Convolutional Neural Network based Automated Attendance System by using Facial Recognition Domain
Shailender Kumar, Dhruv Kathpalia, Dipen Singh, Mandeep Vats · 2020
This project aims to recognize faces in an image, video, or via live camera using a deep learning-based Convolutional Neural Network model that is fast as well as accurate. Face recognition is a process of identifying faces in an image and has practical applications in a variety of domains, including information security, biometrics, access control, law enforcement, smart cards, and surveillance system. Deep Learning uses numerous layers to discover interpretations of data at different extraction levels. It has improved the landscape for performing research in facial recognition. The state-of-the-art implementation has been bettered by the introduction of deep learning in face recognition and has stimulated success in practical applications. Convolutional neural networks, a kind of deep neural network model has been proven to achieve success in the face recognition domain. For real-time systems, sampling must be done before using CNNs. On the other hand, complete images (all the pixel values) are passed as the input to Convolutional Neural Networks. The following steps: feature selection, feature extracti on, and training are performed in each step. This might lead to the assumption, where convolutional neural network implementation has a chance to get complicated and time-consuming.