Program Penerjemah Bahasa Isyarat Indonesia (BISINDO) Secara Real Time Menggunakan Convolutional Neural Network Dan Mediapipe

Therry, Renaldy William Lijaya (2024) Program Penerjemah Bahasa Isyarat Indonesia (BISINDO) Secara Real Time Menggunakan Convolutional Neural Network Dan Mediapipe. Undergraduate thesis, UPN Veteran Jawa Timur.

	Text (Cover) 20081010179-cover.pdf Download (1MB)
	Text (Bab 1) 20081010179-bab1.pdf Download (460kB)
	Text (Bab 2) 20081010179-bab2.pdf Restricted to Repository staff only until 19 September 2026. Download (927kB)
	Text (Bab 3) 20081010179-bab3.pdf Restricted to Repository staff only until 19 September 2026. Download (702kB)
	Text (Bab 4) 20081010179-bab4.pdf Restricted to Repository staff only until 19 September 2026. Download (2MB)
	Text (Bab 5) 20081010179-bab5.pdf Download (438kB)
	Text (Daftar Pustaka) 20081010179-daftarpustaka.pdf Download (411kB)

Abstract

Communication is one of the essential aspects of daily life, but for the deaf and mute community, communication can be a significant challenge. Indonesian Sign Language (BISINDO) is one of the communication tools used by this community, but many people in general still do not understand sign language. This can lead to difficulties in communication and limit the access of the deaf and mute community to participate in social and economic activities. Therefore, developing an accurate sign language detection system is crucial to facilitate communication between the deaf, mute community, and the general public. In this study, the author developed a program to translate Indonesian Sign Language using Convolutional Neural Network (CNN) and MediaPipe. The program has limitations, namely that the detection process is only focused on the right hand and the program can only detect the BISINDO alphabet. The model training results show an accuracy of 98.59%, a loss of 13.03%, a validation accuracy of 93.78%, and a validation loss of 32.15%. In the model evaluation results, the model can detect the Indonesian sign language alphabet with an accuracy of 94%. The detection accuracy test results in four scenarios showed the best detection results in the first scenario, which involved good lighting and a dark background, with an accuracy of 72%, a precision of 69.5%, a recall of 68%, and an F1-score of 68.7%.

Item Type:

Thesis (Undergraduate)

Contributors:

Contribution	Contributors	NIDN/NIDK	Email
Thesis advisor	Junaidi, Achmad	NIDN0710117803	achmadjunaidi.if@upnjatim.ac.id
Thesis advisor	Sihananto, Andreas Nugroho	NIDN0012049005	andreas.nugroho.jarkom@upnjatim.ac.id

Subjects:

P Language and Literature > P Philology. Linguistics > P99 Semiotics. Signs and Symbols
P Language and Literature > P Philology. Linguistics > P99.5 Non Verbal communication
Q Science > Q Science (General)
Q Science > QM Human anatomy
T Technology > T Technology (General)
P Language and Literature > P Philology. Linguistics > P93.5 Visual Communication

Divisions:

Faculty of Computer Science > Departemen of Informatics

Depositing User:

Renaldy William Lijaya Therry

Date Deposited:

20 Sep 2024 03:42

Last Modified:

20 Sep 2024 03:42

URI:

https://repository.upnjatim.ac.id/id/eprint/29637