I have trained an object detection on custom vision Microsoft in order to use the trained model I should have the python library Microsoft Custom Vision Image Processing Lib