Speaker recognition is conceded with ex-
tracting the identity of the person speaking the ut-
terance. In today’s modern world, as speech interac-
tion with smart computers becomes more persistent
in day to day activities, the utility of automatically
recognizing a speaker increases. Speaker recog-
nition can be text dependent in which recognizer
works on pre-trained data, or it may be text-inde-
pendent where there is no constraint on the spoken
words by the speaker.
The techniques for speaker recognition can
be categorized into three major approaches.
38