Exploiting PCA classifiers to speaker recognition

A novel approach to text-independent speaker recognition using a new classifier, called principal component space (PCS) is proposed in this work. This classifier uses the subspaces spanned by the principal components as the criteria. Together with other PCA classifier, it forms a hybrid classifier which is another technique presented here. All of these classifiers were applied to speaker recognition in particular on YOHO corpus. The experimental works show promising results.