Artificial Intelligence (AI) enhances attendance monitoring by using facial recognition to automatically identify student in online classes, ensuring accurate and real-time attendance tracking. AI also support continuous monitoring with periodic check and strengthens communication with parents by providing timely updates on student attendance and performance. Additionally, it ensures data privacy with encryption and adaptive algorithms to adjust to varying conditions.
The existing attendance system uses facial recognition to capture student image at the start and end of a session to record attendance and identify student by comparing capture images with stored data, marking them present based on successful matched. It operates by utilizing the camera’s feed to automate the attendance process, ensuring that student participation is tracked without manual intervention.
The existing attendance system primarily focuses on offline classes and lack adaptability for online environment, with limited scalability and issues with camera performance. We propose a web based application named as “Smart Attendance Verification System (SAVS) aimed at enhancing online class attendance monitoring. It includes cloud-based data management, periodic attendance check and the ability to operate in low light environment using standard cameras.
Introduction
1. Background and Motivation
The rise of online education has increased accessibility and flexibility, but it also introduces new challenges:
Inefficient traditional attendance methods (manual roll calls, self-marking)
Proxy attendance
Lack of continuous monitoring
Difficulty verifying participation in virtual settings
There is a growing need for automated, secure, and scalable attendance solutions tailored specifically for online learning environments.
2. Role of Artificial Intelligence in Attendance
AI-based facial recognition systems offer:
Real-time monitoring using computer vision and deep learning
Reduced human intervention and errors
Adaptive performance under different lighting and environmental conditions
Privacy protection via encrypted data handling
Parental communication through automated attendance reports
However, most existing systems are:
Designed for offline or physical classrooms
Dependent on camera quality, lighting, and internet
Not integrated with live virtual platforms like Zoom or Google Meet
Lacking continuous presence verification
3. Proposed Solution: SAVS
The Smart Attendance Verification System (SAVS) is a web-based application designed for real-time, continuous student attendance monitoring in virtual classrooms using facial recognition.
Key Features of SAVS
Continuous Attendance Checks: Conducted at set intervals throughout the session (not just at login).
Mandatory Camera Access: Students cannot join sessions without enabling video.
Parental Verification: Email registration and automated report delivery for transparency.
Integrated Tools: Session recording, screen sharing, and low-light optimization.
Cloud-Based Architecture: Enables scalability and storage of attendance data (MongoDB Atlas/PostgreSQL).
Real-Time Notifications: For students and parents, enhancing accountability and engagement.
4. Methodology and Architecture
A. AI Model Integration
VGG19 (CNN): Used for deep facial feature extraction.
Extracts hierarchical facial features with resilience to lighting and angle changes.
Uses convolution, ReLU activation, and max-pooling operations.
K-Nearest Neighbors (KNN): Classifies students based on facial features.
Calculates Euclidean distance between known and captured features.
Uses majority voting among nearest neighbors to verify identity.
Benefits of VGG19 + KNN
High accuracy
Low computational cost
Real-time performance, even with limited processing power
5. Comparative Analysis: Existing Systems vs. SAVS
Limitations of Current Systems
Offline-focused with no online platform integration
Inadequate for large-scale use
Poor adaptability to lighting, facial occlusions, and varied camera resolutions
Lack of real-time verification and security protocols
Improvements in SAVS
Web-based design with WebRTC for seamless live class integration
Cloud storage and automated analytics
Security features, including encrypted data and verified video access
Optimized for scalability and dynamic online environments
6. Relevance and Impact
In countries like India, where online learning is still growing and infrastructure varies:
SAVS can reduce absenteeism
Improve student engagement and accountability
Enable inclusive participation across urban and rural regions
Support the transition to hybrid and online education models
Conclusion
The Smart Attendance Verification System (SAVS) presents a transformative approach to attendance monitoring in online education by leveraging AI driven facial recognition and real-time automation. Traditional attendance systems often suffer from manual inefficiencies, proxy attendance, and a lack of scalability, which SAVS effectively addresses through automated attendance marking, periodic verification, and parent notifications. The integration of VGG19 for feature extraction and KNN for processing ensures high accuracy, while cloud-based storage and encryption enhance data security.
The comparative analysis with traditional systems highlights SAVS\' superiority in terms of attendance verification success, engagement, transparency, and efficiency. The system fosters greater student participation and trust-building with parents, ensuring a more accountable learning environment. However, challenges like scalability, network dependency, and processing speed remain areas for optimization. Future enhancements such as LMS integration, advanced deep learning models (ResNet, EfficientNet), real-time student engagement tracking, blockchain for secure storage, and an Android mobile application will make SAVS even more efficient, secure, and accessible. As online education continues to evolve, SAVS stands as a robust and innovative solution, ensuring seamless, accurate, and engaging attendance verification in the digital era, particularly benefiting a country like India, where online education still faces significant challenges.
References
[1] M. S. M. Alburaiki, G. M. Johar, R. A. A. Helmi, and M. H. Alkawaz, “Mobile based attendance system: face recognition and location detection using machine learning,” in Proc. 2021 IEEE 12th Control and System Graduate Research Colloquium (ICSGRC), 2021, pp. 177–182.
[2] A. Anshari, S. A. Hirtranusi, D. I. Sensuse, and R. R. Suryono, “Face Recognition for Identification and Verification in Attendance System: A Systematic Review,” in Proc. 2021 IEEE International Conference on Communication, Networks and Satellite (COMNETSAT), 2021, pp. 316–323.
[3] S. Bussa, A. Mani, S. Bharuka, and S. Kaushik, “Smart attendance system using OPENCV based on facial recognition,” Int. J. Eng. Res. Technol., vol. 9, no. 3, pp. 54– 59, 2020.
[4] S. Dey, S. Barman, R. K. Bhukya, R. K. Das, B. C. Haris, S. M. Prasanna, and R. Sinha, “Speech biometric based attendance system,” in Proc. 2014 Twentieth National Conference on Communications (NCC), 2014, pp. 1–6.
[5] A. Kassem, M. Hamad, Z. Chalhoub, and S. El Dahdaah, “An RFID attendance and monitoring system for university applications,” in Proc. 17th IEEE Int. Conf. on Electronics, Circuits, and Systems (ICECS), 2010.
[6] M. Kassim, H. Mazlan, N. Zaini, and M. K. Salleh, “Web-based student attendance system using RFID technology,” in Proc. 2012 IEEE Control and System Graduate Research Colloquium, 2012, pp. 213–218.
[7] S. Khan, A. Akram, and N. Usman, “Real-time automatic attendance system for face recognition using face API and OpenCV,” Wireless Pers. Commun., vol. 113, no. 1, pp. 469– 480, 2020.
[8] R. Rastogi, A. Tyagi, H. Upadhyay, and D. Singh, “Algorithmic Analysis of Automatic Attendance System Using Facial Recognition: A Revolutionary Approach for Future Education,” Int. J. Decis. Support Syst. Technol. (IJDSST), vol. 14, no. 1, pp. 1–19, 2022.
[9] H. Sultan, M. H. Zafar, S. Anwer, A. Waris, H. Ijaz, and M. Sarwar, “Real-Time Face Recognition Based Attendance System For University Classroom,” in Proc. 2022 2nd Int. Conf. on Artificial Intelligence (ICAI), 2022, pp. 165–168.