard" words 60% - 70%. He calls it the "most robust" of those produced so far.

Jose Hernandez-Rebollar, a researcher from George Washington University, has this year reported his work on the latest variation of "the glove." A 34-year-old native of Mexico, he has conducted public demonstrations of the "AcceleGlove." He came to Washington through the Fulbright Program which provides grants for graduate students and teachers to study abroad. His field is electrical engineering, and the glove was his doctoral project.

Hernandez-Rebollar is not deaf himself, but he said work on his glove was driven by a desire to help others, to " ... produce something ... deaf people can use in everyday life." He spent three years working on it.

The "AcceleGlove" is essentially a wearable computer with super-small electronic circuitry. On each finger and the back of the palm is an accelerometer. Sensors in the glove work with a micro-controller attached to the arm. Information received (signals from hands and fingers) map the placement of the arm and movement of fingers as they "speak" A. S. L.

That information is turned into data a computer can read and then convert to words heard over a loudspeaker or read on a computer screen. It takes only milliseconds to make the sign (detection), recognize the sign (classification), and see or hear the corresponding word.

As things stand now, the glove is not viable as a commercial product. Future research would have to center on: (1) reduction of the hardware’s size and weight; (2) addition of more gestures and translations to the "dictionary" of available words (Cataloging and translating a language is very difficult; the "Oxford English Dictionary" took 40 years); and (3) development of gloves for both hands since some signs require two.

The military is also looking into similar technology to silently help soldiers in combat. If successful, they could send soundless wireless orders to troops with uniforms having small pager motors sewed into them.

There is an estimated 500,000 to 2,000,000 A. S. L. speakers. It is the fourth most-used language in the U. S. and has been around for 200 years.

Many in the deaf community, particularly the "National Association of the Deaf," give the idea a "lukewarm reception." Some simply do not want to use such a device. "For people who are highly educated, how would the glove keep up with the nuances and the changes of the language?" Besides, they say, it can translate only a few hundred words.

Another angle was expressed this way: "The glove would be a setback for the hard of hearing ... Deafness is a unique trait, not a deficiency ... We should push the hearing world to embrace deaf individuals, not force them to use technology (the glove, cochlear implants) to accommodate the hearing world."

Opposing those ideas was this question: "If new techniques can help improve your life and make you feel better about yourself, why not use it if it’s available?"