How to apply augmentation to masks or labels

[gtgrp-user]

Hi,

I have successfully applied the following transforms to my dataloader for the training images. However, I'm unsure how to apply the random horizontal and vertical flips to the masks/labels.

transforms = AugmentationSequential(
    MinMaxNormalize(mins, maxs),
    indices.AppendNDWI(index_green=1, index_nir=3),
    indices.AppendNDVI(index_nir=3, index_red=2),
    K.RandomHorizontalFlip(p=0.5),
    K.RandomVerticalFlip(p=0.5),
    data_keys=["image"],
)

[isaaccorley]

If you modify data_keys to data_keys=["image", "mask"] this will apply the correct transforms to both the image and mask. AugmentationSequential is simply a custom wrapper around kornia.augmentation.AugmentationSequential if you're curious about further documentation.