Hidden Markov Model for Gesture Recognition

Abstract

This report presents a method for developing a gesture-based system using a multi-dimensional hidden Markov (HMM). Instead of using geometric features, gestures are converted into sequential symbols. HMMs are employed to represent the gestures and their parameters are learned from the training data. Based on "the most likely performance" criterion, the gestures can be recognized through evaluating the trained HMMs. We have developed a prototype system to demonstrate the feasibility of the proposed method. The system achieved 99.78% accuracy for an isolated recognition task with nine gestures. Encouraging results were also obtained from experiments of continuous gesture recognition. The proposed method is applicable to any gesture represented by a multi-dimensional signal, and will be a valuable tool in telerobotics and human computer interfaces.