$900,000 NSF Grant For Technology To Study American Sign Language

Oct. 29, 2007

Oct 29, Assistant Professor Vassilis Athitsos and his collaborators have been awarded a three-year, $900,000 National Science Foundation (NSF) grant to develop new technology for learning American Sign Language (ASL). The research will focus on designing novel computer vision and machine learning methods for automatic sign recognition. The project is titled "Large Lexicon Gesture Representation, Recognition, and Retrieval." It is led by Professors Stan Sclaroff and Carol Neidle at Boston University, and Professor Vassilis Athitsos at UT Arlington. In current ASL dictionaries, signs are typically organized according to their English translation, and a user cannot look up a sign without knowing in advance the meaning of that sign. A key goal of the project is to develop methods that would allow users to demonstrate signs in front of a camera and have a computer look up the meaning of those signs based on their visual properties. The BU and UTA researchers also hope to develop a way to perform "keyword-based" searches for American Sign Language, by identifying occurrences of specific signs in large video databases of ASL literature.