Is micropython and tensorflow lite implementation on Portenta limited to images?

Arduino Boards
1

Hi,

I am looking to use the Portenta more centrally for processing decisions using a Tensorflow Neural net.

I’ve successfully converted the model to the lite version however I cannot find commands or examples of how to execute it when it is not passing images to the model, only strings.

is there any examples or sources to documentation for using models and such above what is currently documented for the OpenMV Arduino micropython instance?

tf — Tensor Flow — MicroPython 1.15 documentation (openmv.io)

Regards,
Roy

2

We only support images for our library. You can make an image a 1d array of 8-bit values and then we’ll support that too.