Abstract: In this paper, we explore joint factor analysis (JFA) for text-dependent speaker recognition with random digit strings. The core of the proposed method is a JFA model by which we extract ...