Voice interaction using Gaussian Mixture Models for Augmented Reality applications

This paper addresses the human computer interaction techniques for Augmented Reality (AR) applications. In fact, AR aims at inserting 2D or 3D virtual object generated by the computer in a real video filmed by a camera. On the other hand, the interaction in AR allows the user to take an action and control the virtual objects. In this work, Automatic Speech Recognition (ASR) system based on Gaussian Mixture Models (GMM) is investigated for voice interaction in AR. Experimental results show that good performance of the developed system. Also, the voice interaction provides an intuitive and a natural workspace for interacting with the augmented environment.