An Assistive Body Sensor Network Glove for Speech- and Hearing-Impaired Disabilities

This paper presents a hand-gesture based interface for facilitating communication among speech- and hearing-impaired disabilities. In the system, a wireless sensor glove equipped with five flex sensors and a 3D accelerometer is used as the input device. By integrating the speech synthesizer onto an automatic gesture recognition system, user\´s hand gestures can be translated into sounds. In this study, we proposed a hierarchical gesture recognition framework based on the combined use of multivariate Gaussian distribution, bigram and a set of rules for model and feature set selection, deriving from a detailed analysis of misclassified gestures in the confusion matrix. To illustrate the practical use of the framework, a gesture recognition experiment has been conducted on American Sign Language (ASL) finger spelling gestures with two additional gestures representing space and full stop. The recognition model has been validated on the pangram "The quick brown fox jumps over the lazy dog.".