o oi@sVddlmZddlZddlmZddlmZ    ddddZGdddejZdS)) annotationsN)nn)metrics?-q=meansameimg1 torch.Tensorimg2 window_sizeintmax_valfloateps reductionstrpaddingreturnc Csnt||||||}tjd|dddd}|dkr!t|}|S|dkr,t|}|S|dkr3 |Std ) aCompute a loss based on the SSIM measurement. The loss, or the Structural dissimilarity (DSSIM) is described as: .. math:: \text{loss}(x, y) = \frac{1 - \text{SSIM}(x, y)}{2} See :meth:`~kornia.losses.ssim` for details about SSIM. Args: img1: the first input image with shape :math:`(B, C, H, W)`. img2: the second input image with shape :math:`(B, C, H, W)`. window_size: the size of the gaussian kernel to smooth the images. max_val: the dynamic range of the images. eps: Small value for numerically stability when dividing. reduction : Specifies the reduction to apply to the output: ``'none'`` | ``'mean'`` | ``'sum'``. ``'none'``: no reduction will be applied, ``'mean'``: the sum of the output will be divided by the number of elements in the output, ``'sum'``: the output will be summed. padding: ``'same'`` | ``'valid'``. Whether to only use the "valid" convolution area to compute SSIM to match the MATLAB implementation of original SSIM paper. Returns: The loss based on the ssim index. Examples: >>> input1 = torch.rand(1, 4, 5, 5) >>> input2 = torch.rand(1, 4, 5, 5) >>> loss = ssim_loss(input1, input2, 5) rr)minmaxrsumnonezInvalid reduction option.)rssimtorchclamprrNotImplementedError) r r r rrrrssim_maplossr!F/home/ubuntu/.local/lib/python3.10/site-packages/kornia/losses/ssim.py ssim_losss*  r#cs0eZdZdZ ddfdd ZdddZZS)SSIMLossaCreate a criterion that computes a loss based on the SSIM measurement. The loss, or the Structural dissimilarity (DSSIM) is described as: .. math:: \text{loss}(x, y) = \frac{1 - \text{SSIM}(x, y)}{2} See :meth:`~kornia.losses.ssim_loss` for details about SSIM. Args: window_size: the size of the gaussian kernel to smooth the images. max_val: the dynamic range of the images. eps: Small value for numerically stability when dividing. reduction : Specifies the reduction to apply to the output: ``'none'`` | ``'mean'`` | ``'sum'``. ``'none'``: no reduction will be applied, ``'mean'``: the sum of the output will be divided by the number of elements in the output, ``'sum'``: the output will be summed. padding: ``'same'`` | ``'valid'``. Whether to only use the "valid" convolution area to compute SSIM to match the MATLAB implementation of original SSIM paper. Returns: The loss based on the ssim index. Examples: >>> input1 = torch.rand(1, 4, 5, 5) >>> input2 = torch.rand(1, 4, 5, 5) >>> criterion = SSIMLoss(5) >>> loss = criterion(input1, input2) rrrrr r rrrrrrrNonecs,t||_||_||_||_||_dSN)super__init__r rrrr)selfr rrrr __class__r!r"r(vs  zSSIMLoss.__init__r r r cCst|||j|j|j|j|jSr&)r#r rrrr)r)r r r!r!r"forwardszSSIMLoss.forwardrrrr) r r rrrrrrrrrr%)r r r r rr )__name__ __module__ __qualname____doc__r(r, __classcell__r!r!r*r"r$Us ! r$r-)r r r r r r rrrrrrrrrr ) __future__rrrkorniarr#Moduler$r!r!r!r"s    ;