In this project we designed an embedded system using a RaspberryPi to create a facial recognition door lock. The way this system work is it uses a keypad code (there is one for each designated user) to identify the the set users. With the basis of the passcode, we then used to a facial recognition model to recognise if the right user is opening the lock. If these parameters are passed correctly, the lock opens.
The model was made using convolutional neural networks (CNN) with VGG16 in the Keras machine learning library.