Open AccessMad Reading Law Classification Using Mel Frequency Cepstal Coefficient (MFCC) and Hidden Markov Model (HMM)
Author(s) -
Oddy Virgantara Putra,
Faisal Reza Pradana,
Jordan Istiqlal Qalbi Adiba
Publication year - 2021
Publication title -
procedia of engineering and life science
Language(s) - English
Resource type - Journals
ISSN - 2807-2243
DOI - 10.21070/pels.v2i0.1148
Subject(s) - hidden markov model , mel frequency cepstrum , speech recognition , computer science , sentence , reading (process) , feature extraction , artificial intelligence , pattern recognition (psychology) , feature (linguistics) , linguistics , philosophy
The COVID-19 pandemic is a disaster that hit the world at this time, all activities are limited. This pandemic has also greatly impacted the process of teaching and evaluating the reading of the Koran which was carried out using the talaqqi and musyafahah methods. Machine Learning research has been developed for the legal classification of Quran recitation. This study aims to be able to classify the law of recitation of recitation, especially in the law of Mad recitation of the letter Maryam verses 1 to 15. This study builds a model using the Mel Frequency Ceptral Coefficient (MFCC) feature extraction with the Hidden Markov Model (HMM) algorithm method. MFCC is used for feature extraction in voice that processes voice data in several stages, including pre-emphasis, frame-blocking, windowing, Fast Fourier Transform, Mel Frequency Wrapping, and Ceptral Liftering. HMM is used in speech recognition with standard sentence percentages. The dataset used in this study is voice data taken from the voice of the Quran reciter that has been recognized and has been affiliated. The test results on the model that has been built have an average percentage of 80% accuracy of the test data.