This paper presents a standalone system that converts static gestures into voice. It is also known as GesTALK. The primary input device is the self design and cost effective Data Glove. It operates in two basic modes. In first mode the system speaks a single alphabet for a static gestures made by the user, while in second mode the system speaks complete string by concatenating words. Various gestures of alphabets are made to spell out the words involved in string. This system has been shown to work for both American Sign Language (ASL) and Pakistan Sign Language (PSL) and other language models. Results show that the system recognizes 24 out of 26 letters; an overall accuracy of 90% is achieved. © 2008 IEEE.
Mendeley saves you time finding and organizing research
Choose a citation style from the tabs below