Smart Vision: CNN-Based Attendance System

Abstract

Attendance management plays a vital role in academic institutions and organizations; however, traditional approaches such as manual registers, identity cards, and fingerprint-based systems suffer from inefficiency, lack of scalability, and vulnerability to proxy attendance. To overcome these challenges, this paper proposes Smart Vision, a Convolutional Neural Network (CNN)–based automated attendance system that employs computer vision and deep learning for real-time face detection and recognition. The proposed system captures facial images with a standard camera and processes them with a trained CNN to extract distinctive facial features for accurate identification. Recognized faces are matched against a stored database, and attendance is automatically recorded along with date and time, ensuring minimal human intervention. The system is designed to operate effectively under varying illumination conditions and facial orientations, enhancing robustness and reliability. By eliminating physical contact and manual processes, the proposed solution improves operational efficiency while preventing impersonation and data inconsistencies. Experimental evaluation demonstrates that the system achieves high recognition accuracy with reduced processing time when compared to conventional attendance methods. The architecture is scalable, cost-effective, and adaptable to different environments such as classrooms, offices, and large-scale organizations. The proposed Smart Vision system highlights the practical application of CNN-based facial recognition in intelligent automation and provides a secure, contactless, and efficient solution for modern attendance management systems.