Once a Model has been created and trained, it can be used.

This involves presenting the model with an image, and receiving back its assessment of which category it is most likely to be.

This process is Image Classification.

To classify an image, add a ML - Classify Image event action, typically to a Record Change event of a Table containing the image to be classified. This action can only be executed in the Background.

On the Configuration tab:

Source

Image

Select the image field in the current table that will be classified.

Model

Model

Select the ModelID that will be used to perform the classification.

Destination

Prediction

Select a text  field that will store the name of the category that the model thinks the image is most likely to be.

Probability

Select a number field that will store the probability that the image is in the specified category. This is a number between 0 and 1. The higher the number the greater the confidence.

Diagnostics

This section gives the standard event action diagnostic feedback.

Completed OK

Store in a Checkbox field

Feedback

Store in a Text field