It was a problem for years, even prior to the recent war on terrorism. A deaf person went through airport security, unable to hear the buzz of the metal detector or the security guard’s commands, creating a tense situation and confusion. Given current world events, such a situation could now have disastrous consequences for the passenger involved. Many deaf people are not fluent in English, so posted signs aren’t always helpful. Having a sign language interpreter at each station is not cost-effective.

In the near future, Paula could provide a solution to this communication barrier, as well as many others.

“Paula” is a computer-generated synthetic interpreter developed by a team of faculty and students in DePaul University’s School of Computer Science, Telecommunications and Information Systems (CTI). Taking her name from the university where she was created, Paula uses speech recognition and sophisticated animation to translate spoken English into American Sign Language (ASL), the language of the deaf in North America. ASL is different from English or any other spoken language, featuring its own unique grammar rules. Human ASL translators are in great demand and are quite expensive to hire, if available when needed.

“Most people are not aware that ASL is not simply a signed form of English,” said Rosalee Wolfe, professor of computer science at DePaul and one of the leaders of the research team. “It is a series of hand configurations, hand positions, body positions and movement and facial expressions that are used in certain specific combinations. Hence, creating an animated translator is a very intricate and detailed process.”

The project represents four years and more than 25,000 hours worth of work by the ASL project team. The idea was born in March 1998, when Karen Alkoby, who is deaf and a DePaul CTI graduate student, outlined the concept of an automated ASL synthesizer in her thesis. Faculty members were intrigued by the idea and decided to see if it could be done, and began work on the project that fall. The ASL team includes about eight or nine faculty and about ten graduate students at any given time.

The faculty and students intentionally came from a diverse background of different talents, Wolfe said. “This is such a wide-reaching project that we need people who have different specialties,” she said. “We have people on the team who are experts in linguistics, speech recognition, human computer interaction, animation, graphics and more.”

The system works like this: a hearing person speaks through a headset, which is connected to the computer. The computer processes the command and the animated figure of Paula translates into ASL through hand gestures and facial expressions on the computer screen.

The system will make more information accessible to deaf and hearing-impaired people, and allow them to participate in exchanges with a hearing audience in classrooms, meetings, and other venues. There are hundreds of potential applications for this technology, including hotel desks, grocery stores and banks. However, the first application of this technology is the field of airport security. The ASL airport system utilizes a series of monitors featuring Paula’s translations, allowing a security guard to communicate effectively with deaf passengers as they proceed through the screening process. The team hopes to begin field testing the system in airports soon.

“There has been a long-standing need for a solution of this sort for deaf travelers in airports,” Wolfe said. “Given the current climate in air travel, it is more crucial than ever that airline passengers who are deaf have access to information about security procedures.”

The development of Paula had humble beginnings. A wooden model was initially used to imitate the movements of what would become the animated translator. Since ASL does not just use hand movements, but also finger positions, body movements, posture and facial expression, creating a model that would accurately translate from the system’s voice recognition software was a long and challenging process. “We had to account for a degree of precision that is not present in animation programs used for video games or movie special effects,” Wolfe said. “For example, we found that no one had ever properly animated the true movements of a human thumb, so we had to figure out how to make that work graphically.”

Despite the fact that Wolfe estimates that this system could sell for up to $8,000 per copy in the private sector, the team plans to donate the technology to non-profit organizations in the deaf community. Aside from a few small grants and university funding of hardware for the project, the system was developed by a tremendous volunteer effort on the part of current and past team members, Wolfe said. “The goal was to develop an ASL translator solution with wide access for people in the deaf community,” she added. “We want to see this technology in as many places and forms as necessary, and we don’t want cost to be an issue.”

For team members, the long process of building the system from the ground up has been a rewarding experience. “Even when we’ve hit a roadblock during our research, it’s still been a lot of fun to try and solve the various challenges that have come our way,” said John McDonald, assistant professor for DePaul CTI. “It’s truly been a labor of love for all of us, and the groundbreaking results are the proof of that.”

DePaul CTI is the one of the largest computer science programs in the country. The undergraduate program enrolls 1,949 students and offers six different degrees. More than 2,300 students are enrolled in the graduate program, which offers nine different master’s degrees. DePaul CTI also features a Ph.D. program in computer science. For more information on DePaul CTI, visit www.cs.depaul.edu. DePaul is the largest Catholic university in the nation and ninth-largest private university in the United States. A richly diverse population of 21,363 students attends classes on two city and six suburban campuses.

(NOTE TO EDITORS: For more background information on DePaul’s ASL Project or to see a demonstration, please visit http://asl.cs.depaul.edu.