Authors : R. Visalakshi and P. Dhanalakshmi
Abstract: In this study, researchers have analyzed the performance of a Speaker recognition system based on features extracted from the speech recorded using close speaking and a throat microphone in clean and noisy environment. In general, clean speech performs better for Speaker recognition system. Speaker recognition in noisy environment, using a transducer held at the throat results in a signal that is clean even in noisy. The proposed techniques of Speaker Recognition (SR) are done in two ways such as acoustic feature extraction and classification. Initially, the speech signal is given to various acoustic features that include Linear Predictive Coefficients (LPC), Linear Predictive Cepstral Coefficients (LPCC) and Relative Spectral-Perceptual Linear Predictive (RASTA-PLP). Second the extracted features are given to Auto Associative Neural Networks (AANN). The experimental results show that the performance of RASTA-PLP with AANN Model gives an accuracy of 95% in clean and 89% in noisy environments using throat microphone.
R. Visalakshi and P. Dhanalakshmi, 2014. Acoustic Feature Extraction Methods LPC, LPCC and RASTA-PLP in Speaker Recognition. Asian Journal of Information Technology, 13: 595-598.