Image of DETEKSI BAHASA ISYARAT INDONESIA (BISINDO) SECARA REAL-TIME DENGAN ARSITEKTUR CONVOLUTIONAL NEURAL NETWORK (CNN) MENGGUNAKAN MODEL MOBILENET
Bookmark Share

Skripsi

DETEKSI BAHASA ISYARAT INDONESIA (BISINDO) SECARA REAL-TIME DENGAN ARSITEKTUR CONVOLUTIONAL NEURAL NETWORK (CNN) MENGGUNAKAN MODEL MOBILENET

Indonesian Sign Language (Bisindo) is the primary language used by the Deaf community to communicate. However, the limited understanding of Bisindo among the general public often creates communication barriers. Therefore, this study aims to develop a real-time Bisindo detection system using the Convolutional Neural Network (CNN) architecture with MobileNet and MobileNetV2 models. The models were trained using image datasets of Bisindo alphabets from A to Z. To enhance model performance, hyperparameter tuning was performed using the Random Search method on several parameters, including batch size, image size, dropout, optimizer, and learning rate. The experimental results show that the MobileNet model achieved the best performance with an accuracy of 96.58%, precision of 96.70%, recall of 96.62%, and F1-score of 96.59%, using a batch size of 32, image size of 224×224×3, dropout of 20%, and the Adam optimizer with a learning rate of 0.001. The letter-by-letter prediction test demonstrated a high confidence level above 90%. However, letters with similar hand shapes such as M and N showed a decrease in confidence to 88%, while less distinguishable letters such as F and Q reached 81% and 84%, respectively. These results indicate that the MobileNet model performs very well in detecting Bisindo gestures in real time.

Availability
#
Central Library (Reference) T1855992025
T185599
Available but not for loan - Not for Loan
Detail Information
Series Title
-
Call Number
T1855992025
Publisher
Indralaya : Prodi Sistem Komputer, Fakultas Ilmu Komputer Universitas Sriwijaya.,
Collation
xiv, 96 hlm.; ilus.; tab.; 29 cm.
Language
Indonesia
ISBN/ISSN
-
Classification
419.07
Content Type
Text
Media Type
unmediated
Carrier Type
other (computer)
Edition
-
Subject(s)
Bahasa Isyarat
Prodi Sistem Komputer
Specific Detail Info
-
Statement of Responsibility
Other version/related
TitleEditionLanguage
DETEKSI SISTEM ISYARAT BAHASA INDONESIA (SIBI) MENGGUNAKAN ARSITEKTUR CONVOLUTIONAL NEURAL NETWORK (CNN) DENGAN METODE RESNET-50 SECARA REAL-TIMEid
PENGENALAN BAHASA ISYARAT INDONESIA (BISINDO) MENGGUNAKAN ALGORITMA CONVOLUTIONAL NEURAL NETWORKid
EKSISTENSI DAN URGENSI PENERJEMAH BAHASA ISYARAT "BISU TULI" DALAM MELANCARKAN PROSES PERADILAN PIDANAid
PENGENALAN KATA DALAM BAHASA ISYARAT INDONESIA (BISINDO) SECARA REAL-TIME MENGGUNAKAN METODE YOU ONLY LOOK ONCE (YOLOv5)id
STRATEGI KOMUNIKASI GERKATIN PALEMBANG (GERAKAN KESEJAHTERAAN TUNARUNGU) DALAM MENSOSIALISASIKAN BAHASA ISYARAT INDONESIA ATAU BISINDO KEPADA MASYARAKAT DI KOTA PALEMBANGid
EKSISTENSI DAN URGENSI PENERJEMAH BAHASA ISYARAT "BISU TULI" DALAM MELANCARKAN PROSES PERADILAN PIDANAid
PENGENALAN BAHASA ISYARAT INDONESIA (BISINDO) MENGGUNAKAN ALGORITMA CONVOLUTIONAL NEURAL NETWORKid
PENERAPAN MACHINE LEARNING DENGAN TENSORFLOW UNTUK MENDETEKSI BAHASA ISYARAT BISINDO BERBASIS APLIKASI ANDROIDid
File Attachment
Comments
You must be logged in to post a comment