www.ijcrt.org © 2018 IJCRT | Volume 6, Issue 2 April 2018 | ISSN: 2320-2882 IJCRT1812538 International Journal of Creative Research Thoughts (IJCRT) www.ijcrt.org 861 Speaker Recognition with various Feature extraction and classification Techniques: A review Mrs. Suhasini S Goilkar Assistant Professor Department of Electronics and Telecommunication Engg. Finolex Academy of Management and Technology, Ratnagiri, Maharashtra, India 415639 Abstract: Speech is very natural form of human communication. Speech processing is one of the significant application area of digital signal processing. In speech processing the research developments like speech recognition, speaker recognition, speech synthesis, speaker identification, speech extraction and speech coding. In speaker recognition process who is speaking, is recognized automatically on the basis of individual information provided in speech wave. In speaker recognition technique the speaker’s speech is used to verify their identity a nd recognize using feature extraction techniques. The objective of this review paper is to summarize various feature extraction and classification techniques. Keywords— Analysis, segmentation, Feature extraction, speaker identification, matching, LPC, MFCC, RASTA filtering I INTRODUCTION From last fifty decades speech recognition is the research area. Many of the developments have been made in speech recognition. But still a one complete system is not developed which gives an accurate results. Speech signal contain different levels of information. Speaker recognition is used in many speech processing applications especially in security and authentication. Now a day’s sec urity is a major requirement [1]. Speaker and speech recognition are very closely related systems but these two systems are different. Speech recognition is the process of recognizing what is being said and speaker recognition is the process of recognizing who is speaking [2]. Basic speaker recognition system Input speech signal Pre-processing Feature extraction Classification Output Preliminary signal processing is carried out to improve the characteristics of the signal such as reducing inserted distortions and adjustment of the frequency range. In pre-processing the silent period of speech signal is removed. In feature extraction features are extracted using different techniques. In classification the different classifiers are used for classification. II SPEAKER RECOGNITION TECHNIQUES Different speaker recognition techniques are given in the table and discussed below. TABLE. 1 Analysis Feature extraction Modelling Matching Classification Segmentation LPC Speaker Recognition Whole-word matching DTW Super- segmentation LPCC Speaker identification Sub-word matching VQ Sub- segmentation MFCC Speaker independent HMM RASTA filtering GMM