# TODO: Conv2dLocal
# TODO: Conv2dMap
# TODO: ConvTranspose2dMap
import torch
import torch.nn as nn
import torch.nn.functional as F
[docs]class MLPConv2d(nn.Module):
"""__init__(self, in_channels, out_channels, ksize=None, stride=1, pad=0, activation=relu.relu, conv_init=None, bias_init=None)
Two-dimensional MLP convolution layer of Network in Network.
This is an "mlpconv" layer from the Network in Network paper. This layer
is a two-dimensional convolution layer followed by 1x1 convolution layers
and interleaved activation functions.
Note that it does not apply the activation function to the output of the
last 1x1 convolution layer.
Args:
in_channels (int or None): Number of channels of input arrays.
If it is ``None`` or omitted, parameter initialization will be
deferred until the first forward data pass at which time the size
will be determined.
out_channels (tuple of ints): Tuple of number of channels. The i-th
integer indicates the number of filters of the i-th convolution.
ksize (int or pair of ints): Size of filters (a.k.a. kernels) of the
first convolution layer. ``ksize=k`` and ``ksize=(k, k)`` are
equivalent.
stride (int or pair of ints): Stride of filter applications at the
first convolution layer. ``stride=s`` and ``stride=(s, s)`` are
equivalent.
pad (int or pair of ints): Spatial padding width for input arrays at
the first convolution layer. ``pad=p`` and ``pad=(p, p)`` are
equivalent.
activation (callable):
Activation function for internal hidden units.
You can specify one of activation functions from
:doc:`built-in activation functions </reference/functions>` or
your own function.
It should not be an
:doc:`activation functions with parameters </reference/links>`
(i.e., :class:`~chainer.Link` instance).
The function must accept one argument (the output from each child
link), and return a value.
Returned value must be a Variable derived from the input Variable
to perform backpropagation on the variable.
Note that this function is not applied to the output of this link.
a keyword argument.
.. note:
From v2, `conv_init` and `bias_init` arguments must be specified as
keyword arguments only. We impose this restriction to forbid
users to assume the API for v1 and specify `wscale` option,
that had been between `activation` and `conv_init` arguments in v1.
See: `Network in Network <https://arxiv.org/abs/1312.4400v3>`_.
Attributes:
activation (callable):
Activation function.
See the description in the arguments for details.
""" # NOQA
def __init__(self, in_channels, out_channels, kernel_size=None, stride=1, padding=0,
activation=F.relu, *args, **kwargs):
super(MLPConv2d, self).__init__()
#if kernel_size is None:
# out_channels, kernel_size, in_channels = in_channels, out_channels, None
assert len(out_channels) > 0
self.convs = [nn.Conv2d(in_channels, out_channels[0], kernel_size, stride, padding)]
for n_in, n_out in zip(out_channels, out_channels[1:]):
self.convs.append(nn.Conv2d(n_in, n_out, 1))
self.convs = nn.ModuleList(self.convs)
self.activation = activation
def forward(self, x):
"""Computes the output of the mlpconv layer.
Args:
x (~chainer.Variable): Input image.
Returns:
~chainer.Variable: Output of the mlpconv layer.
"""
f = self.activation
for l in self.convs:
x = f(l(x))
return x
[docs]class UpsampleConvLayer(nn.Module):
"""UpsampleConvLayer
Upsamples the input and then does a convolution. This method gives better results
compared to ConvTranspose2d.
ref: http://distill.pub/2016/deconv-checkerboard/
"""
def __init__(self, in_channels, out_channels, kernel_size=1, stride=1, upsample=None):
super(UpsampleConvLayer, self).__init__()
self.upsample = upsample
reflection_padding = kernel_size // 2
self.reflection_pad = torch.nn.ReflectionPad2d(reflection_padding)
self.conv2d = torch.nn.Conv2d(in_channels, out_channels, kernel_size, stride)
def forward(self, x):
x_in = x
if self.upsample:
x_in = torch.nn.functional.interpolate(x_in, mode='nearest', scale_factor=self.upsample)
out = self.reflection_pad(x_in)
out = self.conv2d(out)
return out