- The Washington Times - Wednesday, September 3, 2003

Long before he became a doctoral candidate at George Washington University, Jose Hernandez-Rebollar had wondered about the possibilities of creating a way for deaf people to translate sign language into sound by electronic means.

However, according to a university publication that first wrote up the story, his dissertation director in the School of Engineering and Applied Science cautioned Mr. Hernandez-Rebollar about taking on the project, calling it risky because it was so experimental.

That was three years ago. The Mexican-born Mr. Hernandez-Rebollar persisted in spite of the warning, and the result is the AcceleGlove, a device that can turn some of American Sign Language (ASL) into sound and text. With the help of the Fulbright scholarship that brought him to GW and, later, an assistantship position in his department, he also has earned his doctorate.

He now works for the Wheaton-based Institute for Disabilities, Research and Training Inc., which is seeking funds for further development of the idea he made a reality. AcceleGlove in its present version is just a prototype. Its name comes from the hardware called accelerometers that act as sensors to pick up movements made by the signing hand as it forms letters and words. A microcontroller analyzes the movements his software program uses to find a matching word.

The challenge at the outset was how to capture the complex motions involved in signing and relay their positions in electronic form to a computer that could produce both a voice translation and a word in writing on the screen.

He used an ordinary thin leather glove, like a driving glove, attached to some small, fancy-looking circuitry fueled by a special software program. The assembly includes a battery and some lightweight elastic bands that hold the wires onto a person’s right arm.

He certainly wasn’t on a crusade to “cure” deafness, as one anonymous critic later assumed. Nor did he think his invention would instantly change the world of people who rely mainly on signing to connect with others. He doesn’t suffer from hearing loss himself, nor did he know well anyone similarly handicapped.

However, his AcceleGlove expands possibilities for those interested in education and communication among the deaf and could be of critical help in emergency situations in which a deaf or hard-of-hearing person otherwise might not be understood.

In spite of the considerable national publicity, the translation device has received during the past few months, it has yet to be considered commercially viable. One reason is that the prototype works on the right hand only, which limits the number of words available to translate to just fewer than 200. ASL users normally depend on both hands for a full-fledged conversation.

“I’ll be happy to see it working even with reduced words for somebody who wants to express himself even with a limited vocabulary,” he says. “It’s not up to me who is going to use this in the future. I just presented this as a solution to a problem. It’s up to anybody to take this and apply it.”

AcceleGlove has some other drawbacks, as well. The apparatus, in its present form, is an awkward and unattractive piece of equipment that requires at least a two-day training period to learn to operate properly.

“I received one e-mail saying, ‘The problem is that [the device] is ugly and no one is going to wear this thing,’” Mr. Hernandez-Rebollar recalls. “And a girl sent an e-mail saying, ‘What made you think I will wear this when I am completely dressed up?’”

The circuitry is very light, he argues; even the battery is not very bothersome. “It wasn’t to solve a fashion problem,” he says, “but [the problem of] recognition of hand signs applied to American Sign Language. You know there are so many different sign languages in the world.”

He says that, in theory anyway, the ASL translation device could be applied to many other sign languages.

Signing works by using fingers and hands to signal words and letters of the English alphabet in rapid-fire motions to another person who understands the signaling system.

Because verb tenses — and sometimes even the letters — of other languages differ somewhat, so do the sounds that words make in other spoken languages. Everyday expressions conveyed in simple — often silent — motions often are different, as well, making the translation challenge a cultural matter, too. For instance, a native of one country points a finger at his chest to indicate that, yes, he is the person being addressed, while the native of another country customarily points to his nose.

This makes it difficult to apply the device universally as it originally was conceived without changing the mechanism involved.

“What I was looking for was the structure,” is how Mr. Hernandez-Rebollar explains his quest. He had to break down the hand shapes used in signing. Signals convey movement, orientation and the position of the hand and fingers in relation to the body. “I saw that there have to be at least four components, and if you know all that, you can describe almost any sign,” he says.

This wasn’t a novel set of observations, he points out. “Somebody else was already thinking about this in the 1960s,” he says. “We were going one step further, developing the instrumentation to prove that concept. My work was to prove the theory and show the instruments.”

Mr. Hernandez-Rebollar, 34, has the patent for his invention and is applying for a trademark on its name.

He discovered the privately run Institute for Disabilities, Research and Training (IDRT) when he needed help learning what he calls the terminology of sign language because his dissertation had to include that information. His relationship with the company will continue even after he leaves for a job he has accepted at the National Institute for Astrophysics, Optics and Electronics in Puebla, Mexico, his hometown, where he completed undergraduate and master’s degrees.

The institute’s primary mission is developing educational materials for people with disabilities, mostly the hard of hearing. Much of its funding comes through grants for research projects from the federal Department of Education, where IDRT President Corinne K. Vinopol says she intends to appeal for help in advancing research on the glove.

“What needs to be done is expand its vocabulary and grammatical ability, refine the electronics and put it onto two arms,” she says.

Mr. Hernandez-Rebollar’s recent project at IDRT was helping hard-of-hearing people who rely on expensive TTY (teletypewriter) machines to communicate phone messages. According to Ms. Vinopol, the institute two years ago developed software that can be loaded onto a PC or laptop and work with any modem — which was unusual, she says. This way, far less costly software replaces the need for a TTY.

The voice modem proved difficult for some people, she notes. “Jose worked to develop a sound card to put into a computer so you don’t have to have a modem at all. He also loaded it onto a hand-held [personal digital assistant] — without having to use the TTY device,” she says. A person using a PDA simply writes a message onto the device and, with the help of a special coupler on his phone that Mr. Hernandez-Rebollar invented, the message appears on a computer screen at the receiving end.



Click to Read More

Click to Hide