:py:mod:`anomalib.models.padim.anomaly_map`
===========================================

.. py:module:: anomalib.models.padim.anomaly_map

.. autoapi-nested-parse::

   Anomaly Map Generator for the PaDiM model implementation.



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

Classes
~~~~~~~

.. autoapisummary::

   anomalib.models.padim.anomaly_map.AnomalyMapGenerator




.. py:class:: AnomalyMapGenerator(image_size: Union[omegaconf.ListConfig, Tuple], sigma: int = 4)

   Generate Anomaly Heatmap.

   :param image_size: Size of the input image. The anomaly map is upsampled to this dimension.
   :type image_size: Union[ListConfig, Tuple]
   :param sigma: Standard deviation for Gaussian Kernel. Defaults to 4.
   :type sigma: int, optional

   .. py:method:: compute_distance(embedding: torch.Tensor, stats: List[torch.Tensor]) -> torch.Tensor
      :staticmethod:

      Compute anomaly score to the patch in position(i,j) of a test image.

      Ref: Equation (2), Section III-C of the paper.

      :param embedding: Embedding Vector
      :type embedding: Tensor
      :param stats: Mean and Covariance Matrix of the multivariate Gaussian distribution
      :type stats: List[Tensor]

      :returns: Anomaly score of a test image via mahalanobis distance.


   .. py:method:: up_sample(self, distance: torch.Tensor) -> torch.Tensor

      Up sample anomaly score to match the input image size.

      :param distance: Anomaly score computed via the mahalanobis distance.
      :type distance: Tensor

      :returns: Resized distance matrix matching the input image size


   .. py:method:: smooth_anomaly_map(self, anomaly_map: torch.Tensor) -> torch.Tensor

      Apply gaussian smoothing to the anomaly map.

      :param anomaly_map: Anomaly score for the test image(s).
      :type anomaly_map: Tensor

      :returns: Filtered anomaly scores


   .. py:method:: compute_anomaly_map(self, embedding: torch.Tensor, mean: torch.Tensor, inv_covariance: torch.Tensor) -> torch.Tensor

      Compute anomaly score.

      Scores are calculated based on embedding vector, mean and inv_covariance of the multivariate gaussian
      distribution.

      :param embedding: Embedding vector extracted from the test set.
      :type embedding: Tensor
      :param mean: Mean of the multivariate gaussian distribution
      :type mean: Tensor
      :param inv_covariance: Inverse Covariance matrix of the multivariate gaussian distribution.
      :type inv_covariance: Tensor

      :returns: Output anomaly score.


   .. py:method:: __call__(self, **kwds)

      Returns anomaly_map.

      Expects `embedding`, `mean` and `covariance` keywords to be passed explicitly.

      Example:
      >>> anomaly_map_generator = AnomalyMapGenerator(image_size=input_size)
      >>> output = anomaly_map_generator(embedding=embedding, mean=mean, covariance=covariance)

      :raises ValueError: `embedding`. `mean` or `covariance` keys are not found

      :returns: anomaly map
      :rtype: torch.Tensor