Face Recognition Attendance System for Online Classes
Bhargavi Poyekar, Rishita Mote, Jahnvi Shah, Surekha Dholay · 2022 13th International Conference on Computing Communication and Networking Technologies (ICCCNT) · 2022
Due to the increasing need for online lectures due to situations like COVID-19 and various online learning platforms, there is a need for a reliable attendance system for online classes. Our system is developed for deploying an easy and secure way of taking attendance without tedious roll-calls and inaccurate participant lists. The teachers will take screenshots of students in the online meets with their video cameras on. This screenshot will be uploaded to our system which will recognize the students and generate a report of their attendance. Our system will allow students and faculty to view the attendance of each lecture and give feedback if any discrepancy is found. We collected the dataset of face images of 94 students which are then augmented to increase the dataset and then used HOG for face detection. We then applied four algorithms namely VGG-16, MobileNet, InceptionV3, and our own CNN model for face recognition. MobileNet gave us the highest accuracy of 97.14%. We, therefore, deployed this model for our website to recognize faces and generate the attendance report.