This project implements an automated face-based attendance system using MTCNN for face detection and alignment, face_recognition for embedding extraction, and cosine-similarity matching to identify students. The system marks attendance in an Excel sheet and includes evaluation metrics such as confusion matrix, precision-recall, and ROC curves. The approach uses pretrained models (transfer learning) and performs supervised face matching without training a custom neural network.
Group project contributors: Aryan Pal, Dikshith Poojary and Tejas Rana.