Implementasi Metode CNN Menggunakan Arsitektur YOLOv8 untuk Menerjemahkan Bahasa Isyarat Indonesia: BISINDO

Farhana, Farizah (2025) Implementasi Metode CNN Menggunakan Arsitektur YOLOv8 untuk Menerjemahkan Bahasa Isyarat Indonesia: BISINDO. Undergraduate thesis, UPN Veteran Jawa Timur.

	Text (Cover) 21082010197-cover.pdf Download (693kB)
	Text (Bab 1) 21082010197-bab1.pdf Download (95kB)
	Text (Bab 2) 21082010197-bab2.pdf Restricted to Repository staff only until 15 July 2027. Download (1MB)
	Text (Bab 3) 21082010197-bab3.pdf Restricted to Repository staff only until 15 July 2027. Download (320kB)
	Text (Bab 4) 21082010197-bab4.pdf Restricted to Repository staff only until 15 July 2027. Download (1MB)
	Text (Bab 5) 21082010197-bab5.pdf Download (10kB)
	Text (Daftar Pustaka) 21082010197-daftarpustaka.pdf Download (91kB)
	Text (Lampiran) 21082010197-lampiran.pdf Restricted to Repository staff only Download (184kB)

Abstract

Many deaf individuals in Indonesia use sign language as their means of communication, one of which is BISINDO. However, the limited understanding of BISINDO sign language becomes an obstacle to effective communication with the general public. Therefore, this study develops a BISINDO hand gesture detection system based on computer vision using the YOLOv8 architecture. The method used in this study is Convolutional Neural Network (CNN) with the You Only Look Once version 8 (YOLOv8) architecture, which is trained using a dataset obtained from the Roboflow framework. This system is built using the Python programming language and the YOLOv8 architecture implemented into the Streamlit interface, with input in the form of images from the camera and uploads. The input will be detected and translated into text and voice using Google Text-To-Speech (gTTS). Testing was conducted and showed that the system was able to detect and translate BISINDO sign language hand gestures in semi-real-time with an accuracy of 89.74%, based on the performance evaluation of the model consisting of TP, FP, FN, and TN values. In addition, the model achieved a precision of 89.28%, recall of 96.15%, and F1-Score of 92.16%. This indicates that YOLOv8 is effective for recognizing BISINDO sign language hand gestures, as it successfully predicted the correct labels, although there were still detection errors under certain conditions.

Item Type:

Thesis (Undergraduate)

Contributors:

Contribution	Contributors	NIDN/NIDK	Email
Thesis advisor	Najaf, Abdul Rezha Efrat	NIDN0029099403	rezha.efrat.sifo@upnjatim.ac.id
Thesis advisor	Permatasari, Reisa	NIDN0014059203	reisa.permatasari.sifo@upnjatim.ac.id

Subjects:

Q Science > QA Mathematics > QA76.87 Neural computers

Divisions:

Faculty of Computer Science > Departemen of Information Systems

Depositing User:

Farizah Farhana

Date Deposited:

22 Jul 2025 03:58

Last Modified:

22 Jul 2025 03:58

URI:

https://repository.upnjatim.ac.id/id/eprint/39556