Sistem Pengenalan Penutur dengan Metode Mel-frequency Wrapping

Authors

  • Ali Mustofa Jurusan Teknik Elektro, Universitas Brawijaya

:

https://doi.org/10.9744/jte.7.2.88-96

Keywords:

speaker, mel-frequency cepstral coefficients, vector quantization, K-mean

Abstract

Speaker recognition is the process of identifying a person based on his voice. Speaker recognition has several useful applications including biometric authentication and intuitive human computer interaction. the Mel Frequency Cepstral Coefficients (MFCC) technique is used to extract features from the speech signal and compare the unknown speaker with the exist speaker in the database. the filter bank is used to wrap the Mel frequency. VQ (vector Quantization) is a process of taking a large set of feature vectors and producing a smaller set of measure vectors that represents the centroids of the distribution. In this method, the K means algorithm is used to do the clustering. In the recognition stage, a distortion measure which based on the minimizing the Euclidean distance was used when matching an unknown speaker with the speaker database. Speech database used 10 different speakers with MFCC 12, 20 codebooks, and 16 centroids. Abstract in Bahasa Indonesia : Pengenalan penutur adalah proses identifikasi suara seseorang.. Pengenalan penutur berguna untuk otentikasi biometrik dan komunikasi antara komputer dengan manusia. Teknik Mel Frequency Cepstral Coefficients (MFCC) digunakan untuk ekstraksi ciri dari sinyal wicara dan membandingkan dengan penutur tak dikenal dengan penutur yang ada dalam database. Filter bank digunakan sebagai pembungkus (wrapping) mel frekuensi. Vector Quantization (VQ) adalah proses meletakkan vektor-vektor ciri yang besar dan menghasilkan ukuran vektor-vektor yang kecil yang berhubungan dengan distribusi centroid. Algoritma K-mean digunakan untuk kluster. Dalam tahap pengenalan, ukuran distorsi berdasarkan minimisasi jarak Euclidean digunakan untuk mencocokkan penutur tak dikenal dengan penutur dalam database. Database wicara menggunakan 10 penutur berbeda dengan MFCC 12, 20 codebook, dan 16 centroid. Kata kunci: penutur, mel-frequency cepstral coefficients, vector quantization, K-mean

Downloads

Published

2008-01-25