Hello - I would like to use the library for offline augmentations (meaning to augment the dataset prior to training rather than during training) - is there a way to use the library without using the batch-generation option? just to pass in images and apply a list of augmentations on both images and their labels and output the n-augmented images/labels?
Another question is, current implementation of the augmentation method assumes a single segmentation? Is there any way to generate augmentation for a pair of image and a list of labels?
Hello - I would like to use the library for offline augmentations (meaning to augment the dataset prior to training rather than during training) - is there a way to use the library without using the batch-generation option? just to pass in images and apply a list of augmentations on both images and their labels and output the n-augmented images/labels?
Another question is, current implementation of the augmentation method assumes a single segmentation? Is there any way to generate augmentation for a pair of image and a list of labels?