IMPLEMENTASI METODE DEEP CONVOLUTIONAL NEURAL NETWORK (CNN) MODEL FACENET SEBAGAI PENCARIAN WAJAH UNTUK REVERSE IMAGE SEARCH

Rachmansyah, Muhammad Rayhan (2025) IMPLEMENTASI METODE DEEP CONVOLUTIONAL NEURAL NETWORK (CNN) MODEL FACENET SEBAGAI PENCARIAN WAJAH UNTUK REVERSE IMAGE SEARCH. Undergraduate thesis, UPN Veteran Jawa Timur.

	Text (Cover) 19081010002-cover.pdf Download (1MB)
	Text (Bab 1) 19081010002-bab1.pdf Download (42kB)
	Text (Bab 2) 19081010002-bab2.pdf Restricted to Repository staff only until 5 December 2027. Download (332kB)
	Text (Bab 3) 19081010002-bab3.pdf Restricted to Repository staff only until 5 December 2027. Download (1MB)
	Text (Bab 4) 19081010002-bab4.pdf Restricted to Repository staff only until 5 December 2027. Download (4MB)
	Text (Bab 5) 19081010002-bab5.pdf Download (97kB)
	Text (Daftar Pustaka) 19081010002-daftarpustaka.pdf Download (161kB)
	Text (Lampiran) 19081010002-lampiran.pdf Restricted to Repository staff only until 5 December 2027. Download (148kB)

Abstract

This study discusses the implementation of the Deep Convolutional Neural Network (CNN) method using the FaceNet model as a facial search function for a reverse image search system. The main motivation behind this research is the increasing need for fast, accurate, and efficient facial identification across various digital application domains. FaceNet was selected due to its ability to transform facial images into vector embeddings whose distances can be measured to determine the degree of similarity between faces. The research methodology includes the processes of face detection, embedding extraction, system training and testing, as well as performance evaluation based on the distances between embeddings. Testing was conducted on sets of images with both same and different identities to assess model performance. The results show that the average embedding distance for same-identity images is 0.3980, while for different identities it is 0.6532. The model also achieved a high accuracy level, with an overall accuracy of 0.95835 across 100 test samples. Based on these findings, the implementation of FaceNet is proven to be effective for facial retrieval in reverse image search systems, demonstrating strong accuracy and consistent capability in distinguishing facial similarity. This study indicates that CNN-based embeddings can serve as a reliable and efficient foundation for visual search systems.

Item Type:

Thesis (Undergraduate)

Contributors:

Contribution	Contributors	NIDN/NIDK	Email
Thesis advisor	Rahmat, Basuki	NIDN0023076907	basukirahmat.if@upnjatim.ac.id
Thesis advisor	Wahanani, Henni Endah	NIDN0022097811	henniendah.if@upnjatim.ac.id

Subjects:

T Technology > T Technology (General)

Divisions:

Faculty of Computer Science > Departemen of Informatics

Depositing User:

Muhammad Rayhan Rachmansyah

Date Deposited:

08 Dec 2025 01:23

Last Modified:

08 Dec 2025 02:15

URI:

https://repository.upnjatim.ac.id/id/eprint/47207

Actions (login required)

View Item