If I have a directory structure like this:
|-data
|-train
|-Bellis perennis
| |-lawn_daisy
| | |-image1.jpg
| |-english_daisy
| |-image1.jpg
|-Tulipa dasystemon
| |-some tulip name
| |-image.jpg
|-etc
and I use Keras flow_from_directory, it will only grab the first tree which is Bellis perennis and Tuplia and use those as labels. Is there a way to use the main directory (Bellis) as one label and its sub_directory(lawn daisy, english_daisy) as a second label and use those labels with the Keras functional API?
Another option is to just change the way the folders are setup, but if there is a way to work with this folder structure I'd like to learn about it.
Edit - After further reading, it seems that you can specify the classes explicitly under train_datagen.flow_from_directory. But can you specify 2 different classes?
