Capstrum Coefficient Features Analysis for Multilingual Speaker Identification System

Abstract

The Capstrum coefficient features analysis plays a crucial role in the overall performance of the multilingual speaker identification system. The objective of the research work to investigates the results that can be obtained when you combine Mel-Frequency Cepstral Coefficients (MFCC) and Gammatone Frequency Cepstral Coefficients (GFCC) as feature components for the front-end processing of a multilingual speaker identification system. The MFCC and GFCC feature components combined are suggested to improve the reliability of a multilingual speaker identification system. The GFCC features in recent studies have shown very good robustness against noise and acoustic change. The main idea is to integrate MFCC & GFCC features to improve the overall multilingual speaker identification system performance. The experiment carried out on recently collected multilingual speaker speech database to analysis of GFCC and MFCC. The speech database consists of speech data recorded from 100 speakers including male and female. The speech samples are collected in three different languages Hindi, Marathi and Rajasthani. The extracted features of the speech signals of multiple languages are observed. The results provide an empirical comparison of the MFCC-GFCC combined features and the individual counterparts. The average language-independent multilingual speaker identification rate 84.66% (using MFCC), 93.22% (using GFCC)and 94.77% (using combined features)has been achieved.