Android-Based Detection Application Of Indonesian Sign Language System (SIBI) Using Rapid Application Development Method

Abstract

Deaf people have limited hearing which causes them to use sign language as a medium of communication. This research continues the development of the object model into an Android-based application to detect gestures of the Indonesian Sign Language System (SIBI) and translate them into spoken language. This research uses the Rapid Application Development (RAD) Method, the RAD Method development process is carried out iteratively through the stages of requirements planning, user design, construction and completion, with the Kotlin programming language and the TensorFlow Lite-based gesture detection model. Testing is done with respondents who are people with normal hearing or lay people and using the Blackbox Testing method. This application consists of three main features, real-time SIBI motion detection, information about SIBI, and translation from Indonesian to SIBI sign language. The test results from 30 respondents showed the usability test of the MySIBI application reached 85.6% which means this application is feasible to use and the results of Black Box Testing with an accuracy value of 100% which means the functionality of this system is very high. This application was successfully developed as an effective communication tool, which can bridge communication between deaf people and the wider community in Indonesia.