Human Action Recognition Based on Fuzzy Support Vector Machines

Abstract

As human action is uncertain and illegible, a human action recognition method basing on fuzzy support vector machine is presented. Fuzzy support vector machine employs the membership function to solve the unclassifiable areas which happens the traditional SVMs' two-class problems extend to the multi-class problems. the method is evaluated on the Weizmann action dataset and received comparative high correct recognition rate. the experimental results show that our approach has efficient recognition performance.