Penerjemah Bahasa Isyarat Indonesia (BISINDO) Ke Media Suara Menggunakan Long Short-Term Memory (LSTM) Berbasis Intel Next Unit Computing (NUC)

Erlangga, I Putu Krisna (2024) Penerjemah Bahasa Isyarat Indonesia (BISINDO) Ke Media Suara Menggunakan Long Short-Term Memory (LSTM) Berbasis Intel Next Unit Computing (NUC). Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5024201055-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2026. Download (33MB) | Request a copy

Abstract

Bahasa isyarat merupakan bahasa yang direpresentasikan dalam gerakan tangan dan ekspresi wajah. Tunarungu menggunakan bahasa isyarat sebagai bahasa komunikasi utama. Dalam berkomunikasi sehari – hari, tunarungu lebih memilih menggunakan BISINDO karena tidak terikat dengan struktur baku bahasa Indonesia dan disertai ekspresi wajah. Menurut GERKATIN terdapat setidaknya 2,9 juta orang penyandang tunarungu. Jumlah penyandang tunarungu yang cukup besar ini tidak diikuti dengan pengetahuan masyarakat umum mengenai bahasa isyarat. Hal ini berdampak pada sulitnya komunikasi tunarungu dengan masyarakat sekitar sehingga adanya keterbatasan dalam peningkatan kualitas hidup mereka. Sistem penerjemah saat ini masih terbatas dalam menerjemahkan dalam bentuk kata saja dan belum adanya upaya dalam membuat sistem yang bersifat inklusif. Pada tugas akhir ini telah dikembangkan sistem penerjemah BISINDO menggunakan arsitektur LSTM. Sistem telah diimplementasikan pada Intel NUC dengan kemampuan dalam menerjemahkan gerakan isyarat secara real time. Pengguna dapat membentuk kalimat - kalimat yang umum digunakan sehari - hari dan mengkonversinya ke media suara dengan bantuan gerakan isyarat kontrol. Berdasarkan pengujian yang telah dilakukan, didapat bahwa sistem dapat beradaptasi dengan adanya perbedaan intensitas cahaya, jarak, serta subjek yang berbeda dengan penulis dengan akurasi tertinggi mencapai 100%. Sistem juga telah dapat berjalan secara realtime dengan performa baik pada Intel Next Unit Computing (NUC). Sistem ini dapat menjadi solusi dalam mengatasi hambatan komunikasi antara tunarungu dengan khalayak umum
======================================================================================================================================
Sign language is a language that is represented in hand movements and facial expressions. Deaf people use sign language as their main language of communication. In everyday communication, deaf people prefer to use BISINDO because it is not tied to the standard structure of Indonesian and is accompanied by facial expressions. According to GERKATIN there are at least 2.9 million deaf people. This large number of deaf people is not accompanied by the general public's knowledge of sign language. This has an impact on the difficulty of communication between deaf people and the surrounding community, resulting in limitations in improving their quality of life. The current translation system is still limited to translating words only and there has been no effort to create an inclusive system. In this final project, a BISINDO translation system has been developed using the LSTM architecture. The system has been implemented on an Intel NUC with the ability to translate gestures in real time. Users can form sentences that are commonly used daily and convert them to voice media with the help of control gestures. Based on the tests that have been carried out, it was found that the system can adapt to differences in light intensity, distance, and different subjects with the author with the highest accuracy reaching 100%. The system can also run in real time with good performance on Intel Next Unit Computing (NUC). This system can be a solution in overcoming communication barriers between the deaf and the general public

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Tunarungu, BISINDO, LSTM, Intel NUC, Deaf, BISINDO, LSTM, Intel NUC
Subjects:	Q Science > QA Mathematics > QA336 Artificial Intelligence Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science) R Medicine > R Medicine (General) > R858 Deep Learning T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Computer Engineering > 90243-(S1) Undergraduate Thesis
Depositing User:	I Putu Krisna Erlangga
Date Deposited:	22 Jul 2024 02:05
Last Modified:	22 Jul 2024 02:05
URI:	http://repository.its.ac.id/id/eprint/108569

Actions (login required)

View Item