mindspore.mint.nn.functional.avg_pool2d

mindspore.mint.nn.functional.avg_pool2d(input, kernel_size, stride=None, padding=0, ceil_mode=False, count_include_pad=True, divisor_override=None)[source]

Applies a 2D average pooling over an input Tensor which can be regarded as a composition of 2D input planes. Typically the input is of shape $(N, C, H_{i n}, W_{i n})$ , outputs regional average in the $(H_{i n}, W_{i n})$ -dimension. Given kernel size $(k H, k W)$ and stride , the operation is as follows.

output (N_{i}, C_{j}, h, w) = \frac{1}{k H * k W} \sum_{m = 0}^{k H - 1} \sum_{n = 0}^{k W - 1} input (N_{i}, C_{j}, s t r i d e [0] \times h + m, s t r i d e [1] \times w + n)

Note

On the Atlas platform, when calculating the input, the precision is degraded from float32 to float16.

Parameters

input (Tensor) – Tensor of shape $(N, C, H_{i n}, W_{i n})$ or $(C, H_{i n}, W_{i n})$ .
kernel_size (Union[int, tuple[int], list[int]]) – The size of kernel used to take the average value. Can be a single number or a tuple $(k H, k W)$ .
stride (Union[int, tuple[int], list[int]], optional) – The distance of kernel moving. Can be a single number or a tuple $(s H, s W)$ . Default: None, where its value is equal to kernel_size.
padding (Union[int, tuple[int], list[int]], optional) – Implicit zero padding to be added on both sides. Can be a single number or a tuple $(p a d H, p a d W)$ . Default: 0.
ceil_mode (bool, optional) – If True, apply ceil instead of floor to compute the output shape. Default: False.
count_include_pad (bool, optional) – If True, include the zero-padding in the averaging calculation. Default: True .
divisor_override (int, optional) – If specified, it will be used as divisor in the averaging calculation, otherwise size of pooling region will be used. Default: None.

Returns

Tensor, with shape $(N, C, H_{o u t}, W_{o u t})$ or $(C, H_{o u t}, W_{o u t})$ .

\begin{array}{r} \begin{array}{ll} H_{o u t} = \frac{H_{i n} + 2 \times p a d d i n g [0] - k e r n e l_s i z e [0]}{s t r i d e [0]} + 1 \\ W_{o u t} = \frac{W_{i n} + 2 \times p a d d i n g [1] - k e r n e l_s i z e [1]}{s t r i d e [1]} + 1 \end{array} \end{array}

Raises

TypeError – If input is not a Tensor.
TypeError – If kernel_size or stride is neither int nor tuple.
TypeError – If ceil_mode or count_include_pad is not a bool.
TypeError – If divisor_override is not an int or None.
ValueError – If the dimension of input is not equal to 3 or 4.
ValueError – If kernel_size or stride is less than 1.
ValueError – If value of padding is less than 0.
ValueError – If kernel_size, padding or stride is a tuple whose length is not equal to 1 or 2.

Supported Platforms:: Ascend

Examples

>>> import mindspore
>>> import numpy as np
>>> from mindspore import Tensor, mint
>>> x = Tensor(np.arange(1 * 3 * 3 * 4).reshape(1, 3, 3, 4), mindspore.float32)
>>> output = mint.nn.functional.avg_pool2d(x, kernel_size=2, stride=1)
>>> print(output)
[[[[ 2.5   3.5   4.5]
   [ 6.5   7.5   8.5]]
  [[14.5  15.5  16.5]
   [18.5  19.5  20.5]]
  [[26.5  27.5  28.5]
   [30.5  31.5  32.5]]]]