vision_unlearning.metrics.image
===============================

.. py:module:: vision_unlearning.metrics.image


Classes
-------

.. autoapisummary::

   vision_unlearning.metrics.image.MetricImage
   vision_unlearning.metrics.image.MetricPaintingStyle
   vision_unlearning.metrics.image.MetricRace
   vision_unlearning.metrics.image.MetricGender
   vision_unlearning.metrics.image.MetricQuality


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

.. py:class:: MetricImage(/, **data: Any)

   Bases: :py:obj:`vision_unlearning.metrics.base.Metric`, :py:obj:`abc.ABC`


   Based only on the image itself
   e.g., image quality, painting style


   .. py:method:: score(image: PIL.Image.Image) -> Dict[str, Any]
      :abstractmethod:



.. py:class:: MetricPaintingStyle(/, **data: Any)

   Bases: :py:obj:`MetricImage`


   Based only on the image itself
   e.g., image quality, painting style


   .. py:attribute:: metrics
      :type:  List[Literal['is_desired_style', 'desired_style_confidence']]
      :value: []



   .. py:attribute:: desired_style
      :type:  str


   .. py:attribute:: top_k
      :type:  int
      :value: 5



   .. py:attribute:: model_path
      :type:  str


   .. py:attribute:: device
      :type:  Optional[torch.device | str | int]
      :value: 'cuda'



   .. py:attribute:: _pipeline
      :type:  Optional[transformers.pipelines.image_classification.ImageClassificationPipeline]
      :value: None



   .. py:method:: model_post_init(__context: Optional[dict] = None) -> None

      Override this method to perform additional initialization after `__init__` and `model_construct`.
      This is useful if you want to do some validation that requires the entire model to be initialized.



   .. py:method:: score(image: PIL.Image.Image) -> Dict[str, bool | float]


.. py:class:: MetricRace(/, **data: Any)

   Bases: :py:obj:`MetricImage`


   Race classification using Hugging Face model:
   syntheticbot/clip-face-attribute-classifier

   Requires the following additional dependencies:
   * tf_keras = "~2.19.0"
   * tensorrt = "~10.13.2"
   * blinker = "~1.9.0"


   .. py:method:: model_post_init(__context: Optional[dict] = None) -> None

      Override this method to perform additional initialization after `__init__` and `model_construct`.
      This is useful if you want to do some validation that requires the entire model to be initialized.



   .. py:method:: score(image: PIL.Image.Image) -> Dict[str, str]


.. py:class:: MetricGender(/, **data: Any)

   Bases: :py:obj:`MetricImage`


   Based only on the image itself
   e.g., image quality, painting style


   .. py:attribute:: device
      :type:  Optional[torch.device | str | int]
      :value: 'cpu'



   .. py:attribute:: _model_name
      :type:  str
      :value: 'prithivMLmods/Realistic-Gender-Classification'



   .. py:attribute:: _id2label


   .. py:attribute:: _model
      :type:  Any


   .. py:attribute:: _processor
      :type:  Any


   .. py:method:: model_post_init(__context)

      Override this method to perform additional initialization after `__init__` and `model_construct`.
      This is useful if you want to do some validation that requires the entire model to be initialized.



   .. py:method:: score(image: PIL.Image.Image) -> Dict[str, Union[Literal['male', 'female'], float]]


.. py:class:: MetricQuality(/, **data: Any)

   Bases: :py:obj:`MetricImage`


   Based only on the image itself
   e.g., image quality, painting style


   .. py:method:: _load_image(img: Union[PIL.Image.Image, numpy.ndarray, str]) -> torch.Tensor


   .. py:method:: score(image: Union[PIL.Image.Image, numpy.ndarray, str]) -> Dict[str, float]


   .. py:method:: score_batch(images: List[Union[PIL.Image.Image, numpy.ndarray, str]]) -> List[Dict[str, float]]