I just started working with machine learning in python (multiclass image classifier model in keras/tensorflow converting it to coreml).
Is there anyone who tried it in xojo or maybe someone who managed to use coreml in xojo ios?
Well, for macOS, I do have a few plugin classes:
Works well here to classify images.
(The plugin would work on iOS, if we would have a plugin SDK there)
I’ve tried macOS (MBS plugin) and I’m impressed! Now working on coming from the RapidMiner to xojo.
I use this code to (local) compile from apple MLmodel to MLmodelc. I think that “MLmodelc” can be used in iOS xojo on the same machine, but haven’t tried that yet.
dim File_to_compile as FolderItem = GetOpenFolderItem(".MLmodel")
dim Error as NSErrorMBS
dim File_to_folder As FolderItem = MLModelMBS.compileModelFile(File_to_compile, Error)
if File_to_folder <> nil then
File_to_folder.MoveFileTo GetSaveFolderItem ( “”,“MLmodelc” )
if Error<>Nil then
I dont know what mlmodelc is but Im using .mlmodel files in swift 4.
I train a pretrained model (imagenet or inception) for my own classes. Convert it to coreml and saving it as mlmodel. I was looking for a method to do these steps in xojo for students and coworkers.
For the use/introduktion of mlmodelc have a look at;