Noise Robust Speaker Identification using PCA based Genetic Algorithm

Abstract

This paper emphasizes text dependent speaker identification system on Principal Component Analysis based Genetic Algorithm which deals with detecting a particular speaker from a known population under noisy environment. At first, the system prompts the user to get speech utterance. Noises are eliminated from the speech utterances by using wiener filtering technique. To extract the features from the speech, various types of feature extraction techniques such as RCC, LPCC, MFCC, MFCC and MFCC have been used. Principal Component Analysis has been used to reduce the dimensionality of the speech feature vector. To classify the speech utterances, Genetic Algorithm has been used. NOIZEOUS speech database has been used to measure the performance of this system under the condition of various SNRs. Experimental results show the superiority of the proposed close-set text dependent speaker identification system which can be used for security and access control purposes.

A text-independent speaker verification model: A comparative analysis

“…(7) Row Feature vector is the matrix with the eigenvectors in the columns transposed so that the eigenvectors are now in the rows, with the most significant eigenvector at the top, and Row Data Adjust is the mean-adjusted data transposed, i.e., the data items are in each column, with each row holding a separate dimension. [2]…”

“…The given voice signal is segmented into equal length voice segments and labels are assigned to identify the speaker. Md Raibul et al [2] have already worked on speaker identification which uses cepstral features and PCA for classification. An enhance study was done by Muda.L [3] [4,5,6] have also done an extensive analysis on feature extraction methods like MFCC, PLP, FFT, LPC and LPCC etc.…”

PCA/LDA approach for text-independent speaker recognition

“…The LDA-GMM classifier is similar to the GMM classifier in Eq. (11). The only difference is the input features X are replaced by feature Y with fewer dimensions reduced by LDA.…”

“…Jin et al 9 reported that the GMM classifier with LDA feature reduction can achieve higher performance with respect to accuracy and efficiency in some circumstances. Other researchers have integrated PCA with GMM 10 and genetic algorithms, 11 and have applied PCA and LDA in conjunction with K-Nearest Neighbors (KNN) algorithms 12 for speaker identification. This paper extends some of this previous work in new directions.…”