Marking attendance is a customary ritual which involved either making entries on a register or punching in using a fingerprint sensor, in the pre-COVID world. Now we must rely on a non-physical medium to mark the attendance for the employees.

Face recognition technology can be used to mark the attendance of employees. In this project, I built a face recognition module (I am not at the liberty to share the architecture of this face recognition module) which can identify people and can be trained with as little as two pictures. This project is up and running at the organisation I currently work in.

Key Features:
1.Integrated Mask detection module to check whether the person is wearing a mask or not, and raises alerts accordingly.
2.Can recognize people even with partial occlusion, i.e., if they have their nose and face covered. Even though the model churns out low confidence predictions, it was tested and the model was quite accurate.
3.automatically "remembers" new faces and labels them as guests until manually specified later.
4.A thermal camera is going to be integrated into this module and the temperature of the employees will be taken automatically.

The best feature about this project is that it can be scaled up to work in bigger organisations, and even in towns.