:py:mod:`anomalib.models.components.layers.sspcab`
==================================================

.. py:module:: anomalib.models.components.layers.sspcab

.. autoapi-nested-parse::

   SSPCAB: Self-Supervised Predictive Convolutional Attention Block for reconstruction-based models.

   Paper https://arxiv.org/abs/2111.09099



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

Classes
~~~~~~~

.. autoapisummary::

   anomalib.models.components.layers.sspcab.AttentionModule
   anomalib.models.components.layers.sspcab.SSPCAB




.. py:class:: AttentionModule(in_channels: int, reduction_ratio: int = 8)

   Bases: :py:obj:`torch.nn.Module`

   Squeeze and excitation block that acts as the attention module in SSPCAB.

   :param channels: Number of input channels.
   :type channels: int
   :param reduction_ratio: Reduction ratio of the attention module.
   :type reduction_ratio: int

   .. py:method:: forward(inputs: torch.Tensor) -> torch.Tensor

      Forward pass through the attention module.



.. py:class:: SSPCAB(in_channels: int, kernel_size: int = 1, dilation: int = 1, reduction_ratio: int = 8)

   Bases: :py:obj:`torch.nn.Module`

   SSPCAB block.

   :param in_channels: Number of input channels.
   :type in_channels: int
   :param kernel_size: Size of the receptive fields of the masked convolution kernel.
   :type kernel_size: int
   :param dilation: Dilation factor of the masked convolution kernel.
   :type dilation: int
   :param reduction_ratio: Reduction ratio of the attention module.
   :type reduction_ratio: int

   .. py:method:: forward(inputs: torch.Tensor) -> torch.Tensor

      Forward pass through the SSPCAB block.