Identifikasi Suara Pengontrol Lampu Menggunakan Mel-Frequency Cepstral Coefficients dan Hidden Markov Model

Abstract

Identification of voice signals can be used to command a computer system. Identification can be made of voice owner and spoken word. Identification of voice owner is used for security, while spoken word identification is often used to execute command on external devices. Previous research on voice identification using Mel-Frequency Cepstral Coefficient and Linear Vector Quantisation, other research using Linear Prediction Cepstral Coefficients and Linear Vector Quantisation. This study create a system that can identify the speakers on 15 seconds. Voice signal extracted using Mel Frequency Cepstral Coefficient, then Hidden Markov Model are used for classification. The classification results are tested to execute the command to turn on lamp.