Skripsi

PENGENALAN BAHASA ISYARAT MENGGUNAKAN SENSOR KINECT BERBASIS DEEP NEURAL NETWORK

---

Penilaian

---

Hand gestures are one of the media that people with disabilities can use to communicate. This difference in the way of communicating results in the deaf and speech impaired having difficulty communicating with normal people. So, a system that can translate sign language into writing is needed. The method used today is still limited to the Indonesian Sign Language System Language (SIBI), while the general language used to communicate is Indonesian Sign Language (BISINDO). In addition, the method used is highly dependent on the accuracy of feature extraction. Thus, in this study, the introduction of BISINDO was developed using themethod Deep Neural Network (DNN). The data used in this study were taken with thesensor Kinect in the form of depth images from 10 respondents with a total of 3600 data obtained consisting of letters and numbers. From the parameters that have been tested, the results of the training show that the use of the stochastic gradient descent optimizer with a hidden layer of 1500 1500, an learning rate RBMof 0.01, and 1500 epochs shows the best accuracy results with thevalue loss lowestcompared to tests using other parameters. This model is then used in tests carried out 5 times for 36 classes. The accuracy results obtained are 95%. Errors that occur can be caused by the similarity of the existing hand sign language, such as the letters D, F, K, O, and R and numbers 2, 4, and 7.

---

Availability

Inventory Code	Barcode	Call Number	Location	Status
2107002878	T58346	T583462021	Central Library (Referens)	Available but not for loan - Not for Loan

Detail Information

Series Title

-

Call Number

T583462021

Publisher

Inderalaya : Jurusan Teknik Elektro, Fakultas Teknik Universitas Sriwijaya., 2021

Collation

xvi, 41 hlm,: ilus.; 29 cm

Language

Indonesia

ISBN/ISSN

-

Classification

681.207

Content Type

Text

Media Type

-

Carrier Type

-

Edition

-

Subject(s)

Sensor
Jurusan Teknik Elektro

Specific Detail Info

-

Statement of Responsibility

MURZ

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment