I'm getting my training data using...
turicreate.image_analysis.load_images('TrainingData', with_path=True)
TrainingData/ has all my training data using the folder names as labels...
data['label'] = data['path'].apply(lambda path: os.path.dirname(path.split('/')[-1]))
That gets the last directory... the problem is that turicreate is going deeper than I'm expecting it to.
The directory structure looks like this...
- TrainingData/
- Model1/
- ADirectory/ // <== here's the problem (I want to ignore this directory)
- img1_s.jpg
- img1.jpg
- img2.jpg
- ADirectory/ // <== here's the problem (I want to ignore this directory)
- Model2/
- Model1/
- runTuriCreate.py
At the moment turicreate is going into TrainingData which is what I'd expect... but it's using all subdirectories as well as the directories I want to model.
Is there a way to do this?
EDIT: I found that you can add recursive=False to the load_images function call. The problem with this is that I AM using recursion... just recursion with one depth level strictly.