Facial Recognition Based Attendance Tracking System

Abstract

Traditional attendance systems are vulnerable to fraudulent practices and inefficiencies. This paper proposes a facial recognition-based attendance tracking system enhanced with anti-spoofing capabilities using YOLOv5. By integrating real-time face detection, recognition, and liveness verification, the system ensures accurate and secure attendance logging. It further integrates with a web portal via ThingSpeak API to notify absenteeism. Built using Python libraries including OpenCV, face_recognition, and TensorFlow, the system captures live video streams, detects faces, and matches them against a pre-encoded database. A GUI using Tkinter allows intuitive interaction, and the system is tested for reliability under varied lighting and angles. The proposed model significantly minimizes manual effort and fraudulent entries while providing scalability and data-driven insights for educational institutions.

Introduction

This project proposes a real-time, AI-powered facial recognition attendance system that addresses the limitations of traditional attendance methods and enhances security, accuracy, and efficiency. It incorporates YOLOv5 for face detection and anti-spoofing techniques to prevent fraudulent entries using photos or videos. The system is designed for use in educational institutions and workplaces, offering a contactless, automated, and tamper-proof attendance solution.

Key Points:

1. Motivation & Challenges:

• Traditional systems are error-prone and vulnerable to proxy attendance.

• Basic facial recognition is easily spoofed with images/videos.

• The solution integrates liveness detection using YOLOv5 to address these concerns.

2. Literature Review Insights:

• Prior work used models like Haar Cascade + LBPH, FaceNet, and CNNs.

• Limitations included poor lighting tolerance, need for frontal faces, and high computational cost.

• Researchers also highlighted privacy, deployment complexity, and dataset bias as challenges.

3. Proposed System Architecture:

• Face Detection: YOLOv5 ensures fast and accurate detection, even in crowded environments.

• Liveness Detection: A deep-learning module detects spoofing attacks.

• Face Recognition: Pre-trained models (FaceNet, VGG-Face) extract high-dimensional facial embeddings.

• Attendance Logging: Recognized faces are logged in real-time with timestamps.

• User Interface: Admin dashboard and student portal provide attendance reports and data management tools.

4. Methodology Overview:

• Real-time video is captured via webcam.

• Each frame undergoes preprocessing, face detection, anti-spoofing, and recognition.

• Facial Features: Extracted using Dlib’s 68-point landmarks and processed with normalization, grayscale conversion, and HAAR features.

• LBPH Algorithm: Used for recognition, training on labeled images to identify individuals.

5. Features & Benefits:

• Highly secure (spoof-proof) and scalable.

• Enables automated attendance marking with minimal manual input.

• Can be integrated with cloud storage, mobile apps, and Learning Management Systems (LMS).

• Provides real-time analytics, error prevention, and better administrative control.

Conclusion

The Face Attendance Detection System is a cutting-edge application of modern technology in attendance management. By integrating machine learning techniques, particularly deep learning for face detection and recognition, the project greatly enhances the traditional attendance taking process. The addition of a face spoofing detection capability ensures the authenticity of applications, deterring potential misuse. The robust combination of tools and frameworks, such as OpenCV and YOLOv5, provides an effective solution that is adaptable across various environments, including education and corporate settings. With real-time feedback and reporting, this system opens new avenues for monitoring and attendance management, significantly reducing administrative workloads and improving accuracy.

References

[1] Ahonen,T.,Hadid,A.,&Pietikäinen,M.(2006).Facerecognitionwithlocalbinary patterns. IEEE Transactions on Pattern Analysis and Machine Intelligence, 28(12), 20372041. [2] Turk, M., & Pentland, A. (1991). Face recognition using eigenfaces. IEEE Transactions on PatternAnalysis and Machine Intelligence, 13(7), 586-591. [3] Zhang,X.,&Gao,Y.(2016).Facerecognitionusingdeeplearningtechniques.IEEE Transactions on Neural Networks and Learning Systems, 27(11), 2367-2376. [4] Satake, S., & Miwa, H. (2017). Face recognition-based attendance system using convolutional neural networks. International Journal ofAdvanced Research in Computer Science, 8(3), 355-362. [5] Singh, R., & Vatsa, M. (2018). Face recognition in unconstrained environments: A review. IEEE Access, 6, 33634-33653. [6] Kaur,A., & Kaur, R. (2019). Automated attendance system using face recognition. 2019 International Conference on Machine Learning, Big Data, and IoT (MLBDI), 1-5. [7] Patel, R., & Patel, S. (2020). Face recognition-based attendance system using deep learning. 2020 International Conference on Emerging Trends in Information Technology (ICETIT), 1-5. [8] Jain,A., & Jain, S. (2018). Face recognition attendance systemusing convolutional neural networks. 2018 International Conference on Computing, Communication and Networking (ICCCN), 1-5. [9] Satake, S. (2017). Face recognition-based attendance system using convolutional neural networks. Master\'s thesis, University of Tokyo.

Copyright

Copyright © 2025 Kiruthika Priya V, Priya Sharma, Kavitha K, Lakshana S, Dr. K. Rajakumari. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.