This dataset contains around 2,000 images of skateboard decks.
The default image size is 128x128, but you can also specify different size during training, e.g with a relativitic GAN:
mantra train relativistic_gan --dataset decks --image-dim 256 256
In the data processing logic, vertical and horizontal flipping have been applied as data augmentation.
To import this dataset to your project, run:
mantra import RJT1990/data/decks
Example usage in code:
from data.decks.data import SkateboardDecks
d = SkateboardDecks()
print(d.X.__class__) ## <class 'numpy.ndarray'>
print(d.X.shape) ## (5680, 128, 128, 3)