Speaker Recognition is the computing task of recognizing a speaker using some speaker- dependent characteristic of his/her voice, known as feature. The recognition process consists of three main stages: feature extraction, speaker modeling and speaker pruning or decision making. In this paper, numerical features of each speaker have been reduced using k-means clustering. The result of k-means clustering is improved by applying Gaussian Mixture Model to reduce the time complexity. The feature of the unknown speaker is compared with that of the speakers stored in the codebook. The speaker is identified from the codebook based on the maximum probability using likelihood function. It is observed that the accuracy of speaker recognition can be improved using this approach.

Speaker Recognition is the computing task of recognizing a speaker using the speaker-dependent part of his voice known as feature. It can be of two types, text-dependent and text-independent. In text-dependent speaker recognition system, the speakers utter a predefined phrase known to the system, whereas text-independent speaker recognition system can process any phrase that makes it robust.

However, nowadays more and more attention is being paid to speaker recognition field. Speaker recognition involves two applications: speaker identification and speaker verification. Speaker identification is the process of automatically recognizing the person speaking on the basis of individual information included in the speech waves. This technique makes it possible to use the speaker’s voice to verify his/her identity and control access to services such as voice dialing, banking by telephone, telephone shopping, database access services, information services, voice mail, security control for confidential information areas, and remote access to computers.

Computer Sciences Journal, Mel Frequency Cepstral Coefficient (MFCC), Gaussian Mixture Model (GMM), Expectation-Maximization (EM) algorithm, k-means clustering, Speaker pruning and Codebook