Ada-Boostbased Gesture Recognition using Time Interval Window

Abstract

----The task of recognizing 3D gesture for controlling equipment is highly challenging due to the propagation of 3D smart TV recently. In this paper, the Ada-Boost algorithm is applied to 3D gesture recognition by using a Kinect sensor. We recognized time-invariant 3D gesture using global and local feature vectors that are normalized. The multi Ada-Boost algorithm is used to train and classify many 3D gesture types. Our experiment shows 95.17% of accuracy and 3.73% of error rate.