APLIKASI PENERJEMAH BAHASA ISYARAT BISINDO MENGGUNAKAN METODE YOLOv5 BERBASIS MOBILE

Abstract

This research aims to develop a mobile application that is able to recognize and translate BISINDO in real-time using the YOLOv5 method. YOLOv5 (You Only Look Once version 5) is an efficient and fast object detection algorithm, making it suitable for implementation on mobile devices. In developing this application, we collected and labeled a typical BISINDO dataset, then trained the YOLOv5 model to detect and recognize these characteristics. The test results show that the implemented model has high accuracy in recognizing various BISINDO characteristics. Apart from that, this application is also designed with an intuitive and easy-to-use user interface, so that users can quickly and easily communicate with people using sign language. With this application, it is hoped that it can expand communication access for people with hearing impaired and increase their social inclusion in society. Further research is needed to continue improving the app's accuracy and efficiency and adding more features that support different user needs.