期刊名称：Romanian Conference on Human-Computer Interaction
印刷版ISSN：2344-1690
出版年度：2018
卷号：RoCHI 2018
页码：84-90
语种：English
出版社：Matrix ROM
摘要：This paper describes a system for gesture recognition based on a smart watch and a mobile phone. Our system also integrates a Web API for saving, retrieving and processing data about the gestures and a web application to visualize the saved data. We propose a button-based segmentation method that uses Dynamic Time Warping and One Nearest Neighbors to classify the data from a single 3-axis accelerometer. The system distinguishes 4 arm gestures using a variable number of labeled templates for the classifier. To emphasize the practical utility of the system the recognized gestures are used to control a music player. This is just an example. This form of user interaction can be exploited by many different types of applications, for instance applications for home automation.
