Can't set source as Camera for Detection.py

[shivenPython2023]

I know you do this in the terminal through the "--input rpi" command, but is there a way to do this in the python file itself? Can you write python code in the detection.py file so that it takes frames from the pi camera and uses these for detection instead? I need to this in the file itself becuase I need to run detection from 2 different cameras simultaneously. I am using the module 3 camera.