A Hand Data Model for Gesture Recognition
This paper describes a novel hand data model as a system evaluator that defines the minimum requirements for a robust recognition system. Interface devices for the user’s hand can be as simple as a switch to highly complex multi sensor gloves. A model is required that can cope with both diversity of the input devices and the rich variety of gesture based sign languages and user interfaces. We adopt an estimation of the Bayes error by calculating the classification error using distribution characteristics of the patterns, and this estimation is useful in classifying the gestures, evaluating the gesture input system, and specifying the gesture vocabulary.
Unable to display preview. Download preview PDF.
- A. Ashbrook. Geometric Histograms for Object Recognition: Developments and Analysis. PhD thesis, University of Sheffield (to be submitted ), 1996.Google Scholar
- S. S. Fels. Building Adaptive Interfaces with Neural Networks: The Glove-Talk Pilot Study. PhD thesis, University of Toronto, 1990.Google Scholar
- M. Skelly. Amer-Ind gestural code based on universal American Indian hand talk. Elsevier, 1979.Google Scholar