Skripsi

IMPLEMENTASI CONVOLUTIONAL NEURAL NETWORK (CNN) DENGAN ARSITEKTUR EFFICIENTNETV2 UNTUK DETEKSI DEEPFAKE PADA GAMBAR WAJAH MANUSIA

---

The rapid advancement of deepfake technology poses significant challenges, as it enables the generation of highly realistic synthetic facial images that are increasingly difficult to distinguish from authentic ones. This development raises substantial concerns regarding information verification and biometric security. This study aims to address these issues by implementing a deepfake detection system utilizing the EfficientNetV2-B2 architecture, which is recognized for its computational efficiency, training stability, and high accuracy in image processing tasks. The research employs two training strategies, namely transfer learning and fine-tuning. The experimental results indicate that the fine-tuning approach consistently outperforms transfer learning. The fine-tuned model achieved an accuracy of 0.91 and an AUC of 0.98 on the first dataset, while obtaining perfect performance with an accuracy and AUC of 1.00 on the second dataset. In contrast, the transfer learning models demonstrated lower performance across both datasets. These findings confirm that EfficientNetV2-B2 combined with fine-tuning is highly effective and reliable for detecting deepfake images with high accuracy.

---

Availability

Detail Information

Series Title

-

Call Number

T1897952025

Publisher

Indralaya : Prodi Sistem Komputer, Fakultas Ilmu Komputer Universitas Sriwijaya., 2025

Collation

xv, 100 hlm.; ilus.; tab.; 29 cm

Language

Indonesia

ISBN/ISSN

-

Classification

006.320 7

Content Type

Text

Media Type

-

Carrier Type

-

Edition

-

Subject(s)

Prodi Sistem Komputer
Artificial Neural Networks

Specific Detail Info

-

Statement of Responsibility

MI

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment