AI Powered Attendance System for Class Rooms using Face Recognition
Authors : Mrs Chandana H M [Assistant Professor at Malnad college of Engineering in the department of Computer Science and Engineering ]
Manoj Banada1, Manoj Maruthi Nayak2, Mohith Gowda3, Yoganandan v4
Affiliations : Malnad College of Engineering, Hassan-573202
November-2024
Abstract
Facial recognition technology is rapidly evolving and becoming a transformative tool in many fields. This paper presents the design and implementation of an AI-powered facial recognition system tailored to a classroom environment. This is different from traditional methods that rely on manual data entry or intrusive biometric systems. Our solution automatically marks attendance using a single frame video capture method. This method significantly reduces the computational cost by guaranteeing high accuracy. This makes it a practical option for educational settings. The proposed system has three main components: a camera module for capturing images in the classroom; A processing module powered by deep learning algorithms for face detection and identification. and a database module for secure attendance record management. This improved workflow This involves detecting faces using extracting data features through an advanced convolutional neural network (CNN) such as FaceNet, and matching the embeddings with stored records using a cosine similarity measurement. It ensures efficient and reliable status awareness. The system was tested in a simulated classroom with 50 students, with a recognition accuracy of 95.2%, an average processing time of 2 seconds per round, and performance consistent with 100 students in a single take. This innovative approach provides a non-intrusive, scalable and efficient solution for attendance automation. Reduce errors by reducing manual effort. Future improvements aim to include real-time monitoring of late arrivals. strength improvement
Keyword :
Facial recognition technology, AI-powered system, classroom environment, attendance automation, video capture method, computational cost, deep learning algorithms, face detection, face identification, Convolutional Neural Network (CNN), FaceNet, data features, cosine similarity, attendance record management, recognition accuracy, processing time, simulation, scalable solution, non-intrusive, real-time monitoring, late arrivals, system performance, database module, secure records
