AI Smart Attendance System Using Face Recognition

AI Smart Attendance System Using Face Recognition

Bhagyashri Wakde1, Mahammad Yunus B2, Mir Naqi Raza3 , Muhammad Adam Khan4 ,Mohammed Muzammil5

1Professor, Dept. of Computer Science Engineering, Rajiv Gandhi Institute of Technology, Bengaluru, India

2Dept. of Computer Science Engineering, Rajiv Gandhi Institute of Technology, Bengaluru, India

3Dept. Of Computer Science Engineering, Rajiv Gandhi Institute of Technology, Bengaluru, India

4Dept. Of Computer Science Engineering, Rajiv Gandhi Institute of Technology, Bengaluru, India

5 Dept. Of Computer Science Engineering, Rajiv Gandhi Institute of Technology, Bengaluru, India

Abstract - In the era of digital transformation and big data, face recognition technology has emerged as one of the most effective biometric solutions for security, authentication, and automation across multiple domains, and this paper presents an advanced AI Smart Attendance System using Face Recognition based on real-time video processing to overcome the limitations of traditional attendance methods such as roll calls and manual signatures, which are time-consuming, error-prone, and susceptible to proxy attendance; the proposed system is designed with a focus on improving recognition accuracy, operational stability, real-time performance, and user-friendly interface design while ensuring scalability and cost-effectiveness. The system operates through four primary phases including database creation, face detection, face recognition, and automated attendance updating, where student facial data is collected and stored securely, faces are detected using Haar Cascade Classifiers, and recognition is performed using Local Binary Pattern Histogram (LBPH) as well as advanced deep learning models such as FaceNet for higher precision, enabling the system to process real-time classroom video streams and identify multiple students simultaneously with high efficiency; experimental results indicate that the system achieves approximately 82% accuracy in standard environments and can reach near 100% accuracy when integrated with cloud-based services such as AWS Rekognition, thereby significantly enhancing reliability and robustness. The implementation can be further extended using modern cloud infrastructure including Amazon S3 for secure storage, AWS Lambda for serverless execution, and DynamoDB for efficient database management, ensuring faster processing, scalability, and data security, while also reducing truancy rates by nearly 60% and minimizing instances of absenteeism, early departures, and classroom disruption. In addition to core functionality, the system introduces several advanced features such as real-time attendance dashboards, automated email notifications to faculty, analytics-based performance reports, and a parent access module, which allows parents or guardians to monitor their child’s attendance through a web or mobile interface, receive instant alerts for absences, and track attendance trends, thereby increasing transparency, accountability, and parental involvement in education; further enhancements include QR-based backup attendance, anti-spoofing mechanisms to prevent fake face detection, mask detection compatibility, mobile camera-based attendance access, and integration with institutional ERP systems. Compared to other biometric systems like fingerprint and iris recognition, the proposed face recognition system is contactless, non-invasive, faster, and more convenient, eliminating congestion and physical interaction, making it highly suitable for modern smart classrooms and organizations; overall, the system provides a comprehensive, intelligent, and scalable solution that automates attendance management, improves academic discipline, enhances monitoring through parental involvement, and contributes significantly to the advancement of smart education systems and institutional productivity.

Key Words: Video Processing, Face Recognition Technology, Face Detection, Face Recognition Attendance System, Automated Attendance System, Real-Time Video Recognition, Haar Cascade Classifier, Local Binary Pattern Histogram (LBPH), Convolutional Neural Network (CNN), Face Identification, OpenCV, Raspberry Pi 3B+, AWS Rekognition, Cloud-Based Attendance, Smart Classroom, Biometric Authentication, Anti-Proxy System, Attendance Analytics, Parent Access and Parental Monitoring, Student Attendance Tracking, Real-Time Notifications, Education Management System.