tensorflow::ops::SpaceToBatchND
tensorflow::ops::SpaceToBatchND
#include <array_ops.h>
SpaceToBatch for N-D tensors of type T.
Summary
This operation divides "spatial" dimensions [1, ..., M]
of the input into a grid of blocks of shape block_shape
, and interleaves these blocks with the "batch" dimension (0) such that in the output, the spatial dimensions [1, ..., M]
correspond to the position within the grid, and the batch dimension combines both the position within a spatial block and the original batch position. Prior to division into blocks, the spatial dimensions of the input are optionally zero padded according to paddings
. See below for a precise description.
Arguments:
- scope: A Scope object
- input: N-D with shape
input_shape = [batch] + spatial_shape + remaining_shape
, where spatial_shape hasM
dimensions. - block_shape: 1-D with shape
[M]
, all values must be >= 1. - paddings: 2-D with shape
[M, 2]
, all values must be >= 0.paddings[i] = [pad_start, pad_end]
specifies the padding for input dimensioni + 1
, which corresponds to spatial dimensioni
. It is required thatblock_shape[i]
dividesinput_shape[i + 1] + pad_start + pad_end
.
This operation is equivalent to the following steps:
- Zero-pad the start and end of dimensions
[1, ..., M]
of the input according topaddings
to producepadded
of shapepadded_shape
. -
Reshape
padded
toreshaped_padded
of shape:[batch] + [padded_shape[1] / block_shape[0], block_shape[0], ..., padded_shape[M] / block_shape[M-1], block_shape[M-1]] + remaining_shape - Permute dimensions of
reshaped_padded
to producepermuted_reshaped_padded
of shape:block_shape + [batch] + [padded_shape[1] / block_shape[0], ..., padded_shape[M] / block_shape[M-1]] + remaining_shape -
Reshape
permuted_reshaped_padded
to flattenblock_shape
into the batch dimension, producing an output tensor of shape:[batch * prod(block_shape)] + [padded_shape[1] / block_shape[0], ..., padded_shape[M] / block_shape[M-1]] + remaining_shape
Some examples:
(1) For the following input of shape [1, 2, 2, 1]
, block_shape = [2, 2]
, and paddings = [[0, 0], [0, 0]]
:
``` x = [[[[1], [2]], [[3], [4]]]] ```
The output tensor has shape [4, 1, 1, 1]
and value:
``` [[[[1]]], [[[2]]], [[[3]]], [[[4]]]] ```
(2) For the following input of shape [1, 2, 2, 3]
, block_shape = [2, 2]
, and paddings = [[0, 0], [0, 0]]
:
``` x = [[[[1, 2, 3], [4, 5, 6]], [[7, 8, 9], [10, 11, 12]]]] ```
The output tensor has shape [4, 1, 1, 3]
and value:
``` [[[1, 2, 3]], [[4, 5, 6]], [[7, 8, 9]], [[10, 11, 12]]] ```
(3) For the following input of shape [1, 4, 4, 1]
, block_shape = [2, 2]
, and paddings = [[0, 0], [0, 0]]
:
``` x = [[[[1], [2], [3], [4]], [[5], [6], [7], [8]], [[9], [10], [11], [12]], [[13], [14], [15], [16]]]] ```
The output tensor has shape [4, 2, 2, 1]
and value:
``` x = [[[[1], [3]], [[9], [11]]], [[[2], [4]], [[10], [12]]], [[[5], [7]], [[13], [15]]], [[[6], [8]], [[14], [16]]]] ```
(4) For the following input of shape [2, 2, 4, 1]
, block_shape = [2, 2]
, and paddings = [[0, 0], [2, 0]]
:
``` x = [[[[1], [2], [3], [4]], [[5], [6], [7], [8]]], [[[9], [10], [11], [12]], [[13], [14], [15], [16]]]] ```
The output tensor has shape [8, 1, 3, 1]
and value:
``` x = [[[[0], [1], [3]]], [[[0], [9], [11]]], [[[0], [2], [4]]], [[[0], [10], [12]]], [[[0], [5], [7]]], [[[0], [13], [15]]], [[[0], [6], [8]]], [[[0], [14], [16]]]] ```
Among others, this operation is useful for reducing atrous convolution into regular convolution.
Returns:
-
Output
: The output tensor.
Constructors and Destructors | |
---|---|
SpaceToBatchND(const ::tensorflow::Scope & scope, ::tensorflow::Input input, ::tensorflow::Input block_shape, ::tensorflow::Input paddings) |
Public attributes | |
---|---|
output |
Public functions | |
---|---|
node() const | ::tensorflow::Node * |
operator::tensorflow::Input() const | |
operator::tensorflow::Output() const |
Public attributes
output
::tensorflow::Output output
Public functions
SpaceToBatchND
SpaceToBatchND( const ::tensorflow::Scope & scope, ::tensorflow::Input input, ::tensorflow::Input block_shape, ::tensorflow::Input paddings )
node
::tensorflow::Node * node() const
operator::tensorflow::Input
operator::tensorflow::Input() const
operator::tensorflow::Output
operator::tensorflow::Output() const
© 2017 The TensorFlow Authors. All rights reserved.
Licensed under the Creative Commons Attribution License 3.0.
Code samples licensed under the Apache 2.0 License.
https://www.tensorflow.org/api_docs/cc/class/tensorflow/ops/space-to-batch-n-d.html