Design of a Machine Learning System for Sign Language Recognition
Team Member(s): Michael Parlato, William Mackie Semester: Spring 2013
An interactive software system which utilizes a user calibration to extract the hand from a video frame, compiles a mathematical model of the hand, and attempts to recognize the gesture in the image. The final product recognizes a set of 25 gestures with an accuracy of 99.52%. Development involved the use of artificial neural networks in recognizing patterns among the 25 classes.