Trainable Weka Segmentation - “Cannot train without at least 2 sets of examples!”

Hi all,
I’m totally new both to this forum and to Trainable Weka Segmentation plugin; as far as I could see, my question hasn’t been answered previously (although this error was mentioned in one post); if it has, I apologize.

So, I’m trying to get to know the plugin by putting in a time lapse image stack of a few cells overexpressing RFP or FRET construct. I obviously want to segment the images into individual cells for further analysis (they change morphology over time). Yet, I can’t even start, I just keep getting the message “Cannot train without at least 2 sets of examples!”, and accordingly, “The training did not finish”.
It must be something very simple that I’m doing wrong?
I apologize again if this has been discussed before.
Many thanks in advance to anyone who can help!

According to the source code, this means that you don’t have training instances for at least two classes. To train a classifier, you need labelled examples for all the classes you want to distinguish. In the case of segmentation, the default classification is binary at the pixel level: a pixel is part of the ROI or not. So you need to provide two sets of annotated pixels as training data. You can find more info in the documentation.

Thank you so much!
Best regards,
Elena