PERANCANGAN DAN IMPLEMENTASI MODEL REKOGNISI WAJAH MANUSIA SEDERHANA MENGGUNAKAN METODE MOBILENETV2 DAN ARCFACE

HADIPUTRA, SUMARNO and Ubaya, Huda (2021) PERANCANGAN DAN IMPLEMENTASI MODEL REKOGNISI WAJAH MANUSIA SEDERHANA MENGGUNAKAN METODE MOBILENETV2 DAN ARCFACE. Undergraduate thesis, Sriwijaya University.

[thumbnail of RAMA_56201_09011381722084.pdf] Text
RAMA_56201_09011381722084.pdf - Accepted Version
Restricted to Repository staff only
Available under License Creative Commons Public Domain Dedication.

Download (7MB) | Request a copy
[thumbnail of RAMA_56201_09011381722084_TURNITIN.pdf] Text
RAMA_56201_09011381722084_TURNITIN.pdf - Accepted Version
Restricted to Repository staff only
Available under License Creative Commons Public Domain Dedication.

Download (13MB) | Request a copy
[thumbnail of RAMA_56201_09011381722084_0216068101_01_front_ref.pdf]
Preview
Text
RAMA_56201_09011381722084_0216068101_01_front_ref.pdf - Accepted Version
Available under License Creative Commons Public Domain Dedication.

Download (1MB) | Preview
[thumbnail of RAMA_56201_09011381722084_0216068101_02.pdf] Text
RAMA_56201_09011381722084_0216068101_02.pdf - Accepted Version
Restricted to Repository staff only
Available under License Creative Commons Public Domain Dedication.

Download (623kB) | Request a copy
[thumbnail of RAMA_56201_09011381722084_0216068101_03.pdf] Text
RAMA_56201_09011381722084_0216068101_03.pdf - Accepted Version
Restricted to Repository staff only
Available under License Creative Commons Public Domain Dedication.

Download (632kB) | Request a copy
[thumbnail of RAMA_56201_09011381722084_0216068101_04.pdf] Text
RAMA_56201_09011381722084_0216068101_04.pdf - Accepted Version
Restricted to Repository staff only
Available under License Creative Commons Public Domain Dedication.

Download (2MB) | Request a copy
[thumbnail of RAMA_56201_09011381722084_0216068101_05.pdf] Text
RAMA_56201_09011381722084_0216068101_05.pdf - Accepted Version
Restricted to Repository staff only
Available under License Creative Commons Public Domain Dedication.

Download (31kB) | Request a copy
[thumbnail of RAMA_56201_09011381722084_0216068101_06_ref.pdf] Text
RAMA_56201_09011381722084_0216068101_06_ref.pdf - Bibliography
Restricted to Repository staff only
Available under License Creative Commons Public Domain Dedication.

Download (100kB) | Request a copy
[thumbnail of RAMA_56201_09011381722084_0216068101_07_lamp.pdf] Text
RAMA_56201_09011381722084_0216068101_07_lamp.pdf - Accepted Version
Restricted to Repository staff only
Available under License Creative Commons Public Domain Dedication.

Download (1MB) | Request a copy

Abstract

When recognized in an image or photo, it is easy for us as humans, but not with a computer. In order to recognize a human face, special treatment is needed so that when given an input image or photo, the computer can recognize that face recognition is a face recognition technique. Biometrics utilizes an image of a human face which is then measured using a systematic calculation so that it can be recognized by the system. This study focuses on human face recognition using the MobileNetV2 and ArcFace methods. The experiment was carried out with 4 face classes and the highest success was able to recognize faces of 10 out of 10 recognized faces. While the tuning process performs 3 tests with different tuning parameter models where model 3 uses Batch Size 32 parameters, Learning Rate 10e-5, and 100 epochs. Based on the results of model 3 using MobileNetV2 and ArcFace to get an accuracy of 85% and a loss of 0.53%, and for GPU performance there is a simple model 2 which gets an average GPU performance during tuning of 24%.

Item Type: Thesis (Undergraduate)
Uncontrolled Keywords: rekognisi wajah, mobilenetv2, arcface
Subjects: Q Science > Q Science (General) > Q334-342 Computer science. Artificial intelligence. Algorithms. Robotics. Automation.
Divisions: 09-Faculty of Computer Science > 56201-Computer Systems (S1)
Depositing User: Users 5579 not found.
Date Deposited: 06 Aug 2021 01:28
Last Modified: 06 Aug 2021 01:28
URI: http://repository.unsri.ac.id/id/eprint/51733

Actions (login required)

View Item View Item