pyiqa.archs.clipiqa_arch
========================

.. py:module:: pyiqa.archs.clipiqa_arch

.. autoapi-nested-parse::

   CLIP-IQA metric, proposed by

   Exploring CLIP for Assessing the Look and Feel of Images.
   Jianyi Wang, Kelvin C.K. Chan, Chen Change Loy.
   AAAI 2023.

   Ref url: https://github.com/IceClear/CLIP-IQA
   Re-implemented by: Chaofeng Chen (https://github.com/chaofengc) with the following modification:
       - We assemble multiple prompts to improve the results of clipiqa model.







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

.. py:data:: default_model_urls

.. py:class:: PromptLearner(clip_model, n_ctx=16)

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


   PromptLearner class for learning prompts for CLIP-IQA.

   Disclaimer:
       This implementation follows exactly the official codes in: https://github.com/IceClear/CLIP-IQA.
       We have no idea why some tricks are implemented like this, which include:
           1. Using n_ctx prefix characters "X"
           2. Appending extra "." at the end
           3. Insert the original text embedding at the middle


   .. py:method:: get_prompts_with_middle_class()

      Get prompts with the original text embedding inserted in the middle.

      :returns: The generated prompts.
      :rtype: torch.Tensor



   .. py:method:: forward(clip_model)

      Forward pass for the PromptLearner.

      :param clip_model: The CLIP model.
      :type clip_model: nn.Module

      :returns: The output features.
      :rtype: torch.Tensor



.. py:class:: CLIPIQA(model_type='clipiqa', backbone='RN50', pretrained=True, pos_embedding=False)

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


   CLIPIQA metric class.

   :param model_type: The type of the model. Default is 'clipiqa'.
   :type model_type: str
   :param backbone: The backbone model. Default is 'RN50'.
   :type backbone: str
   :param pretrained: Whether to load pretrained weights. Default is True.
   :type pretrained: bool
   :param pos_embedding: Whether to use positional embedding. Default is False.
   :type pos_embedding: bool


   .. py:method:: forward(x)

      Forward pass for the CLIPIQA model.

      :param x: Input tensor with shape (N, C, H, W).
      :type x: torch.Tensor

      :returns: The output probabilities.
      :rtype: torch.Tensor