Dear Ilastik users and devs,
I recently start looking into using machine learning for segmentation and quite amazed with the ease and efficiency of the Ilastik!
However I can’t find the option to save the trained model for later use. Is it possible? Is there a way to use it then in other environments (e.g. python).
Cheers
Hello @Anton_Kamnev,
the classifier is saved inside the project file, which itself is just a hdf5 file (with a different ending, ilp). So you can open it with all the hdf5 libraries available. E.g. if your project file was myproject.ilp then you could find the trained classifier at myproject.ilp/PixelClassification/ClassifierForests. However, in order to apply this classifier to new data, you’d have to calculate the features for each image, just as it is done in ilastik, load the classifier via the appropriate library and construct the pipeline.
If you just want to apply your trained classifier to new data, you can do so by using the headless mode of ilastik.
Thanks for the detailed answer Dominik!