SBIR Phase I: A Sign Parameter-to-Text Input Software for People Who Are Deaf

This Small Business Innovation Research (SBIR) Phase I project proposes to develop assistive technology software that provides an accessible input interface for people who are Deaf. Many Deaf individuals have difficulty with reading and writing English text. Although text is within the visual modality, it is not always accessible from a language perspective. The objective of this project will be to develop Sign Parameter-to-Text technology, which will allow Deaf individuals, who are fluent in American Sign Language, the ability to search for signs based on linguistic parameters such as handshake, location, and movement. Once the user finds the sign, semantically equivalent English words will be displayed for selection as the text input. This research will be done by preparing a corpus of signs using a phonological notation system and performing statistical analysis on the data to discover the discriminatory features which can optimize the search capabilities. Based on these features, this project will develop search algorithms, a proposed graphical user interface and exploration of hardware input devices which correspond intuitively with the interface. The proposed interface features and functionality will be discussed with a focus group for design feedback. Design requirements and feasibility studies will be documented. The commercialization of this technology will result in a product that provides an accessible input interface to off-the-shelf-software such as word processing, email and instant messaging programs and will complement other writing supports such as automated spelling and grammar checking and word prediction as well as English writing tutorials. Additionally, this technology has educational value and may increase the English literacy and writing skills of Deaf individuals. This research will add a distinct input component to VCom3D's current Text-to-Sign Animation output system. This research will also contribute to corpus analysis and search engine algorithms based on sign language phonological parameters.