This paper presents an automated attendance mark- ing system specifically designed for a classroom environment utilizing deep learning techniques. The system employs a collection of Convolutional Neural Networks (CNN), ResNet, and Histogram of Oriented Gradients (HOG) for accurate face detection and recognition. By implementing a webcam-based face detection with AI capabilities, the system automatically identifies students and marks attendance, storing data in a secure Excel- based database accessible through different user interfaces for administrators and students. The implementation demonstrates significant advantages over traditional methods, particularly in handling varying light conditions and partial face obstruction, while maintaining high accuracy in attendance records. This solution includes a web-based interface for fast access and management of attendance data in a structured format.