mindspore.nn.AdaptiveMaxPool3d

class mindspore.nn.AdaptiveMaxPool3d(output_size, return_indices=False)[source]

Applies a 3D adaptive max pooling over an input signal composed of several input planes.

The output is of size \((D, H, W)\), for any input size. The number of output features is equal to the number of input planes.

Parameters

output_size (Union[int, tuple]) – The target output size is \((D, H, W)\). ouput_size can be a tuple with 3 elements, or a single D for \((D, D, D)\). \(D\), \(H\) and \(W\) can be int or None which means the output size is the same as that of the input.
return_indices (bool) – If return_indices is True, the indices of max value would be output. Default: False.

Inputs:

x (Tensor) - Tensor, has shape of \((C, D, H, W)\) or \((N, C, D, H, W)\) . The suppoerted dtypes are int8, int16, int32, int64, uint8, uint16, uint32, uint64, float16, float32 and float64 data type.

Outputs:

y (Tensor) - Tensor, has the same number of dims and data type as the x .
argmax (Tensor) - Tensor, the indices of the maximum values along with the outputs, has the same shape as y and a dtype of int32. Return this only when return_indices is True.

Raises

TypeError – If x is not a Tensor.
ValueError – If the dimensions number of x is not 4 or 5.
TypeError – If dtype of x is not int8, int16, int32, int64, uint8, uint16, uint32, uint64, float16, float32 or float64.
ValueError – If output_size is neither an int nor a tuple with shape (3,).

Supported Platforms:: GPU CPU

Examples

>>> x = Tensor(np.arange(0,36).reshape((1, 3, 3, 4)).astype(np.float32))
>>> output_size = (1, 1, 2)
>>> net = nn.AdaptiveMaxPool3d(output_size, True)
>>> output = net(x)
>>> print(output[0].asnumpy())
[[[[33. 35.]]]]
>>> print(output[1].asnumpy())
[[[[33 35]]]]