VGG-16 ARCHITECTURE ON CNN FOR AMERICAN SIGN LANGUAGE CLASSIFICATION

Mutiara Dolla Meitantya (Faculty of Computer Science, Universitas Dian Nuswantoro, Semarang, Indonesia)
Christy Atika Sari (Faculty of Computer Science, Northern Technical University, Mosul, Iraq)
Eko Hari Rachmawanto (Faculty of Computer Science, Universitas Dian Nuswantoro, Semarang, Indonesia)
Rabei Raad Ali (Faculty of Computer Science, Northern Technical University, Mosul, Iraq)

Jurnal Teknik Informatika (JUTIF)
Vol. 5 No. 4 (2024): JUTIF Volume 5, Number 4, August 2024

Every country has its sign language such as in Indonesia there are 2 types namely Indonesian Sign Language System called SIBI and BISINDO (Indonesian Sign Language). American Sign Language (ASL) is a sign language that is widely used in the world. In this research, the classification of American Sign Language (ASL) using the Convolutional Neural Network (CNN) method using VGG-16 architecture with Adam optimizer. The data used is 14000 ASL image data with 28 classes consisting of letters A to Z plus space and nothing with a division of 90% training data and 10% validation data. From this research, the overall accuracy is obtained with a value of 98% and the accuracy value of validation data evaluation is 89.07%.