Skip to content
Open access

IMPLEMENTATION OF DEEP LEARNING-BASED FACE RECOGNITION FOR A WEB-BASED EMPLOYEE ATTENDANCE SYSTEM

Sep 2026 · Jurnal Pilar Nusa Mandiri · 0 citations

Abstract

Employee attendance recording is a critical component of effective human resource management, yet conventional methods such as attendance cards and manual logging remain vulnerable to data manipulation and recording errors. Universitas Janabadra currently relies on a single fingerprint device, which is inconvenient for staff working far from the attendance point and limited in its reporting capability. This study aims to design and implement a face-recognition-based employee attendance system that leverages deep learning to improve the efficiency, security, and accuracy of attendance recording. The system was developed through five stages: problem identification, requirement analysis, design, implementation, and testing. Face detection and recognition were implemented using the FaceMatcher method from the Face-API.js library, which is based on pre-trained SSD MobileNetV1 and built on TensorFlow.js, while the web-based attendance application was developed using the Laravel framework. Testing was conducted through test cases and compatibility testing under various lighting conditions, distances, and facial accessories. The results show that all functional test scenarios performed successfully, and a quantitative evaluation using 200 recognition trials, comprising 150 attempts from 15 registered employees and 50 attempts from 5 unregistered individuals, achieved an overall accuracy of 79.00%, a precision of 98.21%, a recall of 73.33%, an F1-score of 83.97%, a false acceptance rate (FAR) of 4.00%, and a false rejection rate (FRR) of 26.67%. This system has the potential to become a practical solution for reducing human error and attendance data manipulation in work environments such as Universitas Janabadra, although the relatively high false rejection rate indicates room for improvement in recognition reliability.

Read PDF

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.