 "Photo: Kristyn Ulanday")
Boston University researchers are trying to point the way to the future of deaf studies and make it easier to learn and analyze American Sign Language data.
Linguistics professors and computer scientists have teamed up to use a $900,000 National Science Foundation grant over the next three years by collecting and analyzing raw data of people signing before video cameras that capture signs and entire sentences from multiple angles. The researchers are exploring the possibility of creating an online tool that will make it possible for people to see a thorough analysis of sign motions, ASL Linguistic Research Project Director Carol Neidle said.
"It's a basic research project," College of Arts and Sciences computer sciences Chairman Stanley Sclaroff, co-director of the Image and Computing Group at BU, said. "The online dictionary is definitely a long-term goal, but there are a number of computer science challenges that need to be solved in order for that to be a reality."
Neidle's group is researching ASL structure with SignStream, a program that allows users to see multiple views of signing at once onscreen and that could some day be used to research other sign languages or gesture components of spoken languages.
"[Sign Stream] has been very useful for our computer-based work on sign language recognition," Neidle said. "It has provided a ground truth that's enabling us to test and refine computer vision and algorithms. The goal is for the computer to detect what's linguistically significant in the visual images of signing."
A collection of ASL data that could be accessed by anyone online through a database Sclaroff nicknamed "Sloogle," would benefit not only the ASL community, Neidle said.
"We certainly think that it would be a good educational tool not only for things like ASL classes, but for parents of deaf kids," Neidle said. "The possibilities go on and on."
Ben Li, a School of Education junior specializing in deaf studies, said he thinks the research has potential.
"If it's successful, I think that the project would be a valuable asset for many people," Li said. "But I know that there are so many factors that need to be considered when transcribing languages. I'm looking forward to seeing their accomplishments."
Li said he thought the potential programs would find more use outside the classroom.
"It would be really useful for students at home as opposed to in class," Li said. "The strength of our programs at BU comes from the fact that our students are fully immersed within a signing environment the minute they step into the classroom."
Jaimee DiMarco, an SED sophomore who takes deaf studies classes, said she agreed an ASL search engine would undoubtedly enhance deaf studies.
"If the research leads to programs, then students of ASL could use it as a reference for signs they may need and don't know," DiMarco said.
Sclaroff said he and his fellow participants hope the research makes an important impression on the ASL user community, and computer scientists come away with the ability to conduct "broad analysis on human motion."
"Everyone wants his or her research to have an impact," Sclaroff said. "It'd be nice to use the techniques we have to enrich the lives of the ASL community."
Be the first to comment on this article!