The SVM, SimpSVM and RVM on sign language recognition problem

Human gesture recognition is a rather new field and many challenges, sign language recognition is a concrete example of gesture recognition. In this paper, we study the feasibility and effectiveness of vector machine learning methods, namely Support Vector Machine (SVM), Simplification of Support Vector Machine (SimpSVM) and Relevance Vector Machine (RVM) to the sign language recognition problem. At the same time, we also give some comparative results of SVM, SimpSVM and RVM for that problem. The experimental results on the Auslan data set show SVM, SimpSVM and RVM could achieve the state-of-the-art predictive performance, also pointed out that prediction behaviors of them are similar in terms of the prediction accuracy when the number of feature changed and sign discrimination. However, SimpSVM and RVM require fewer testing time than SVM in testing phase.