Real-Time Recognition of Body Motion for Virtual Dance Collaboration System

Abstract

A method of real-time recognition of body motion for a virtual dance collaboration system is described. Fourteen feature values are extracted from motion-captured body motion data, and the dimensions of the data are reduced by using principal component analysis (PCA). In the training phase, a dictionary for motion recognition is constructed from training samples of several types of motion. In the recognition phase, feature values obtained from a live dancer's motion data are projected to the subspace obtained by PCA, and the system recognizes the live dancer's motions by comparing them with the motion dictionary. In this paper, we present recognition experiments on seven kinds of basic motions and breakdance motions. The recognition experiment proved that the method was effective for motion recognition. A preliminary experiment in which a live dancer and a virtual dancer collaborate with body motion was also carried out.