A Study on the Use of Kinect Sensor in Traditional Folk Dances Recognition via Posture Analysis

Abstract

In this paper, we evaluate the performance of widely applied soft computing classifiers, in folk dance recognition problems, emphasizing on posture identification. In particular, the goal is to identify postures which are characteristic for the dance performed, based on exploiting simultaneously the information of 24 body joints, acquired by a Kinect II sensor. The data sets described 6 folk dances, and their variations, originating from Greece.