Pengenalan Bahasa Isyarat Indonesia Dinamis Menggunakan Long Short-Term Memory

Abstract

Sign language is a way for people with special needs, or people with disabilities, to communicate. Sign Language Recognition (SLR) is the field of research and technology focused on developing computer systems that can recognize, interpret, and translate sign language. There have been many advances in recognizing static characters in Indonesian Sign Language. However, this research has only focused on static sign language and hasn't addressed dynamic sign language involving hand movements or gestures. This study develops a method for recognizing characters in Indonesian Sign Language, considering not only static characters but also dynamic ones involving hand movements or gestures using LSTM. The average accuracy of real-time testing is about 92%, with an average testing time of 4 seconds.