Wiimote squash : comparing DTW and WFM techniques for 3D gesture recognition

This paper compares Dynamic Time Warping (DTW) and Waveform Matching (WFM), the two gesture recognition techniques, applied on a specific Squash game application we have developed. Our application gets accelerometer readings by moving the Nintendo™ Wiimote in a 3D space. This application manipulates the Wiimote gestures in the form of signals. These signals are the effect of the movements, the positioning and the orientations of Squash racket (actually Wiimote) on the computer screen. We implemented five Squash shots (i.e. 3D uni-stroke gestures) in order to compare DTW and WFM. The results indicate that there is not a significant difference in the detection of gestures by both techniques (DTW and WFM) but there exists a strong degree of association between both techniques.