Irfan Maulana, . (2026) RANCANG BANGUN APLIKASI KLASIFIKASI BAHASA ISYARAT MENGGUNAKAN METODE MOBILENET-SSD BERBASIS ANDROID. Skripsi thesis, Universitas Pembangunan Nasional Veteran Jakarta.
![[img]](http://repository.upnvj.ac.id/style/images/fileicons/text.png) |
Text
ABSTRAK.pdf Download (218kB) |
|
Text
AWAL.pdf Download (917kB) |
|
Text
BAB I.pdf Restricted to Repository UPNVJ Only Download (265kB) |
|
Text
BAB II.pdf Restricted to Repository UPNVJ Only Download (566kB) |
|
Text
BAB III.pdf Restricted to Repository UPNVJ Only Download (332kB) |
|
Text
BAB IV.pdf Restricted to Repository UPNVJ Only Download (1MB) |
|
Text
BAB V.pdf Download (256kB) |
|
Text
DAFTAR PUSTAKA.pdf Download (239kB) |
|
Text
RIWAYAT HIDUP.pdf Restricted to Repository staff only Download (180kB) |
|
Text
LAMPIRAN.pdf Restricted to Repository UPNVJ Only Download (2MB) |
|
Text
HASIL PLAGIARISME.pdf Restricted to Repository staff only Download (30MB) |
|
Text
ARTIKEL KI.pdf Restricted to Repository staff only Download (802kB) |
Abstract
Communication using the Indonesian Sign Language System (SIBI) remains challenging due to limited public understanding of sign language. This study aims to develop an Android-based sign language classification application using the MobileNet-SSD method. The model was developed with the TensorFlow Object Detection API through transfer learning on SSD MobileNetV2 FPNLite using a dataset of 5,760 images representing 24 SIBI alphabet classes (A to Y, excluding J and Z). The research stages included data preprocessing, an 80:20 training and testing split, model training, evaluation using mean Average Precision (mAP) and recall, TensorFlow Lite conversion, and implementation in an Android application developed with Kotlin. The trained model achieved an mAP of 0.75 to 0.78, mAP@0.50IoU above 0.95, and a recall of 0.88 to 0.90. Testing on the evaluation dataset produced an accuracy of 99.48% with 1,146 correct predictions out of 1,152 test samples, while realtime testing on the Android application achieved an average accuracy of 85.58%. Black Box Testing and User Acceptance Testing confirmed that all application functions operated correctly and were well accepted by users. The application is suitable as a learning medium and a communication aid for Indonesian sign language.
| Item Type: | Thesis (Skripsi) |
|---|---|
| Additional Information: | [No.Panggil: 2010511057] [Pembimbing 1: Ridwan Raafi’udin] [Pembimbing 2: Kharisma Wiati Gusti] [Penguji 1: Musthofa Galih Pradana] [Penguji 2: Ichsan Mardani] |
| Uncontrolled Keywords: | Android. Deep learning, Indonesian Sign Language System (SIBI), MobileNet-SSD, TensorFlow Lite |
| Subjects: | Q Science > QA Mathematics > QA75 Electronic computers. Computer science Q Science > QA Mathematics > QA76 Computer software |
| Divisions: | Fakultas Ilmu Komputer > Program Studi Informatika (S1) |
| Depositing User: | IRFAN MAULANA |
| Date Deposited: | 25 Aug 2026 15:20 |
| Last Modified: | 25 Aug 2026 15:20 |
| URI: | http://repository.upnvj.ac.id/id/eprint/51911 |
Actions (login required)
 |
View Item |