Mel Frequency Cepstral Coefficients for Text-Independent Speaker Recognition Using Artificial Neural Network Model

다국어 초록 (Multilingual Abstract)

This paper demonstrates a security of speaker recognition system using ANN-GDA as training model. Speaker Recognition (SP) is the use of a machine to recognize a person from a spoken phrase. These systems can operate in two modes: to identify a particular person or to verify a person’ claimed identity. Speech processing and the basic components of automatic, In Text-Independent Speaker Recognition (TI-SP), there is no constraint on the text during enrollment and perform a recognition system does not know text spoken by person. Test utterances can be completely different from enrollment utterances. In this paper, we study the applicability of Artificial Neural Network (ANNs), the technique Gradient descent with adaptive lr back-propagation perform a training of the network can be envisaged as the minimization of an error function En as core classifiers for Mel Frequency Cepstral Coefficients (MFCC). The Artificial Neural Network show better performance for speech and need less training data [1] and experiment result shows that the ANN model achieved highest accuracy.

목차 (Table of Contents)

• Abstract
• 1. INTRODUCTION
• 2. PRE-PROCESSING
• 3. FEATURE EXTRACTION
• 4. PATTERN MATCHING
• 5. EXPERIMENT
• 6. CONCLUSIONS
• REFERENCES