INOVASI MEDIA ICT DETEKSI WAJAH UNTUK PRESENSI DIGITAL MAHASISWA: ANALISIS AKURASI DAN PERFORMA
DOI:
https://doi.org/10.51878/academia.v6i4.13792Keywords:
Media ICT Pembelajaran, Deteksi Wajah, Presensi Digital MahasiswaAbstract
ABSTRACT
The digital transformation of higher education demands ICT-based learning media innovations that support accurate, efficient, and contactless classroom governance, particularly in student attendance validation. This study analyzes the accuracy and performance of two web-based face detection libraries, OpenCV.js and face-api.js, as ICT media for student digital attendance processed entirely on the client side without biometric data transmission, in line with the privacy-by-design principle. The method employed is a controlled laboratory experiment with a 36-scenario factorial design: three lighting levels (100, 300, 500 lux) × four pose angles (0°, 15°, 30°, 45°) × three device tiers (low-end, mid-range, high-end), involving 50 student participants aged 18–19 years. Results across 180 observations per library show that face-api.js excels in detection accuracy (recall 97.8%; F1-Score 98.9%) compared to OpenCV.js (recall 81.7%; F1-Score 89.9%), while OpenCV.js excels in speed (latency 163.5 ms; FPS-eq 6.1) compared to face-api.js (269.5 ms; FPS-eq 3.7). Both libraries achieved 100% precision in a closed-set scenario with latency well below 500 ms, validating technical feasibility as a Proof of Concept at Technology Readiness Level 3. Both client-side pathways were 1.65–2.51 times faster than the server-side baseline (410.1 ms). These findings contribute to lightweight, cost-effective, and privacy-oriented ICT learning media for digital attendance in higher education.
ABSTRAK
Transformasi digital pendidikan tinggi menuntut inovasi media ICT yang mendukung tata kelola pembelajaran secara lebih akurat, efisien, dan nirkontak, khususnya dalam validasi kehadiran mahasiswa. Penelitian ini menganalisis akurasi dan performa dua pustaka deteksi wajah berbasis web, yaitu OpenCV.js dan face-api.js, sebagai media ICT presensi digital mahasiswa yang diproses sepenuhnya di sisi klien (client-side) tanpa transmisi data biometrik, sejalan dengan prinsip privacy-by-design. Metode yang digunakan adalah eksperimen laboratorium terkendali dengan desain faktorial 36 skenario: tiga tingkat pencahayaan (100, 300, 500 lux) × empat sudut pose (0°, 15°, 30°, 45°) × tiga kelas perangkat (low-end, mid-range, high-end), dengan 50 mahasiswa berusia 18–19 tahun sebagai partisipan. Hasil pengujian pada 180 observasi per pustaka menunjukkan bahwa face-api.js unggul dalam akurasi deteksi (recall 97,8%; F1-Score 98,9%) dibandingkan OpenCV.js (recall 81,7%; F1-Score 89,9%), sementara OpenCV.js unggul dalam kecepatan (latensi 163,5 ms; FPS-eq 6,1) dibandingkan face-api.js (269,5 ms; FPS-eq 3,7). Kedua pustaka mencatat precision 100% dalam skenario closed-set dan latensi jauh di bawah 500 ms, memvalidasi kelayakan teknis sebagai Proof of Concept TKT Level 3. Kedua jalur client-side juga 1,65–2,51 kali lebih cepat daripada baseline server-side (410,1 ms). Temuan ini berkontribusi pada pengembangan media ICT pembelajaran yang ringan, hemat biaya, dan berorientasi privasi bagi presensi digital di perguruan tinggi.
Downloads
References
Budiman, A., Fabian, Yaputera, R. A., Achmad, S., & Kurniawan, A. (2023). Student attendance with face recognition (LBPH or CNN): Systematic literature review. Procedia Computer Science, 216, 31–38. https://doi.org/10.1016/j.procs.2022.12.108
Ennajar, S., & Bouarifi, W. (2024). Deep transfer learning approach for student attendance system during the COVID-19 pandemic. Journal of Computer Science, 20(3), 229–238. https://doi.org/10.3844/jcssp.2024.229.238
Fang, Z., & Zhou, Z. (2023). Studies advanced in robust face recognition under complex light intensity. Dalam Proceedings of the 2023 International Conference on Image, Algorithms and Artificial Intelligence (ICIAAI) (pp. 1005–1012). https://doi.org/10.2991/978-94-6463-300-9_101
George, A., Ecabert, C., Otroshi Shahreza, H., Kotwal, K., & Marcel, S. (2024). EdgeFace: Efficient face recognition model for edge devices. IEEE Transactions on Biometrics, Behavior, and Identity Science, 6(2), 158–168. https://doi.org/10.1109/TBIOM.2024.3352164
Goh, H.-A., Ho, C.-K., & Abas, F. S. (2022). Front-end deep learning web apps development and deployment: A review. Applied Intelligence, 53(12), 15923–15945. https://doi.org/10.1007/s10489-022-04278-6
Huang, K., Teng, Y., Chen, Y., & Wang, Y. (2024). From pixels to principles: A decade of progress and landscape in trustworthy computer vision. Science and Engineering Ethics, 30, 26. https://doi.org/10.1007/s11948-024-00480-6
Huang, Y., Wu, Z., Chen, J., & Xiang, H. (2024). Privacy-preserving face recognition method based on randomization and local feature learning. Journal of Imaging, 10(3), 59. https://doi.org/10.3390/jimaging10030059
Justad, V. M. (2023). face-api.js: A JavaScript face recognition API for the browser [Perangkat lunak komputer]. GitHub. https://github.com/justadudewhohacks/face-api.js
Kamil, M. H. M., Zaini, N., Mazalan, L., & Ahamad, A. H. (2023). Online attendance system based on facial recognition with face mask detection. Multimedia Tools and Applications, 82(22), 34437–34457. https://doi.org/10.1007/s11042-023-14842-y
Kementerian Riset dan Teknologi/Badan Riset dan Inovasi Nasional. (2022). Pedoman penyusunan Tingkat Kesiapterapan Teknologi (TKT). Kemenristek/BRIN.
Nielsen, J. (1993). Usability engineering. Morgan Kaufmann.
OpenCV.org. (2024). OpenCV.js: Computer vision library for the web. https://docs.opencv.org/4.x/d5/d10/tutorial_js_root.html
Oroceo, P. P., Kim, J.-I., Caliwag, E. M. F., Kim, S.-H., & Lim, W. (2022). Optimizing face recognition inference with a collaborative edge–cloud network. Sensors, 22(21), 8371. https://doi.org/10.3390/s22218371
Sholi, R. T., Sarker, M. F. H., Sohel, M. S., Islam, M. K., Tamal, M. A., Bhuiyan, T., Shakil, S. M. K. A., & Ahmed, M. F. (2024). Application of computer vision and mobile systems in education: A systematic review. International Journal of Interactive Mobile Technologies, 18(1), 168–187. https://doi.org/10.3991/ijim.v18i01.46483
Sismadi, W., Indra, Martono, B. A., & Widyastuti, T. (2022). Comparative analysis of Codeigniter, Laravel and Ktupad frameworks: Case study online exam applications. Indonesian Journal of Applied Research, 3(3), 207–219. https://iojs.unida.info/index.php/IJAR/article/view/236
Sismadi, W., Martono, B. A., Susanto, Y., & Muzaeni, A. (2025). Perbandingan performa framework MVC dalam sistem kehadiran berbasis pengenalan wajah. EDUTECH: Jurnal Inovasi Pendidikan Berbantuan Teknologi, 5(1), 1–7. https://doi.org/10.51878/edutech.v5i1.4405
Song, Z., Wang, G., Yang, W., Li, Y., Yu, Y., Wang, Z., Zheng, X., & Yang, Y. (2025). Privacy-preserving method for face recognition based on homomorphic encryption. PLOS ONE, 20(2), e0314656. https://doi.org/10.1371/journal.pone.0314656
Ulhaq, M. R. D., Zaidan, M. A., & Firdaus, D. (2023). Pengenalan ekspresi wajah secara real-time menggunakan metode SSD Mobilenet berbasis Android. Journal of Technology Informatics (JoTI), 5(1), 48–52. https://doi.org/10.37802/joti.v5i1.387
Wang, Q., Jiang, S., Chen, Z., Cao, X., Li, Y., Li, A., Ma, Y., Cao, T., & Liu, X. (2025). Anatomizing deep learning inference in Web browsers. ACM Transactions on Software Engineering and Methodology, 34(2), Article 47, 1–43. https://doi.org/10.1145/3688843
Zhao, X., Wang, L., Zhang, Y., Han, X., Deveci, M., & Parmar, M. (2024). A review of convolutional neural networks in computer vision. Artificial Intelligence Review, 57, 99. https://doi.org/10.1007/s10462-024-10721-6
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 Wawan Sismadi, Abdul Manan, Estuti Fitri Hartini
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
