The transpose of
input, filters, output_shape, strides, padding='SAME',
data_format='NHWC', dilations=None, name=None
This operation is sometimes called "deconvolution" after
(Zeiler et al., 2010), but is really the transpose (gradient) of
atrous_conv2d rather than an actual deconvolution.
Tensor of type
float and shape
[batch, height, width,
NHWC data format or
[batch, in_channels, height,
NCHW data format.
Tensor with the same type as
input and shape
width, output_channels, in_channels].
must match that of
Tensor representing the output shape of the
An int or list of
ints that has length
stride of the sliding window for each dimension of
input. If a single
value is given it is replicated in the
W dimension. By default
C dimensions are set to 0. The dimension order is determined
by the value of
data_format, see below for details.
Either the |