Abstract:
The goal of this work is the use of the formant method for advance
tolerable accuracy in the safety of text dependent speaker
verification, that is succeeding speaker recognition from voice.
The parameters of safety consist of false acceptance rate and
false rejection rate.