Open AccessPerformance analysis method of dynamic time warping and k-nearest neighbor in sound based presence system
Author(s) -
Herbet Simangungsong,
Herman Mawengkang,
Maya Silvi Lidya
Publication year - 2020
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/725/1/012127
Subject(s) - dynamic time warping , computer science , biometrics , speech recognition , fast fourier transform , k nearest neighbors algorithm , pattern recognition (psychology) , artificial intelligence , matching (statistics) , test data , mathematics , algorithm , statistics , programming language
Verification and identification of a person using biometrics have been widely used as the retina of the eye, the face and voice. In this experiment, such as voice biometrics to identify a person who is used to the system presence. Voice recognition is done by pattern matching between training data and test data. In this study used methods Dynamic Time Warping (DTW), K-Nearest neigbors (KNN) and Fast Frequency Transform (FFT) for voice recognition. DTW is used as a method of pattern recognition, while KNN is used for sound classification. Before testing conducted prior extraction using FFT method. This study uses 100 votes out of 10 people with the amount of each 10 people. Presentations were used as training data by 70% and 30% of test data. Results obtained by dividing the recognized voice to the overall sound. From the results 83.33 % voice recognition.