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

In the simplest case, the output value of the layer with input size $(N, C, D, H, W)$ , output $(N, C, D_{o u t}, H_{o u t}, W_{o u t})$ and kernel_size $(k D, k H, k W)$ can be precisely described as:

Usage

nn_max_pool3d(
  kernel_size,
  stride = NULL,
  padding = 0,
  dilation = 1,
  return_indices = FALSE,
  ceil_mode = FALSE
)

Arguments

kernel_size: the size of the window to take a max over
stride: the stride of the window. Default value is kernel_size
padding: implicit zero padding to be added on all three sides
dilation: a parameter that controls the stride of elements in the window
return_indices: if TRUE, will return the max indices along with the outputs. Useful for torch_nn.MaxUnpool3d later
ceil_mode: when TRUE, will use ceil instead of floor to compute the output shape

Details

$\begin{array}{ll} out (N_{i}, C_{j}, d, h, w) = & max_{k = 0, \dots, k D - 1} max_{m = 0, \dots, k H - 1} max_{n = 0, \dots, k W - 1} \\ input (N_{i}, C_{j}, stride[0] \times d + k, stride[1] \times h + m, stride[2] \times w + n) \end{array}$

If padding is non-zero, then the input is implicitly zero-padded on both sides for padding number of points. dilation controls the spacing between the kernel points. It is harder to describe, but this link_ has a nice visualization of what dilation does. The parameters kernel_size, stride, padding, dilation can either be:

a single int – in which case the same value is used for the depth, height and width dimension
a tuple of three ints – in which case, the first int is used for the depth dimension, the second int for the height dimension and the third int for the width dimension

Shape

Input: $(N, C, D_{i n}, H_{i n}, W_{i n})$
Output: $(N, C, D_{o u t}, H_{o u t}, W_{o u t})$ , where $D_{o u t} = ⌊ \frac{D_{i n} + 2 \times padding [0] - dilation [0] \times (kernel\_size [0] - 1) - 1}{stride [0]} + 1 ⌋$

$H_{o u t} = ⌊ \frac{H_{i n} + 2 \times padding [1] - dilation [1] \times (kernel\_size [1] - 1) - 1}{stride [1]} + 1 ⌋$

$W_{o u t} = ⌊ \frac{W_{i n} + 2 \times padding [2] - dilation [2] \times (kernel\_size [2] - 1) - 1}{stride [2]} + 1 ⌋$

Examples

if (torch_is_installed()) {
# pool of square window of size=3, stride=2
m <- nn_max_pool3d(3, stride = 2)
# pool of non-square window
m <- nn_max_pool3d(c(3, 2, 2), stride = c(2, 1, 2))
input <- torch_randn(20, 16, 50, 44, 31)
output <- m(input)
}