The Development of Android for Indonesian Sign Language Using Tensorflow Lite and CNN: An Initial Study

Abstract

Deaf people in the world have different sign language according to their living place. They tend to use simple sign language to communicate with others. Whereas every country has their own sign language. Usually a country has one national sign language, but there are 2 kinds of sign languages (SIBI and BISINDO) in Indonesia. However, BISINDO (Indonesian Sign Language) had not recognized yet as a formal sign language like as SIBI (Indonesian Sign Language System), it is only a daily sign language among them and with normal people. Thus, researcher propose to promote BISINDO by developing BisAndro (BISINDO using Android) application. At the start, researcher prepare to compare BISINDO systems in cyberspace, develope basic BISINDO website by PHP and MySQL, train some gestures in TensorFlow lite, and make CNN (Convolutional Neural Network) plan before developing BisAndro application. BisAndro application will be built by Python programming, Tensorflow as an image recognition technology, and Android mobile phone as the launcher device. The application is expected can be used by deaf people to communicate with others easily because it facilitates two ways communication through mobile phone. So, the preparation is needed to know about initial requirement of the next BISINDO learning application using Android smartphone in order to be more useful for deaf and normal people directly.