An Enhanced Optical-Flow Based Attendance Tracking System for Campus Environment

Manual attendance marking consumes significant time during classes, as it involves calling out students' names individually and recording their presence either physically or digitally. Although efforts to automate this process have been made, such as using biometric data like fingerprints, these methods are still time-consuming due to the need for students to queue up for scanning. A computer vision-based attendance system offers a more efficient and seamless alternative by leveraging advanced face recognition algorithms to identify students or employees automatically. This approach uses convolutional neural networks (CNNs) to extract facial features from images and match them with pre-registered data. A pipeline combining the Multi-task Cascaded Convolutional Networks (MTCNN) for face detection and alignment with the FaceNet model for recognition and classification is employed to ensure high accuracy. The system marks attendance in the backend, eliminating the need for manual intervention. Additionally, the model's performance is compared with traditional recognition and classification algorithms to demonstrate its superior efficiency and reliability, making it an ideal solution for automating attendance systems.