Pembangkitan Citra Wajah Dari Sketch Wajah Menggunakan Cyclegan

Rasyid, Muhammad Azki (2022) Pembangkitan Citra Wajah Dari Sketch Wajah Menggunakan Cyclegan. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 07211840000072-Undergraduate_Thesis.pdf Restricted to Repository staff only Download (4MB)

Abstract

Penggunaan sketsa wajah merupakan alat bantu yang digunakan lembaga penegak hukum dalam melakukan proses identifikasi tersangka tindak kriminal. Sketsa wajah digunakan ketika tidak terdapat foto dari tersangka tindak kriminal di tempat kejadian perkara. Sketsa wajah digunakan dalam proses identifikasi mugshot pada database dengan menggunakan sistem face recognition, dikarenakan sketsa wajah memiliki modalitas yang berbeda dengan citra wajah seperti halnya tekstur wajah, maka dibangkitkanlah citra wajah baru dari input sketsa wajah yang dimiliki sehingga dapat memiliki tekstur yang dapat menyerupai citra wajah. CycleGAN merupakan metode yang digunakan dalam melakukan tugas image-to-image translation, metode tersebut dapat digunakan dalam melakukan style transfer. Oleh karena itu, dalam penelitian tugas akhir ini dikembangkan sebuah model yang berfungsi untuk membangkitkan citra wajah dari sketsa wajah sehingga dapat mengolah sketsa wajah menjadi citra wajah yang memilki tekstur wajah. Kata Kunci: Deep Learning, GAN, Generative Adversarial Networks, CycleGAN, Style Transfer, Image-to-Image Translation, Face Recognition
==============================================================================================================================
Facial sketches i a tool used by the law enforcement agencies in carrying out the process of identifying criminal suspects. The face sketch is used when there is no photo of the suspected criminal at the scene. Face sketches are used in the system to identify the mugshot in the database using a face recognition. The fact that facial sketches have different modalities from an image such as the texture of the face, so why don’t we just generate a new face image from the face sketch as an input so that they can have the same modality like the facial image. CycleGAN is a method that has been used for image-to-image translation, and proved good to perform style transfer. Therefore, in this final project we developed a model that used to evoke the facial image from the face sketches so that they can be seen as a real face. Keywords: Deep Learning, GAN, Generative Adversarial Networks, CycleGAN, Style Transfer, Image-to-Image Translation, Face Recognition

Item Type:	Thesis (Other)
Additional Information:	RSKom 006.42 Ras p-1 2022
Uncontrolled Keywords:	Deep Learning. GAN. Generative Adversarial Networks. CycleGAN. Style Transfer. Image-to-Image Translation. Face Recognition. Deep Learning. GAN. Generative Adversarial Networks. CycleGAN. Style Transfer. Image-to-Image Translation. Face Recognition.
Subjects:	Q Science > QA Mathematics > QA75 Electronic computers. Computer science. EDP
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Computer Engineering > 90243-(S1) Undergraduate Thesis
Depositing User:	Mr. Marsudiyana -
Date Deposited:	17 Jun 2026 06:01
Last Modified:	17 Jun 2026 06:01
URI:	http://repository.its.ac.id/id/eprint/133854

Actions (login required)

View Item