WIBAWA, BHAKAS ADHITYA ANUGRAH (2026) PEMODELAN MT-GCNET INTEGRASI MTCNN, GLCM, CNN BERBASIS FACENET UNTUK VERIFIKASI KEHADIRAN REMOTE PEKERJA LAPANGAN. S1 thesis, Universitas Mercu Buana Jakarta.
|
Text (HAL COVER)
Cover.pdf Download (437kB) | Preview |
|
![[img]](https://repository.mercubuana.ac.id/style/images/fileicons/text.png) |
Text (BAB I)
Bab 1.pdf Restricted to Registered users only Download (39kB) |
|
|
Text (BAB II)
Bab 2.pdf Restricted to Registered users only Download (722kB) |
|
|
Text (BAB III)
Bab 3.pdf Restricted to Registered users only Download (227kB) |
|
|
Text (BAB IV)
Bab 4.pdf Restricted to Registered users only Download (1MB) |
|
|
Text (BAB V)
Bab 5.pdf Restricted to Registered users only Download (81kB) |
|
|
Text (DAFTAR PUSTAKA)
Daftar Pustaka.pdf Restricted to Registered users only Download (107kB) |
|
|
Text (LAMPIRAN)
Lampiran.pdf Restricted to Registered users only Download (1MB) |
Abstract
The increasing mobility of field personnel and the widespread adoption of remote working arrangements have intensified the demand for attendance systems that ensure high accuracy, reliability, and robust performance under dynamic environmental conditions. Conventional attendance systems remain constrained by limitations in identity verification and vulnerability to fraudulent manipulation. This study proposes and implements a remote attendance system based on facial recognition by introducing a novel hybrid architecture, MT-GCNet, which integrates Multi-task Cascaded Convolutional Neural Network (MTCNN) for face detection, Gray Level Co-occurrence Matrix (GLCM) for texture feature extraction, and a FaceNet-based Convolutional Neural Network (CNN) for facial representation. The proposed framework was evaluated under varying illumination conditions and partial facial occlusions caused by protective mask usage, while also comparing the effectiveness of GLCM features using Softmax and Multilayer Perceptron (MLP) classifiers. Experimental results demonstrate that MT-GCNet with the MLP classifier achieved 98.53% accuracy, 98.62% precision, 98.52% recall, and an F1-score of 98.51%, representing a 0.74% improvement over the baseline model. In real-time deployment, the system attained an average recognition accuracy of 89.8% with a mean response time of 5.2 seconds under bright indoor illumination (76 lux). Furthermore, verification experiments involving highly similar facial pairs (look-alike subjects) achieved 100% verification accuracy with a False Acceptance Rate (FAR) of 0% and a False Rejection Rate (FRR) of 0% under both Full Face and Half Face conditions. These findings demonstrate that the integration of texture descriptors and facial embeddings within the proposed MT-GCNet architecture substantially enhances identity discrimination while preserving robust verification performance under facial occlusion, thereby establishing its potential as a secure, reliable, and adaptive remote attendance solution for field personnel. Kata kunci: Face Recognition, MTCNN, GLCM, FaceNet, MLP Peningkatan mobilitas karyawan lapangan dan penerapan sistem kerja jarak jauh menuntut sistem presensi yang memiliki akurasi dan reliabilitas tinggi serta mampu beroperasi pada kondisi lingkungan yang dinamis. Presensi konvensional masih memiliki keterbatasan, terutama dalam validitas identitas dan ketahanan terhadap manipulasi data. Penelitian ini bertujuan merancang dan mengimplementasikan sistem presensi jarak jauh berbasis pengenalan wajah dengan mengusulkan arsitektur hibrida MT-GCNet, yang mengintegrasikan Multi-task Cascaded Convolutional Neural Network (MTCNN) sebagai detektor wajah, Gray Level Cooccurrence Matrix (GLCM) sebagai ekstraksi fitur tekstur, dan Convolutional Neural Network (CNN) berbasis FaceNet. Sistem dievaluasi pada berbagai kondisi iluminasi dan oklusi parsial akibat penggunaan masker pelindung, serta membandingkan efektivitas fitur GLCM pada pengklasifikasi Softmax dan Multilayer Perceptron (MLP). Hasil penelitian menunjukkan bahwa MT-GCNet dengan pengklasifikasi MLP menghasilkan akurasi 98,53%, precision 98,62%, recall 98,52%, dan F1-score 98,51%, meningkat 0,74% dibandingkan model baseline. Pada implementasi real-time, sistem mencapai rata-rata akurasi pengenalan 89,8% dengan waktu respons 5,2 detik pada kondisi iluminasi dalam ruangan terang (76 lux). Pengujian verifikasi pada pasangan subjek dengan tingkat kemiripan wajah tinggi (look-alike) menunjukkan akurasi 100% dengan False Acceptance Rate (FAR) 0% dan False Rejection Rate (FRR) 0% pada kondisi Full Face maupun Half Face. Hasil tersebut menunjukkan bahwa integrasi fitur tekstur dan embedding wajah pada MT-GCNet mampu meningkatkan kemampuan diskriminasi identitas serta mempertahankan keandalan verifikasi pada kondisi wajah beroklusi, sehingga berpotensi menjadi solusi presensi jarak jauh yang tangguh, aman, dan adaptif bagi pekerja lapangan. Kata kunci: Pengenalan Wajah, MTCNN, GLCM, FaceNet, MLP
| Item Type: | Thesis (S1) |
|---|---|
| NIM/NIDN Creators: | 41522010265 |
| Uncontrolled Keywords: | Pengenalan Wajah, MTCNN, GLCM, FaceNet, MLP |
| Subjects: | 000 Computer Science, Information and General Works/Ilmu Komputer, Informasi, dan Karya Umum > 000. Computer Science, Information and General Works/Ilmu Komputer, Informasi, dan Karya Umum > 004 Data Processing, Computer Science/Pemrosesan Data, Ilmu Komputer, Teknik Informatika 100 Philosophy and Psychology/Filsafat dan Psikologi > 150 Psychology/Psikologi > 153 Conscious Mental Process and Intelligence/Intelegensia, Kecerdasan Proses Intelektual dan Mental > 153.1 Memory and Learning/Memori dan Pembelajaran > 153.12 Memory/Memori > 153.124 Recognition/Pengenalan |
| Divisions: | Fakultas Ilmu Komputer > Informatika |
| Depositing User: | khalimah |
| Date Deposited: | 01 Sep 2026 06:59 |
| Last Modified: | 01 Sep 2026 06:59 |
| URI: | http://repository.mercubuana.ac.id/id/eprint/103541 |
Actions (login required)
 |
View Item |
