How add real and spoof data in dataset

[kundanSingh11]

As per the document we need to call this command to prepare data

python scripts/prepare_data.py
--orig_dir
--crop_dir
--size
--bbox_expansion_factor
--spoof_types [ ...]

do we need to do this for all --spoof_types ?

or how to give --spoof_types for every image or a dataset and do we need to run this command for real and every spoof attack ?

[kundanSingh11]

is the prepare_data based on CelebA file stracture ?

[kundanSingh11]

At least let me know do we need only real face for training??