Skip to content
heliaEDGE
Reference
HELIA

random_crop

Random crops sharing sampled offsets across spatially aligned leaves.

Machine-readable model

  • RandomCrop1DclassTraining-only crop; inference leaves the original duration unchanged.
  • RandomCrop2DclassTraining-only image crop; inference leaves the original extent unchanged.
class

RandomCrop1D

Python

Training-only crop; inference leaves the original duration unchanged.

helia_edge/layers/preprocessing/random_crop.py:18

RandomCrop1D(duration: int, unique_batch: bool = False, **kwargs={})

Training-only crop; inference leaves the original duration unchanged.

Base class: BaseAugmentation1D.

Inherited from BaseAugmentation: augment_masks(), augment_sample(), augment_targets(), batch_augment(), call().

class

RandomCrop2D

Python

Training-only image crop; inference leaves the original extent unchanged.

helia_edge/layers/preprocessing/random_crop.py:46

RandomCrop2D(height: int, width: int, unique_batch: bool = False, **kwargs={})

Training-only image crop; inference leaves the original extent unchanged.

Base class: BaseAugmentation2D.

Inherited from BaseAugmentation: augment_masks(), augment_sample(), augment_targets(), batch_augment(), call().