A Gesture Recognition Design Toolkit Everyday Gesture Library

Abstract

Creating gestural recognition system is a challenging task which requires skills and updated applications. It is required for designer to be skillful and innovative in order to create interesting and acceptable gestures. Gesture Recognition Design Toolkit (GRDT) is set of tools designed to simplify the gesture creation process for non experts. From designing a new gesture, to suggesting the best fit, to test it for false positive activation, till final selection, are the step-by-step functions of Gesture Recognition Design Toolkit. Creating gestural recognition system, interface designers encounter many challenges throughout the process. Selecting the best gesture for a particular motion with low probability of false positives is one of many. This study explores ways to improve the functionality and enhance the efficiency of gestural library using machine learning techniques and data mining General Terms Multiple Action Gesture Interface Creation, Human computer interaction, Symbolic aggregate approximation, Nearest neighbor dynamic time wrapping, Hidden markov model, Dynamic time wrapping, Extract, transform and load