PENGARUH PERUBAHAN MODEL YOLOV5 DALAM AKURASI DETEKSI EKSPESI WAJAH MENGGUNAKAN KDEF DATASET

Azhar, Febrian Selva (2023) PENGARUH PERUBAHAN MODEL YOLOV5 DALAM AKURASI DETEKSI EKSPESI WAJAH MENGGUNAKAN KDEF DATASET. S2 thesis, Universitas Mercu Buana - Menteng.

[img] Text (Cover)
55420120013-Febrian Selva Azhar-01 Cover - Tugas Kuliah Yolo.pdf

Download (320kB)
[img] Text (Abstrak)
55420120013-Febrian Selva Azhar-02 Abstrak - Tugas Kuliah Yolo.pdf

Download (75kB)
[img] Text (Bab 1)
55420120013-Febrian Selva Azhar-03 BAB 1 - Tugas Kuliah Yolo.pdf
Restricted to Registered users only

Download (145kB)
[img] Text (Bab 2)
55420120013-Febrian Selva Azhar-04 BAB 2 - Tugas Kuliah Yolo.pdf
Restricted to Registered users only

Download (775kB)
[img] Text (Bab 3)
55420120013-Febrian Selva Azhar-05 BAB 3 - Tugas Kuliah Yolo.pdf
Restricted to Registered users only

Download (1MB)
[img] Text (Bab 4)
55420120013-Febrian Selva Azhar-06 BAB 4 - Tugas Kuliah Yolo.pdf
Restricted to Registered users only

Download (295kB)
[img] Text (Bab 5)
55420120013-Febrian Selva Azhar-07 BAB 5 - Tugas Kuliah Yolo.pdf
Restricted to Registered users only

Download (53kB)
[img] Text (Daftar Pustaka)
55420120013-Febrian Selva Azhar-09 daftar pustaka - Tugas Kuliah Yolo.pdf
Restricted to Registered users only

Download (72kB)
[img] Text (Form pernyataan)
02-Form Pernyataan Keaslian dan Persetujuan Unggah dan Publikasi Tugas Akhir_revisi2 - Tugas Kuliah Yolo.pdf
Restricted to Registered users only

Download (91kB)

Abstract

Pengenalan ekspresi wajah merupakan merupakan kecerdasan buatan yang sedang banyak dilakukan penelitian. Pengenalan ekspresi wajah digunakan untuk mengetahui keadaan emosional seseorang, baik senang, sedih, ketakutan, dan keadaan emosional lain pada manusia. kedepan nantinya penelitian pada bidang ini dapat dimanfaatkan untuk dapat digunakan dalam kehidupan manusia. Pada penelitian ini dilakukan pengenalan wajah pada dataset KDEF dengan model deteksi yolov5. dimana dataset KDEF ini terdiri dari 7 class ekspresi yaitu angry, disgust, fear, happy, neutral, sad, dan surprise dengan masing-masing class terdiri dari 420 foto. Dan yolov5 merupakan deep learning berbasis citra yang memiliki beberapa model dengan tipe konvolusi, anchor dan backbone yang berbeda. Pada penelitian ini dilakukan pembandingan model yolov5 untuk mendeteksi dataset ekspresi wajah KDEF. Sehingga akan diketahui mana model yolov5 yang paling optimal untuk digunakan pada deteksi ekspresi wajah dengan dataset KDEF. Kata kunci : Yolov5, KDEF, Ekspresi wajah Recognition of facial expressions is an artificial intelligence is being familiar for researched. Facial expression recognition used to for determine a person's emotional, like happy, sad, scared, and other emotional in human being. On the future, research in this subject can be used full in human life. In this study, facial recognition will carried out on the KDEF dataset with yolov5 detection model. where KDEF dataset consists of 7 expression classes namely angry, disgust, fear, happy, neutral, sad, and surprise with each class consisting of 420 photos. And yolov5 is an image-based deep learning that has several models with different types of convolutions, anchors and backbones. In this study, will compared between the yolov5 model to detect the KDEF facial expression dataset. So that it will be known which yolov5 model is the most optimal for use in facial expression detection with the KDEF dataset. Keyword: Yolov5, KDEF, facial expressions

Item Type: Thesis (S2)
NIM/NIDN Creators: 55420120013
Uncontrolled Keywords: Yolov5, KDEF, Ekspresi wajah Yolov5, KDEF, facial expressions
Subjects: 600 Technology/Teknologi > 620 Engineering and Applied Operations/Ilmu Teknik dan operasi Terapan > 621 Applied Physics/Fisika terapan
Divisions: Pascasarjana > Magister Teknik Elektro
Depositing User: ALFINA DHEA NOVA
Date Deposited: 14 Apr 2023 04:17
Last Modified: 14 Apr 2023 04:17
URI: http://repository.mercubuana.ac.id/id/eprint/76504

Actions (login required)

View Item View Item