IMPLEMENTASI ALGORITMA YOLOv8 PADA SISTEM PENGENALAN HURUF ABJAD BAHASA ISYARAT AMERICAN SIGN LANGUAGE UNTUK KOMUNIKASI NON-VERBAL BERBASIS ANDROID

Muhammad Faturrahman, . (2024) IMPLEMENTASI ALGORITMA YOLOv8 PADA SISTEM PENGENALAN HURUF ABJAD BAHASA ISYARAT AMERICAN SIGN LANGUAGE UNTUK KOMUNIKASI NON-VERBAL BERBASIS ANDROID. Skripsi thesis, Universitas Pembangunan Nasional Veteran Jakarta.

[img] Text
ABSTRAK.pdf
Download (540kB)
[img] Text
AWAL.pdf
Download (1MB)
[img] Text
BAB 1.pdf
Restricted to Repository UPNVJ Only
Download (2MB)
[img] Text
BAB 2.pdf
Restricted to Repository UPNVJ Only
Download (4MB)
[img] Text
BAB 3.pdf
Restricted to Repository UPNVJ Only
Download (1MB)
[img] Text
BAB 4.pdf
Restricted to Repository UPNVJ Only
Download (4MB)
[img] Text
BAB 5.pdf
Download (425kB)
[img] Text
DAFTAR PUSTAKA.pdf
Download (1MB)
[img] Text
DAFTAR RIWAYAT HIDUP.pdf
Restricted to Repository UPNVJ Only
Download (136kB)
[img] Text
LAMPIRAN.pdf
Restricted to Repository UPNVJ Only
Download (3MB)
[img] Text
HASIL PLAGIARISME.pdf
Restricted to Repository staff only
Download (458kB)
[img] Text
ARTIKEL KI.pdf
Restricted to Repository staff only
Download (424kB)

Abstract

The use of sign language as the primary communication tool for the deaf and mute faces challenges in social interaction and access to information, especially without adequate technological support. Technology can help overcome these challenges by enhancing accessibility and communication capabilities. This study aims to develop an Android application that detects the alphabet of American Sign Language (ASL) in real-time using the YOLOv8 model converted to TensorFlow Lite (TFLite) format. Data from the Kaggle platform were used, consisting of 1512 photos for training and testing. The model was trained for 100 epochs, producing a mAP score 0.995, precision 0.986, recall 0.985, and F1 score 0.985. The model was converted to TFLite using Google Colab for efficiency on mobile devices. The Android application was developed with Kotlin, utilizing TFLite for real-time ASL letter inference with 80% accuracy for 26 classes. This study demonstrates the effectiveness of YOLOv8 on mobile devices for ASL letter detection, but improvements in the dataset, data augmentation, and more interactive interface features are needed

Item Type: Thesis (Skripsi)
Additional Information: [No.Panggil:2010511120] [Pembimbing 1: Indra Permana Solihin] [Pembimbing 2: Zatin Niqotaini]] [Penguji 1: Nur Hafifah Matondang] [Penguji 2: Muhammad Adrezo]
Uncontrolled Keywords: sign language, YOLOv8, TensorFlow Lite, Android application, ASL letter detection
Subjects: Q Science > Q Science (General)
T Technology > T Technology (General)
Divisions: Fakultas Ilmu Komputer > Program Studi Informatika (S1)
Depositing User: MUHAMMAD FATURRAHMAN
Date Deposited: 04 Sep 2024 06:48
Last Modified: 04 Sep 2024 06:48
URI: http://repository.upnvj.ac.id/id/eprint/31680

Actions (login required)

View Item View Item