A robust, segmental method for text independent speaker identification | IEEE Conference Publication | IEEE Xplore