cristinel.ababei, april.2024,2026

This is a modified version of the "person detection" example from 
the latest version (2024-25) of the Arduino_TensorFlowLite library.
The original version of this example is inside:
  person_detection_Arduino_TensorFlowLite/,
but as of April.15.2024 does not have support for capturing data 
from the camera OV7675 available on the Arduino Nano 33 BLE Sense.

The modifications I made include:

1)
Essentially bringing the support for the OV7675 camera from the older 
version (2019) of this example from the library Harvard_TinyMLx, which 
I uninstalled because it was not updated for a long time and it was bundled 
with an older version of TensorFlow - making recent g_model.cc models generated with
more recent TensorFlow (in Google Colab) damage Arduino boards.
The older version is inside:
  person_detection_Harvard_TinyMLx/

2)
Copying camera related files from Harvard_TinyMLx->TO->Arduino_TensorFlowLite
The files needed are backed-up here: src/additional_OV7675_files/

But, to make the exaple compile you need to copy them from:
  C:\Users\Cristinel Ababei\Documents\Arduino\libraries\Harvard_TinyMLx\src
To:
  C:\Users\Cristinel Ababei\Documents\Arduino\libraries\Arduino_TensorFlowLite\src\additional_OV7675_files
Or, just place them locally inside the scketch's itself folder into a new folder called "src".
Either way, it should work...


So, long story short: this is a combination of code from 
"person detection" example from more recent Arduino_TensorFlowLite library
with code from "person detection" example from the older Harvard_TinyMLx library.