Make a system that allows the facial recognition model to be self-learning over time

[AkshayVemulapalli]

At the early stages of the model, it might not perform very well because of limited access to training data (labeled pictures of student's faces).

We should make a system for students to notify the model if they are accidentally misclassified, and select their correct name. This will allow the model to continue learning from its mistakes over time and improve its performance.

It can also gather training data over time by taking more pictures of people's faces every time they are marked present, until it has enough pictures to consistently perform well.

[VJCS185]

This can be done by taking pictures of the students as they are scanned by the system. These pictures will keep being added to the database and the software can keep training off the new images every day. This allows the system to adapt to the changing look of each student as well as be more accurate.