Speaker identification by anchor models with PCA/LDA post-processing | IEEE Conference Publication | IEEE Xplore