This is code to evaulate if a face is really a face or not on a raspberry pi interface by using a pretrained model researched by Becca. The model is included in the picameraeval.py file.Make sure your tensorflow version is 1.x.
Items needed:
Raspberry Pi (3b was used here)
Camera that works with raspberry pi
Set-up: Make sure to allow your raspberry pi to be used - here is a good guide; https://thepihut.com/blogs/raspberry-pi-tutorials/16021420-how-to-install-use-the-raspberry-pi-camera
Install:
python3 -m pip install tensorflow
python3 -m pip install numpy
python3 -m pip install pandas
To run:
Open the file called run.py any pre-installed python IDEs on raspberry pi such as Thorny or Geanie, and run!
You can also try python3 picameraeval.py on the terminal.
The results will be in the first directory called "results_withpredcnn2.csv" (maybe can be improved in the future)
"classes:" 1 - face
"classes:" 0 - not a face
Check out "demo" folder for smaller python files I used to create this!