mindspore.nn.ChannelShuffle

class mindspore.nn.ChannelShuffle(groups)

Divide the channels in a tensor of shape $(\ast ,C,H,W)$ into $g$ group and rearrange them as $(\ast ,\frac{C}{g},g,H\ast W)$, while retaining the original tensor shape in the final output.

Parameters

groups (int) – Number of groups to divide channels in, must be greater than 0. Refer to $g$ in the above formula.

Inputs:

• x (Tensor) - Tensor of shape $(\ast ,C_{in},H_{in},W_{in})$.

Outputs:

Tensor, with the same type and shape as the x.

Raises

• TypeError – If groups is not a positive integer.

• ValueError – If groups is less than 1.

• ValueError – If dims of x is less than 3.

• ValueError – If number of channels can not be divisible by groups.

Supported Platforms:

Ascend GPU CPU

Examples

>>> import mindspore as ms
>>> import numpy as np
>>> channel_shuffle = ms.nn.ChannelShuffle(2)
>>> x = ms.Tensor(np.arange(16).astype(np.int32).reshape(1, 4, 2, 2))
>>> print(x)
[[[[ 0  1]
   [ 2  3]]
  [[ 4  5]
   [ 6  7]]
  [[ 8  9]
   [10 11]]
  [[12 13]
   [14 15]]]]
>>> output = channel_shuffle(x)
>>> print(output)
[[[[ 0  1]
   [ 2  3]]
  [[ 8  9]
   [10 11]]
  [[ 4  5]
   [ 6  7]]
  [[12 13]
   [14 15]]]]