<![CDATA[Indonesian Sign Language (BISINDO) and the Indonesian Sign Language System (SIBI) are forms of communication used by the deaf community in Indonesia. However, the use of BISINDO is considered less effective in the sign language translation system due to variations in body movements in each community. On the other hand, SIBI is considered more effective because it is an adaptation of American Sign Language (ASL) and has been officially recognized by the Indonesian government. This research aims to develop a deep learning-based sign language translation system to support communication with the deaf community using Indonesian Sign Language (SIBI). The research methodology involves acquiring a sign language data set, preprocessing the data using the Mediapipe library, training the model using Gated Recurrent Neural Networks (GRU), and evaluating model performance using the Confusion Matrix method. The test results show that the developed model succeeded in achieving an accuracy level of 94% in classifying SIBI sign language signs. This shows the potential of the system in assisting communication and increasing accessibility for deaf people who use Indonesian Sign Language. This research makes a significant contribution to technological developments aimed at improving the quality of life and social inclusion for the deaf community]]>
Copyrights © 2024
