Open AccessAPPLICATION OF GAUSSIAN SUPERVECTOR IN SPEECH ANALYSIS
Author(s) -
Kauleshwar Prasad,
Piyush Lotia
Publication year - 2015
Publication title -
international journal of electrical and electronics engineering
Language(s) - English
Resource type - Journals
ISSN - 2231-5284
DOI - 10.47893/ijeee.2015.1158
Subject(s) - mel frequency cepstrum , computer science , speech recognition , mixture model , matlab , speaker recognition , pattern recognition (psychology) , identity (music) , feature (linguistics) , expectation–maximization algorithm , identification (biology) , feature extraction , maximum likelihood , gaussian , artificial intelligence , cepstrum , mathematics , statistics , linguistics , philosophy , physics , botany , quantum mechanics , acoustics , biology , operating system
The idea of the Speaker Identification is to implement a recognizer using Matlab which can identify a person by processing his/her voice. The basic goal of the paper is to classify and recognize the speeches of different persons. This classification is mainly based on extracting several key features like Mel Frequency Cepstral Coefficients (MFCC) from the speech signals of those persons by using the process of feature extraction using MATLAB. The above features may consists of pitch, amplitude, frequency etc. Using a statistical model like Gaussian mixture model (GMM) and features extracted from those speech signals we build a unique identity for each person who enrolled for speaker recognition. There is an elegant and powerful method for finding the maximum likelihood and that method is called Expectation and Maximization algorithm. The performance of the technique has been measured by three parameters: Number of Speakers in Database, Number of Persons Tested and the % Error.