Speech interaction : comprehensive design and development of time in improving performance of speaker recognition from front end to back end

Speaker recognition is a process where a person is recognized on the basis of his/her voice signals [1]. Speaker recognition can be further broken into two categories: speaker identification and speaker verification. Identification takes the speech signal from an unknown speaker and compares this with a set of valid users. The best match is then used to identify the unknown speaker. Similarly, in verification the unknown speaker first claims identity, and the claimed model is then used for identification. If the match is above a predefined threshold, the identity is accepted. A complete speaker recognition system consists of front end and back end process. The front end which contain preprocessing and feature extraction while the back end which contain pattern classification techniques.