PixelShuffle

class torch.nn.PixelShuffle(upscale_factor: int)

Rearranges elements in a tensor of shape $(*, C \times r^2, H, W)$ to a tensor of shape $(*, C, H \times r, W \times r)$ .

This is useful for implementing efficient sub-pixel convolution with a stride of $1/r$ .

Look at the paper: Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network by Shi et. al (2016) for more details.

Parameters

upscale_factor (int) – factor to increase spatial resolution by

Shape:

• Input: $(N, L, H_{in}, W_{in})$ where $L=C \times \text{upscale\_factor}^2$

• Output: $(N, C, H_{out}, W_{out})$ where $H_{out} = H_{in} \times \text{upscale\_factor}$ and $W_{out} = W_{in} \times \text{upscale\_factor}$

Examples:

>>> pixel_shuffle = nn.PixelShuffle(3)
>>> input = torch.randn(1, 9, 4, 4)
>>> output = pixel_shuffle(input)
>>> print(output.size())
torch.Size([1, 1, 12, 12])