minerva.transforms.moco_augmentations
=====================================

.. py:module:: minerva.transforms.moco_augmentations


Classes
-------

.. autoapisummary::

   minerva.transforms.moco_augmentations.MoCoRandomAugmentations


Module Contents
---------------

.. py:class:: MoCoRandomAugmentations

   Default MoCo augmentations.

   Based on the paper's technical details:
   - 224×224-pixel crop from a randomly resized image
   - Random color jittering
   - Random horizontal flip
   - Random grayscale conversion

   Initialize MoCo's default augmentation pipeline.


   .. py:method:: __call__(image)

      Apply MoCo augmentations to a PIL Image.

      Parameters
      ----------
      image : PIL.Image.Image
          Input PIL image

      Returns
      -------
      Tuple[torch.Tensor, None]
          Augmented tensor and None (for compatibility with other augmentation classes)



   .. py:attribute:: augment