#! /usr/bin/python
# -*- coding: utf-8 -*-
import tensorlayerx as tlx
from tensorlayerx import logging
from tensorlayerx.nn.core import Module
__all__ = [
'GroupConv2d',
]
[docs]class GroupConv2d(Module):
"""The :class:`GroupConv2d` class is 2D grouped convolution, see `here <https://blog.yani.io/filter-group-tutorial/>`__.
Parameters
--------------
out_channels : int
The number of filters.
kernel_size : tuple or int
The filter size.
stride : tuple or int
The stride step.
n_group : int
The number of groups.
act : activation function
The activation function of this layer.
padding : str
The padding algorithm type: "SAME" or "VALID".
data_format : str
"channels_last" (NHWC, default) or "channels_first" (NCHW).
dilation : tuple or int
Specifying the dilation rate to use for dilated convolution.
W_init : initializer or str
The initializer for the weight matrix.
b_init : initializer or None or str
The initializer for the bias vector. If None, skip biases.
in_channels : int
The number of in channels.
name : None or str
A unique layer name.
Examples
---------
With TensorLayer
>>> net = tlx.nn.Input([8, 24, 24, 32], name='input')
>>> groupconv2d = tlx.nn.GroupConv2d(
... out_channels=64, kernel_size=(3, 3), stride=(2, 2), n_group=2, name='group'
... )(net)
>>> print(groupconv2d)
>>> output shape : (8, 12, 12, 64)
"""
def __init__(
self,
out_channels=32,
kernel_size=(1, 1),
stride=(1, 1),
n_group=1,
act=None,
padding='SAME',
data_format="channels_last",
dilation=(1, 1),
W_init='truncated_normal',
b_init='constant',
in_channels=None,
name=None
):
super().__init__(name, act=act)
self.out_channels = out_channels
self.kernel_size = self.check_param(kernel_size)
self._stride = self.stride = self.check_param(stride)
self.n_group = n_group
self.padding = padding
self.data_format = data_format
self._dilation_rate = self.dilation = self.check_param(dilation)
self.W_init = self.str_to_init(W_init)
self.b_init = self.str_to_init(b_init)
self.in_channels = in_channels
if self.in_channels:
self.build(None)
self._built = True
logging.info(
"Conv2d %s: out_channels: %d kernel_size: %s stride: %s n_group: %d pad: %s act: %s" % (
self.name, out_channels, str(kernel_size), str(stride), n_group, padding,
self.act.__class__.__name__ if self.act is not None else 'No Activation'
)
)
def __repr__(self):
actstr = self.act.__class__.__name__ if self.act is not None else "No Activation"
s = (
'{classname}(in_channels={in_channels}, out_channels={out_channels}, kernel_size={kernel_size}'
', stride={stride}, n_group = {n_group}, padding={padding}'
)
if self.dilation != (1, ) * len(self.dilation):
s += ', dilation = {dilation}'
if self.b_init is None:
s += ', bias=False'
s += (',', +actstr)
if self.name is not None:
s += ', name=\'{name}\''
s += ')'
return s.format(classname=self.__class__.__name__, **self.__dict__)
def build(self, inputs_shape):
if self.data_format == 'channels_last':
if self.in_channels is None:
self.in_channels = inputs_shape[-1]
self._stride = [1, self._stride[0], self._stride[1], 1]
self._dilation_rate = [1, self._dilation_rate[0], self._dilation_rate[1], 1]
elif self.data_format == 'channels_first':
if self.in_channels is None:
self.in_channels = inputs_shape[1]
self._stride = [1, 1, self._stride[0], self._stride[1]]
self._dilation_rate = [1, 1, self._dilation_rate[0], self._dilation_rate[1]]
else:
raise Exception("data_format should be either channels_last or channels_first")
if self.n_group < 1:
raise ValueError(
"The n_group must be a integer greater than or equal to 1, but we got :{}".format(self.n_group)
)
if self.in_channels % self.n_group != 0:
raise ValueError(
"The channels of input must be divisible by n_group, but we got: the channels of input"
"is {}, the n_group is {}.".format(self.in_channels, self.n_group)
)
if self.out_channels % self.n_group != 0:
raise ValueError(
"The number of filters must be divisible by n_group, but we got: the number of filters "
"is {}, the n_group is {}. ".format(self.out_channels, self.n_group)
)
# TODO channels first filter shape [out_channel, in_channel/n_group, filter_h, filter_w]
self.filter_shape = (
self.kernel_size[0], self.kernel_size[1], int(self.in_channels / self.n_group), self.out_channels
)
self.filters = self._get_weights("filters", shape=self.filter_shape, init=self.W_init)
self.b_init_flag = False
if self.b_init:
self.biases = self._get_weights("biases", shape=(self.out_channels, ), init=self.b_init)
self.bias_add = tlx.ops.BiasAdd(self.data_format)
self.b_init_flag = True
self.group_conv2d = tlx.ops.GroupConv2D(
strides=self._stride, padding=self.padding, data_format=self.data_format, dilations=self._dilation_rate,
out_channel=self.out_channels, k_size=(self.kernel_size[0], self.kernel_size[1]), groups=self.n_group
)
self.act_init_flag = False
if self.act:
self.act_init_flag = True
def forward(self, inputs):
if self._forward_state == False:
if self._built == False:
self.build(tlx.get_tensor_shape(inputs))
self._built = True
self._forward_state = True
outputs = self.group_conv2d(inputs, self.filters)
if self.b_init_flag:
outputs = self.bias_add(outputs, self.biases)
if self.act_init_flag:
outputs = self.act(outputs)
if not self._nodes_fixed and self._build_graph:
self._add_node(inputs, outputs)
self._nodes_fixed = True
return outputs