Implementation of rgb2ycbcr function in Matlab under PyTorch References from：`https://en.wikipedia.org/wiki/YCbCr#ITU-R_BT.601_conversion` A

Convert the image data type to the Tensor (NCWH) data type supported by PyTorch Args: image (np.ndarray): The image data read by ``OpenCV

Load pre-trained model weights Args: model (nn.Module): model compile_state (bool): model compilation state, `False` means not co

Check if the dimensions of the two tensors are the same Args: raw_tensor (np.ndarray or Tensor): tensor flow of images to be compared, RG

Restore training model weights Args: model (nn.Module): model ema_model (nn.Module): EMA model optimizer (nn.optim): opti

Load model weights and parameters Args: model (nn.Module): model compile_mode (bool): Enable model compilation mode, `False` mean

Args: crop_border (int, optional): how many pixels to crop border. Default: 0 only_test_y_channel (bool, optional):

Implementation of `calculate_weights_indices` function in Matlab under Python language. Args: in_length (int): Input length. out_

Implementation of `cubic` function in Matlab under Python language. Args: x: Element vector. Returns: Bicubic interpolation

PyTorch implements the MSE (Mean Squared Error, mean square error) function Args: raw_tensor (Tensor): tensor flow of images to be compar

Reshape the tensor to the shape of (B, C, K, H, W) or (B, C, H, K, W) for 1D convolution. Args: tensor (Tensor): Tensor to be reshaped.

Resize the given tensor to the given size. Args: tensor (Tensor): Tensor to be resized. dim (int): Dimension to be resized.

Randomly intercept two images in the specified area Args: gt_images (ndarray | Tensor | list[ndarray] | list[Tensor]): ground truth image

Randomly flip the image up and down Args: gt_images (ndarray): ground truth images read by the PyTorch library lr_images (ndarray

Randomly rotate the image Args: gt_images (ndarray | Tensor | list[ndarray] | list[Tensor]): ground truth images read by the PyTorch libr

Randomly flip the image left and right Args: gt_images (ndarray): ground truth images read by the PyTorch library lr_images (ndar

Convert the Tensor(NCWH) data type supported by PyTorch to the np.ndarray(WHC) image data type Args: tensor (Tensor): Data types supporte

Casts the input tensor to the correct data type and stores the original data type. Args: tensor (Tensor): Input tensor. Returns:

PyTorch implements the fspecial_gaussian() function in MATLAB Args: window_size (int): Gaussian filter size sigma (float): sigma

Get padding size and padded tensor Args: tensor (Tensor): shape (b, c, h, w) kernel_size (int): kernel size x_size (int):

Get weight for each pixel Args: tensor (Tensor): shape (b, c, h, w) kernel_size (int): kernel size kernel (str): kernel t

PyTorch implements PSNR (Peak Signal-to-Noise Ratio, peak signal-to-noise ratio) function Args: raw_tensor (Tensor): tensor flow of image

Reflect padding for 2-dim tensor Args: tensor (Tensor): shape (b, c, h, w) dim (int): 2 or -2 pad_pre (int): padding size

Reshape the input tensor to 4-dim tensor Args: tensor (Tensor): shape (b, c, h, w) or (c, h, w) or (h, w) Returns: tensor (T

PyTorch implements SSIM (Structural Similarity) function Args: raw_tensor (Tensor): tensor flow of images to be compared, RGB format, dat

Convert the input to a tuple Args: dim (int): the dimension of the input

Implementation of bgr2ycbcr function in Matlab under Python language. Args: image (np.ndarray): Image input in BGR format only_us

Implementation of `imresize` function in Matlab under Python language. Args: image: The input image. scale_factor (float): Scale

Split the image into multiple small images. Args: args (dict): Custom parameter dictionary.

Args: crop_border (int, optional): how many pixels to crop border. Default: 0 only_test_y_channel (bool, optional):

Args: window_size (int): Gaussian filter size, must be an odd number, default: ``11`` gaussian_sigma (float): sigma

Args: gt_images_dir (str): Ground-truth image address. lr_images_dir (str, optional): Low resolution image address.

Args: paired_gt_images_dir: The address of the ground-truth image after registration paired_lr_images_dir: The addre

Resize image with given kernel and sigma. Args: x (Tensor): Input image with shape (b, c, h, w) scale_factor (float): Scale facto

Implementation of bgr2ycbcr function in Matlab under PyTorch References from：`https://en.wikipedia.org/wiki/YCbCr#ITU-R_BT.601_conversion` A

Crop small image patches from one image center area. Args: image (np.ndarray): The input image for `OpenCV.imread`. image_size (i