In this paper, we describe a method for representing the acoustic similarity of a target
speaker with respect to a set of known speakers as a feature for speaker verification. We …