Electronic Glove Translates Sign Language

From: Betterhumans, Canada - Aug 18, 2003

Gabe Romain
Betterhumans Staff

An electronic glove designed to translate the complex hand movements of American Sign Language into spoken word and text is under development and could be on the market next year.

Doctoral candidate Jose Hernandez-Rebollar from The George Washington University in Washington, DC has created what he calls the AcceleGlove to decode the fast hand movements of ASL .

The glove uses sensors to map the movement, position and orientation of the hand and fingers.

This information is then analyzed by a microcontroller to find the correct letter, word or phrase associated with the hand movement.

Once the sign is recognized, the hand movements are converted to text or speech.

ASL-to-English conversion

Hernandez-Rebollar, although not deaf himself, realized that while there are dictionaries to translate almost every spoken language, there was no electronic means of translating ASL.

ASL is the fourth most used language in the US, and the National Campaign for Hearing Health says that 28 million people in the US alone suffer from some form of hearing loss.

"I started thinking about an electronic translation method for ASL before I came to GW," says Hernandez-Rebollar. "The language has been around for almost 200 years, yet unlike most other languages we do not have electronic translation for ASL."

Letters, words and phrases

The AcceleGlove can translate signs that correspond to every letter of the alphabet, allowing any word to be spelled out.

So far the glove can also translate 173 words, as well as a few expressions such as "What's the matter?" and "I'll help you."

The glove accurately translates easy words more than 95% of the time and has an accuracy rate of 60% to 70% for harder words.

Military and teaching

The glove could be used in applications other than helping the hearing understand the hearing impaired.

Communicating through hand gestures could be used to help soldiers in the military, for example, and for teaching ASL.

"The thing that makes Jose's research so interesting is that it is applicable to so many different areas," says Robert Lindeman, assistant professor of computer science at GWU.

Hernandez-Rebollar plans to extend his research by improving the algorithms used to recognize hand and finger movements.

He thinks that a right hand glove could be on the market by next year, and he plans to develop a two-handed version that could be ready as early as 2005.

Copyright © 2002-2003 Betterhumans