def __init__(self,
in_channels,
out_channels,
strides,
padding=1,
dilation=1,
conv1_stride=False,
bottleneck_factor=4,
data_format="channels_last",
**kwargs):
super(ResBottleneck, self).__init__(**kwargs)
mid_channels = out_channels // bottleneck_factor
self.conv1 = conv1x1_block(in_channels=in_channels,
out_channels=mid_channels,
strides=(strides if conv1_stride else 1),
data_format=data_format,
name="conv1")
self.conv2 = conv3x3_block(in_channels=mid_channels,
out_channels=mid_channels,
strides=(1 if conv1_stride else strides),
padding=padding,
dilation=dilation,
data_format=data_format,
name="conv2")
self.conv3 = conv1x1_block(in_channels=mid_channels,
out_channels=out_channels,
activation=None,
data_format=data_format,
name="conv3")
def __init__(self,
in_channels,
out_channels,
strides,
bottleneck_factor=4,
data_format="channels_last",
**kwargs):
super(ResNeStABottleneck, self).__init__(**kwargs)
self.resize = (strides > 1)
mid_channels = out_channels // bottleneck_factor
self.conv1 = conv1x1_block(in_channels=in_channels,
out_channels=mid_channels,
data_format=data_format,
name="conv1")
self.conv2 = saconv3x3_block(in_channels=mid_channels,
out_channels=mid_channels,
data_format=data_format,
name="conv2")
if self.resize:
self.pool = AvgPool2d(pool_size=3,
strides=strides,
padding=1,
data_format=data_format,
name="pool")
self.conv3 = conv1x1_block(in_channels=mid_channels,
out_channels=out_channels,
activation=None,
data_format=data_format,
name="conv3")
def __init__(self,
in_channels,
out_channels,
strides,
cardinality,
bottleneck_width,
data_format="channels_last",
**kwargs):
super(SENetBottleneck, self).__init__(**kwargs)
mid_channels = out_channels // 4
D = int(math.floor(mid_channels * (bottleneck_width / 64.0)))
group_width = cardinality * D
group_width2 = group_width // 2
self.conv1 = conv1x1_block(in_channels=in_channels,
out_channels=group_width2,
data_format=data_format,
name="conv1")
self.conv2 = conv3x3_block(in_channels=group_width2,
out_channels=group_width,
strides=strides,
groups=cardinality,
data_format=data_format,
name="conv2")
self.conv3 = conv1x1_block(in_channels=group_width,
out_channels=out_channels,
activation=None,
data_format=data_format,
name="conv3")
def __init__(self,
in_channels,
out_channels,
strides,
conv1_ibn,
data_format="channels_last",
**kwargs):
super(IBNResBottleneck, self).__init__(**kwargs)
mid_channels = out_channels // 4
self.conv1 = ibn_conv1x1_block(in_channels=in_channels,
out_channels=mid_channels,
use_ibn=conv1_ibn,
data_format=data_format,
name="conv1")
self.conv2 = conv3x3_block(in_channels=mid_channels,
out_channels=mid_channels,
strides=strides,
data_format=data_format,
name="conv2")
self.conv3 = conv1x1_block(in_channels=mid_channels,
out_channels=out_channels,
activation=None,
data_format=data_format,
name="conv3")
def __init__(self,
in_channels,
out_channels,
strides,
data_format="channels_last",
**kwargs):
super(ResNeStADownBlock, self).__init__(**kwargs)
self.pool = AvgPool2d(pool_size=strides,
strides=strides,
ceil_mode=True,
data_format=data_format,
name="pool")
self.conv = conv1x1_block(in_channels=in_channels,
out_channels=out_channels,
activation=None,
data_format=data_format,
name="conv")
def __init__(self,
in_channels,
out_channels,
strides,
padding=1,
dilation=1,
use_bias=False,
use_bn=True,
bottleneck=True,
conv1_stride=False,
data_format="channels_last",
**kwargs):
super(ResUnit, self).__init__(**kwargs)
self.resize_identity = (in_channels != out_channels) or (strides != 1)
if bottleneck:
self.body = ResBottleneck(in_channels=in_channels,
out_channels=out_channels,
strides=strides,
padding=padding,
dilation=dilation,
conv1_stride=conv1_stride,
data_format=data_format,
name="body")
else:
self.body = ResBlock(in_channels=in_channels,
out_channels=out_channels,
strides=strides,
use_bias=use_bias,
use_bn=use_bn,
data_format=data_format,
name="body")
if self.resize_identity:
self.identity_conv = conv1x1_block(in_channels=in_channels,
out_channels=out_channels,
strides=strides,
use_bias=use_bias,
use_bn=use_bn,
activation=None,
data_format=data_format,
name="identity_conv")
self.activ = nn.ReLU()
def __init__(self,
in_channels,
out_channels,
strides,
cardinality,
bottleneck_width,
identity_conv3x3,
data_format="channels_last",
**kwargs):
super(SENetUnit, self).__init__(**kwargs)
self.resize_identity = (in_channels != out_channels) or (strides != 1)
self.body = SENetBottleneck(in_channels=in_channels,
out_channels=out_channels,
strides=strides,
cardinality=cardinality,
bottleneck_width=bottleneck_width,
data_format=data_format,
name="body")
self.se = SEBlock(channels=out_channels,
data_format=data_format,
name="se")
if self.resize_identity:
if identity_conv3x3:
self.identity_conv = conv3x3_block(in_channels=in_channels,
out_channels=out_channels,
strides=strides,
activation=None,
data_format=data_format,
name="identity_conv")
else:
self.identity_conv = conv1x1_block(in_channels=in_channels,
out_channels=out_channels,
strides=strides,
activation=None,
data_format=data_format,
name="identity_conv")
self.activ = nn.ReLU()
def __init__(self,
in_channels,
out_channels,
strides,
conv1_ibn,
data_format="channels_last",
**kwargs):
super(IBNResUnit, self).__init__(**kwargs)
self.resize_identity = (in_channels != out_channels) or (strides != 1)
self.body = IBNResBottleneck(in_channels=in_channels,
out_channels=out_channels,
strides=strides,
conv1_ibn=conv1_ibn,
data_format=data_format,
name="body")
if self.resize_identity:
self.identity_conv = conv1x1_block(in_channels=in_channels,
out_channels=out_channels,
strides=strides,
activation=None,
data_format=data_format,
name="identity_conv")
self.activ = nn.ReLU()