Im doing a project for my university where we have to make an AI Based Attendance System using raspberry pi 5. The main ambiguity i am having is that we want our raspberry pi to detect faces from its camera, compare them to the database stored in the cloud. If the match is met, the student should get a notification on an app that their attendance has been marked, and teachers can check in the end of the class how many students are present. the cloud we want to use to store our database is firebase. Thus i wanted to ask if their is any way to link firebase with the facial recognition model on pi 5, so that it could compare the faces it is capturing from the pictures placed in the database on firebase
1 Like
Hey there, @Hanzalah310571, and welcome to the forum. Glad to have you here.
I must admit that I’ve never worked with Firebase before, but this seems like it would be perfectly doable. There’s a good guide on accessing Firebase with python here:
https://medium.com/full-stack-engineer/the-simplest-firebase-api-6521d4307cb7
Based on the guide, however, you would need to modify the file model_training.py so that instead of going through the local files in dataset, it would then check the corresponding names and pictures in the Firebase API. That would probably involve a fair bit of restructuring the code.
It’s also not terribly efficient since it means that everytime you set up the code again you would need to remunerate. It feels like it would be a lot easier to have the faces and names on the pi itself.
