IEEE 14th Signal Processing and Communications Applications, Antalya, Türkiye, 16 - 19 Nisan 2006, ss.57-58
A new decision making algorithm in speaker verification is presented. First, a baseline system is formed using mel-frequency cepstral coefficients (MFCC) as features, and for speaker modeling a radial basis function (RBF) neural network is used. Then, a two-stage decision making algorithm is proposed, which has the aim to eliminate the qualified impostors in the second stage.