A fast speaker verification method with global speaker model
Yiying Zhang, Xiaoyan Zhu · 2002
A fast text-independent speaker verification method is proposed based on likelihood score normalization. A global speaker model is established to represent the universal features of speech. The likelihood score is subtracted by the score produced by this model so that the common factors of speech are removed and the differences between speakers are brought into prominence. Compared to the conventional method and the popular likelihood score normalization approach, the proposed technique decreases the equal error rate significantly, and the verification speed is fast.