Use CNN model to classify the users face and combine with Arduino.
First one, we must to create 2 empty folders, named "data" and "model", which will be saved parameter files and used by the codes.
Second, use the codes to follow the next steps:
step01: Run createdb.py to create a database.
step02: Use catch_pic_db.py to catch user face pictures and register into step01 database.
step03: Use showdb.py to make sure the database you have to load in.
step04: Modify the prepare_data.py file, which in load_dataset() function with many labels(e.g. if you have 5 users, you can modify the 5 labels in label_dict{[labels]}).
step05: Run the tf1_train.py, remember to modify parameter about nb_classes, batch_size, and nb_epoch, also can choose data_augmentation true or false.
step06: Remember prepare a Arduino to set the led then create a empty folder named "led_test" and put the led_test.ino in it.
step07: Run the face_predict_gui.py and execute the GUI buttons.
step08: Finally, use showdb.py code to make sure the database to show users have checked in successfully.