Sistem Pengenalan Ucapan Bahasa Daerah Menggunakan Metode Mel Frequency Cepstral Coefficient (MFCC) dan Adaptive Neuro-Fuzzy Inference System (ANFIS)

  • Ade Hastuty Universitas Muhammadiyah Parepare
  • Basri Muh Universitas Muhammadiyah Parepare
  • Asfiad Amir Universitas Muhammadiyah Parepare
Keywords: Bugis Language, dialects, classification, recognition, MFCC, ANFIS

Abstract

Bugis language is one of the traditional languages in South Sulawesi which has various dialects. Different ways of speaking and tone of voice that cause voice of each person also vary. The design of the Bugis language speech recognition system is one way to digitize and indicate the Bugis regional language. This study uses Mel Frequency Cepstral Coefficient (MFCC) method as a feature extraction and Adaptive Neuro-Fuzzy Inference System (ANFIS) as an algorithm for classification and recognition. This research was conducted by collecting various voice data from various age groups and regions. Voice data is collected in wav audio format and then extracted into numerical results to use as database on system. Test was carried out using various database models and the amount of voice data. The test results show that databases with single data produce a higher level of accuracy with a percentage of truth above 50%. While the test results with a database of 37 speakers with a total data of 740 have a low level of accuracy and are below 50%. Testing in outdoors has a lower accuracy around 10%. Error value in the data training process in the range of 0.5 - 1.5.

References

[1] W.S.M. Sanjaya dan Z. Salleh. 2014. “Implementasi Pengenalan Pola Suara Menggunakan Mel-Frequency Cepstrum Coefficients (MFCC) Dan Adaptive Neuro-Fuzzy Inferense System (ANFIS) Sebagai Kontrol Lampu Otomatis”. Al-HAZEN Jurnal of Physics, vol 1, no.1, hlm. 43-54, Desember 2014
[2] T. Nasution. “Metoda Mel Frequency Cepstrum Coefficients (MFCC) untuk Mengenali Ucapan pada Bahasa Indonesia”. Jurnal Sains dan Teknologi Informasi, vol. 1, no. 1, hlm. 22-31, Juni 2012
[3] N.R. Pradipta, Tasripan, H. dan Kusuma. “Perancangan Perangkat Antarmuka Berbasis Pengenalan Suara pada Purwarupa Mesin Cetak Huruf Braille ITS untuk Operator Tunanetra”. Jurnal Teknik ITS, vol. 7, no. 2, hlm. A401-A407, 2019.
[4] M.B. Gunawan. “Konversi Suara ke Teks Menggunakan Metode Hidden Markov Model”. Skripsi, Universitas Sanata Dharma, Yogyakarta, Agustus 2010.
[5] David “Penerapan Algoritma Levenberg-Marquadt dan Backpropagation Neural Network Untuk Klasifikasi Suara Manusia”. Jurnal Buana Informatika, vol. 4, no. 1, Januari 2013
[6] M.N. Rabbani, A. Rizal, dan F.Y. Suratman. “Implementasi Kunci Berbasis Suara Menggunakan Metode Mel Frequency Cepstral Coefficient (MFCC)”. e-Proceeding of Engineering, vol. 3, no. 3, hlm. 3998-4007, Desember 2016.
[7] R. Umar, I. Riadi dan A. Hanif. “Analisis Bentuk Pola Suara Menggunakan Ekstraksi Ciri Mel-Frequency Cepstral Coefficients (MFCC)”. Cogito Smart Journal, vol. 4, no.2, hlm. 294-304, 2019.
[8] U. Sutisna, R. Hidayat, dan Litasari. “Pengenalan Tutur Terisolasi Dalambahasa Indonesia Menggunakan MFCC, FCM, Dan ANFIS”. Seminar Nasional Teknologi Informasi dan Multimedia 2013, hlm. 22-7 – 22-12.
.
Published
2021-05-06
Section
Articles