a  dX@s2ddlmZmZmZddlZddlmZddlmZddl m Z ddl m Z e eejeejefeddd Ze ee eeejeejefed d d Zee eeefd ddZdee e e edddZdee e edddZee ed ddZe ee eeejejed ddZee e e edddZdS))SequenceTupleUnionN)Tensor) functional)rank_zero_warn)_TORCH_GREATER_EQUAL_1_10) kernel_sizesigmadtypedevicereturncCsTtjd|dd|dd||d}tt||d d}||jddS)aYCompute 1D gaussian kernel. Args: kernel_size: size of the gaussian kernel sigma: Standard deviation of the gaussian kernel dtype: data type of the output tensor device: device of the output tensor Example: >>> _gaussian(3, 1, torch.float, 'cpu') tensor([[0.2741, 0.4519, 0.2741]]) )startendstepr r rdim)torcharangeexppowsum unsqueeze)r r r r distgaussrm/var/www/html/stable-diffusion-webui/venv/lib/python3.9/site-packages/torchmetrics/functional/image/helper.py _gaussian s$r)channelr r r r r cCsTt|d|d||}t|d|d||}t||}||d|d|dS)aCompute 2D gaussian kernel. Args: channel: number of channels in the image kernel_size: size of the gaussian kernel as a tuple (h, w) sigma: Standard deviation of the gaussian kernel dtype: data type of the output tensor device: device of the output tensor Example: >>> _gaussian_kernel_2d(1, (5,5), (1,1), torch.float, "cpu") tensor([[[[0.0030, 0.0133, 0.0219, 0.0133, 0.0030], [0.0133, 0.0596, 0.0983, 0.0596, 0.0133], [0.0219, 0.0983, 0.1621, 0.0983, 0.0219], [0.0133, 0.0596, 0.0983, 0.0596, 0.0133], [0.0030, 0.0133, 0.0219, 0.0133, 0.0030]]]]) rr)rrmatmultexpand)r r r r r gaussian_kernel_xgaussian_kernel_ykernelrrr_gaussian_kernel_2dsr')inputs window_sizer cCsBtjdd|||j|jd}||d}tjd|j|jd}||fS)zConstruct uniform weight and bias for a 2d convolution. Args: inputs: Input image window_size: size of convolutional kernel Return: The weight and bias for 2d convolution r)r r r)rZonesr r zeros)r(r) kernel_weight kernel_biasrrr_uniform_weight_bias_conv2d>s  r-)r(rpad outer_padr c Csl|j|}t||t|ddd|j}t||t|d|||d|j}t|||f|S)a,Apply single-dimension reflection padding to match scipy implementation. Args: inputs: Input image dim: A dimension the image should be padded over pad: Number of pads outer_pad: Number of outer pads Return: Image padded over a single dimension r)shaperZ index_selectrtor cat)r(rr.r/Z_maxxyrrr_single_dimension_padOs $,r6)r(r.r/r cCsdD]}t||||}q|S)zApply reflection padding to the input image. Args: inputs: Input image pad: Number of pads outer_pad: Number of outer pads Return: Padded image )r)r6)r(r.r/rrrr_reflection_pad_2dbs r8csLt|d|dt|\tjfddtjdDddS)zApply uniform filter with a window of a given size over the input image. Args: inputs: Input image window_size: Sliding window used for rmse calculation Return: Image transformed with the uniform input rcs0g|](}tjdd|fdddqS)Nrr)padding)FZconv2dr).0r r(r,r+rr sz#_uniform_filter..rr)r8r-rr3ranger1)r(r)rr<r_uniform_filterss  r?c Cst|d|d||}t|d|d||}t|d|d||}t||}t|ddd|d||d|d|d} | |d|d|d|dS)a6Compute 3D gaussian kernel. Args: channel: number of channels in the image kernel_size: size of the gaussian kernel as a tuple (h, w, d) sigma: Standard deviation of the gaussian kernel dtype: data type of the output tensor device: device of the output tensor rrrr0)rrr!r"mulrrepeatr#) r r r r r r$r%Zgaussian_kernel_zZ kernel_xyr&rrr_gaussian_kernel_3ds rB)r(pad_hpad_wpad_dr cCsZtr"tj|||||||fdd}n4tdt|||gD]\}}t||d|dd}q8|S)aUReflective padding of 3d input. Args: inputs: tensor to pad, should be a 3D tensor of shape ``[N, C, H, W, D]`` pad_w: amount of padding in the height dimension pad_h: amount of padding in the width dimension pad_d: amount of padding in the depth dimension Returns: padded input tensor Zreflect)modezkAn older version of pyTorch is used. For optimal speed, please upgrade to at least PyTorch v1.10 or higher.rr)r/)rr:r.r enumerater6)r(rCrDrErr.rrr_reflection_pad_3ds rH)r)r)typingrrrrrZtorch.nnrr:Ztorchmetrics.utilitiesrZtorchmetrics.utilities.importsrintfloatr r strrr'r-r6r8r?rBrHrrrrs*    "