The EPSRC funded GestAID project (Gesture-driven Adaptive Interfaces for the Disabled) ran from March 2000 until March 2003. It investigated the use of computer vision systems to provide a means of computer access for users who have difficulty with traditional mouse-keyboard interfaces. Gesture recognition systems were developed which learned to recognise user-specific gestures from examples. They were relatively fast to train and did not require markers or any other equipment to be worn by the users. Computer users with motor impairment defined their own gesture vocabularies and trained the computer to recognise them. Such vocabularies are easier to remember and more comfortable to perform than predefined signs and gestures. Several methods for learning and recognition using trajectory-based and history-based visual representation were implemented. These were evaluated with a range of users and over 1000 video sequences in collaboration with volunteers from the MacKinnon Centre, Dundee.

The project was based in the Department of Applied Computing at the University of Dundee.

The principal investigator was Dr. Stephen McKenna.
Research Assistant: Kenny Morrison

Publications

Morrison K and McKenna S J, "An Experimental Comparison of Trajectory-based and History-based Representation for Gesture Recognition", 5th International Workshop on Gesture and Sign Language based Human-Computer Interaction, Genoa, Italy, April, 2003, 12 pages.

Morrison K and McKenna S J, "The Dundee Gesture Database for User-Specified, Motor-Impaired Computer Input", 5th International Workshop on Gesture and Sign Language based Human-Computer Interaction, Genoa, Italy, April, 2003

Morrison, K and McKenna S J, "Automatic Visual Recognition of Gestures made by Motor-Impaired Computer Users", Journal of Technology and Disability, IOS Press (ISSN 1055-4181), Vol. 14, No. 4, 2002, 197-203

Morrison K and McKenna S J, "Contact-Free Recognition of User-Defined Gestures as a Means of Computer Access for the Physically Disabled", Proceedings of the 1st Cambridge Workshop on Universal Access and Assistive Technology, University of Cambridge, 25-27 March, 99-103

Morrison K and McKenna S J, "Vision-based Gesture Recognition as an Assistive Technology for Computer Access", Recent Advances in Assistive Technology and Engineering, Birmingham, 5-6 November 2001

In Review

McKenna S J and Morrison K, “Vision-based Recognition of User-Specific Gestures for Computer Access”, IEEE Transactions on Neural and Rehabilitation Engineering