Real-Time Facial Verification with Liveness Detection for Mobile-Based Human Resource Information Systems
Journal of Intelligent Computing and Engineering · 01.05.2026 · 2 (1) · 1-10
DOI: 10.67561/jice.v2i1.1 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Intelligent Computing and Engineering, online erschienen am 01.05.2026. Verfasst von Aditya Prapanca Prapanca und 1 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-SA 4.0).
Abstract (Originalsprache)
The increasing adoption of mobile-based Human Resource Information Systems (HRIS) requires reliable and efficient authentication mechanisms to support employee attendance management. This study proposes a real-time facial verification framework that integrates deep facial embeddings, cosine similarity, and eye-blink-based liveness verification within a mobile HRIS application. The proposed method employs a pretrained VGGFace2 model to extract 2048-dimensional facial embeddings, while TensorFlow Lite enables lightweight inference suitable for mobile deployment. During enrollment, facial samples from multiple orientations are collected to represent intra-personal facial variations. During verification, the query embedding is compared with the stored reference representation using cosine similarity, followed by eye-blink detection as an additional liveness condition. The system was evaluated using multiple similarity thresholds and facial acquisition conditions, including forward-facing, dim-lighting, and tilted-face scenarios. The results indicate that verification performance is affected by threshold selection, illumination, and facial orientation. A 70% similarity threshold was selected as the operational configuration based on the evaluated scenarios. Functional testing confirmed that the integrated HRIS features operated according to their expected scenarios. Furthermore, usability evaluation involving 81 respondents achieved an average System Usability Scale score of 78.49, indicating acceptable usability. The findings demonstrate the feasibility of combining lightweight deep facial verification and liveness detection for mobile HRIS-based attendance authentication.
Zitieren
APA
Prapanca, A. P., & Nusantoko, Y. R. (2026). Real-Time Facial Verification with Liveness Detection for Mobile-Based Human Resource Information Systems. Journal of Intelligent Computing and Engineering, 2(1), 1-10. https://doi.org/10.67561/jice.v2i1.1
Vancouver
Prapanca AP, Nusantoko YR. Real-Time Facial Verification with Liveness Detection for Mobile-Based Human Resource Information Systems. Journal of Intelligent Computing and Engineering. 2026;2(1):1-10. doi:10.67561/jice.v2i1.1
BibTeX
@article{Prapanca2026,
author = {Prapanca, Aditya Prapanca and Nusantoko, Yuliarta Rizki},
title = {Real-Time Facial Verification with Liveness Detection for Mobile-Based Human Resource Information Systems},
journal = {Journal of Intelligent Computing and Engineering},
year = {2026},
volume = {2},
number = {1},
pages = {1--10},
doi = {10.67561/jice.v2i1.1}
}
EUVIMED