Human voice recognition system with backpropagation neural network method

Prayugo, Mohammad Bagus Dimas, Nafisa, Nanda Azzahrotun, Yulianas, Azis and Fahmi, Hisyam ORCID: https://orcid.org/0000-0002-2665-1536 (2023) Human voice recognition system with backpropagation neural network method. Presented at The 12th International Conference on Green Technology (ICGT 2022), 26-27 Oct 2022, Malang, Indonesia.

Preview

Text 16012.pdf - Published Version Available under License Creative Commons Attribution Non-commercial. Download (566kB) | Preview

Abstract

The system on the computer can make everything run quickly and efficiently, so that it becomes a tool in information processing. One of the computer systems is an Artificial Neural Network (ANN). Along with technological advances, events that require computational models to perform speech recognition can be useful for science, as well as for making practical applications such as voice-based security systems. Artificial neural network is a method of grouping and separating data that has a working system like a neural network in humans. Artificial neural networks can pick up patterns that have been perfectly studied and well received. Backpropagation is a systematic method for training multiple layers of artificial neural networks. The backpropagation network model is composed of an input layer, at least one hidden layer and an output layer. Voice data in the form of signals is converted into discrete data by LPC and FFT methods. The activation function used is the sigmoid function, 2 hidden layers and the number of neurons 15. Optimal training was obtained in the 4th experiment with an MSE error of 0.19413 with a time of 11 s with 678 iterations. System accuracy to training data is 90%, and accuracy to test data is 40%. This means that the level of system accuracy can run well.

Item Type:	Conference (Paper)
Keywords:	artificial neural network; backpropagation; voice; Mean Square Error (MSE)
Subjects:	08 INFORMATION AND COMPUTING SCIENCES > 0801 Artificial Intelligence and Image Processing > 080108 Neural, Evolutionary and Fuzzy Computation 08 INFORMATION AND COMPUTING SCIENCES > 0801 Artificial Intelligence and Image Processing > 080109 Pattern Recognition and Data Mining
Divisions:	Faculty of Mathematics and Sciences > Department of Mathematics
Depositing User:	Hisyam Fahmi
Date Deposited:	03 Nov 2023 09:40

Downloads

Origin of downloads

Actions (login required)

View Item