def __init__(self,
nclass,
backbone='mobilenet_v2_1_0',
aux=True,
height=None,
width=None,
base_size=520,
crop_size=480,
pretrained_base=True,
norm_layer=nn.BatchNorm,
norm_kwargs=None,
**kwargs):
super(FCNMobileNet, self).__init__(nclass,
aux,
backbone,
height,
width,
base_size,
crop_size,
pretrained_base,
norm_layer=norm_layer,
norm_kwargs=norm_kwargs)
with self.name_scope():
self.head = FCNHead(nclass=nclass,
in_channels=self.base_channels[3],
norm_layer=norm_layer,
norm_kwargs=norm_kwargs)
if self.aux:
self.auxlayer = FCNHead(nclass=nclass,
in_channels=self.base_channels[2],
norm_layer=norm_layer,
norm_kwargs=norm_kwargs)
def __init__(self, nclass, capacity=512, attention=False, drop=.4,
norm_layer=nn.BatchNorm, norm_kwargs=None, height=120, width=120):
super(_CAHead, self).__init__()
self.up_kwargs = {'height': height, 'width': width}
self.attention = attention
self.gamma = 1.0
height = height // 2
width = width // 2
with self.name_scope():
# Chained Context Aggregation Module
self.gp = GlobalFlow(capacity, 2048, norm_layer, norm_kwargs,
height=height, width=width)
self.cp1 = _ContextFlow(capacity, stride=2, norm_layer=norm_layer,
norm_kwargs=norm_kwargs, height=height, width=width)
self.cp2 = _ContextFlow(capacity, stride=4, norm_layer=norm_layer,
norm_kwargs=norm_kwargs, height=height, width=width)
self.cp3 = _ContextFlow(capacity, stride=8, norm_layer=norm_layer,
norm_kwargs=norm_kwargs, height=height, width=width)
self.cp4 = _ContextFlow(capacity, stride=16, norm_layer=norm_layer,
norm_kwargs=norm_kwargs, height=height, width=width)
if self.attention:
self.selection = _FeatureSelection(256, in_channels=capacity, norm_layer=norm_layer,
norm_kwargs=norm_kwargs)
else:
self.proj = ConvBlock(256, 3, 1, 1, in_channels=capacity, norm_layer=norm_layer,
norm_kwargs=norm_kwargs)
self.drop = nn.Dropout(drop) if drop else None
# decoder
self.decoder = ConvBlock(48, 3, 1, 1, norm_layer=norm_layer, norm_kwargs=norm_kwargs)
self.conv3x3 = ConvBlock(256, 3, 1, 1, norm_layer=norm_layer, norm_kwargs=norm_kwargs)
# segmentation head
self.seg_head = FCNHead(nclass, norm_layer=norm_layer, norm_kwargs=norm_kwargs)
def __init__(self,
nclass,
decoder_capacity,
input_height,
input_width,
norm_layer=nn.BatchNorm,
norm_kwargs=None):
super(_LadderHead, self).__init__()
with self.name_scope():
self.conv_c4 = ConvBlock(decoder_capacity,
1,
norm_layer=norm_layer,
norm_kwargs=norm_kwargs,
activation='relu')
self.fusion_c3 = _LadderFusion(decoder_capacity,
input_height // 16,
input_width // 16,
norm_layer=norm_layer,
norm_kwargs=norm_kwargs)
self.fusion_c2 = _LadderFusion(decoder_capacity,
input_height // 8,
input_width // 8,
norm_layer=norm_layer,
norm_kwargs=norm_kwargs)
self.fusion_c1 = _LadderFusion(decoder_capacity,
input_height // 4,
input_width // 4,
norm_layer=norm_layer,
norm_kwargs=norm_kwargs)
self.seg_head = FCNHead(nclass,
decoder_capacity,
norm_layer,
norm_kwargs,
activation='relu')
def __init__(self,
nclass,
backbone='densnet169',
aux=True,
height=None,
width=None,
base_size=520,
crop_size=480,
pretrained_base=False,
norm_layer=nn.BatchNorm,
norm_kwargs=None,
**kwargs):
super(LadderDenseNet, self).__init__(nclass,
aux,
backbone,
height,
width,
base_size,
crop_size,
pretrained_base,
norm_layer=norm_layer,
norm_kwargs=norm_kwargs)
decoder_capacity = 128
with self.name_scope():
self.head = _LadderHead(nclass,
decoder_capacity,
norm_layer=norm_layer,
norm_kwargs=norm_kwargs,
input_height=self._up_kwargs['height'],
input_width=self._up_kwargs['width'])
if self.aux:
self.auxlayer = FCNHead(nclass=nclass,
in_channels=decoder_capacity,
norm_layer=norm_layer,
norm_kwargs=norm_kwargs)
def __init__(self,
nclass,
backbone='xception65',
aux=True,
height=None,
width=None,
base_size=520,
crop_size=480,
pretrained_base=True,
norm_layer=nn.BatchNorm,
norm_kwargs=None,
**kwargs):
super(DeepLabv3Plus, self).__init__(nclass,
aux,
backbone,
height,
width,
base_size,
crop_size,
pretrained_base,
dilate=True,
norm_layer=norm_layer,
norm_kwargs=norm_kwargs)
with self.name_scope():
self.head = _DeepLabHead(nclass,
2048,
norm_layer,
norm_kwargs,
height=self._up_kwargs['height'] // 4,
width=self._up_kwargs['width'] // 4)
if self.aux:
self.auxlayer = FCNHead(nclass, 728, norm_layer, norm_kwargs)
def __init__(self, nclass, backbone='resnet50', aux=True, height=None, width=None,
base_size=520, crop_size=480, pretrained_base=False,
norm_layer=nn.BatchNorm, norm_kwargs=None, **kwargs):
super(DenseASPP, self).__init__(nclass, aux, backbone, height, width,
base_size, crop_size, pretrained_base, dilate=True,
norm_layer=norm_layer, norm_kwargs=norm_kwargs)
with self.name_scope():
self.head = _DenseASPPHead(nclass, 2048, norm_layer, norm_kwargs)
if self.aux:
self.auxlayer = FCNHead(nclass, 1024, norm_layer, norm_kwargs)
def __init__(self, nclass, backbone='resnet50', aux=True, height=None, width=None,
base_size=520, crop_size=480, pretrained_base=False,
norm_layer=nn.BatchNorm, norm_kwargs=None, **kwargs):
super(CANet, self).__init__(nclass, aux, backbone, height, width, base_size, crop_size,
pretrained_base, dilate=True, norm_layer=norm_layer,
norm_kwargs=norm_kwargs)
with self.name_scope():
self.head = _CAHead(nclass, norm_layer=norm_layer, norm_kwargs=norm_kwargs,
height=self._up_kwargs['height'] // 4,
width=self._up_kwargs['width'] // 4)
if self.aux:
self.auxlayer = FCNHead(nclass, 1024, norm_layer, norm_kwargs)
def __init__(self, nclass, backbone='resnet18', aux=True, height=None, width=None,
base_size=520, crop_size=480, pretrained_base=False, norm_layer=nn.BatchNorm,
norm_kwargs=None, **kwargs):
super(BiSeNet, self).__init__(nclass, aux, backbone, height, width, base_size,
crop_size, pretrained_base, dilate=False,
norm_layer=norm_layer, norm_kwargs=norm_kwargs)
self.head = _BiSeNetHead(nclass, norm_layer=norm_layer, norm_kwargs=norm_kwargs,
height=self._up_kwargs['height'] // 8,
width=self._up_kwargs['width'] // 8)
if self.aux:
self.auxlayer = FCNHead(nclass, norm_layer=norm_layer, norm_kwargs=norm_kwargs,
drop_out=.0)
def __init__(self, nclass, backbone='resnet50', aux=True, height=None, width=None,
base_size=520, crop_size=480, pretrained_base=False, dilate=True,
norm_layer=nn.BatchNorm, norm_kwargs=None, **kwargs):
super(DeepLabv3, self).__init__(nclass, aux, backbone, height, width, base_size,
crop_size, pretrained_base, dilate,
norm_layer=norm_layer, norm_kwargs=norm_kwargs)
self.output_stride = 8 if dilate else 32
with self.name_scope():
self.head = _DeepLabHead(nclass, self.base_channels[3], norm_layer, norm_kwargs,
height=self._up_kwargs['height'] // self.output_stride,
width=self._up_kwargs['width'] // self.output_stride)
if self.aux:
self.auxlayer = FCNHead(nclass, self.base_channels[2], norm_layer, norm_kwargs)
def __init__(self, nclass, height, width, norm_layer=nn.BatchNorm, norm_kwargs=None):
super(_BiSeNetHead, self).__init__()
self.up_kwargs = {'height': height, 'width': width}
with self.name_scope():
self.spatial_path = _SpatialPath(128, norm_layer=norm_layer, norm_kwargs=norm_kwargs)
self.global_flow = GlobalFlow(128, in_channels=512, norm_layer=norm_layer,
norm_kwargs=norm_kwargs, height=height, width=width)
self.refine_c4 = _ARModule(128, in_channels=512, norm_layer=norm_layer,
norm_kwargs=norm_kwargs)
self.refine_c3 = _ARModule(128, in_channels=256, norm_layer=norm_layer,
norm_kwargs=norm_kwargs)
self.proj = ConvBlock(128, 3, 1, 1, norm_layer=norm_layer, norm_kwargs=norm_kwargs)
self.fusion = _FFModule(256, norm_layer, norm_kwargs, reduction=1)
self.seg = FCNHead(nclass, in_channels=256, norm_layer=norm_layer, norm_kwargs=norm_kwargs,
drop_out=.0, reduction=4)
def __init__(self,
nclass,
num_scale=2,
height=60,
width=60,
norm_layer=nn.BatchNorm,
norm_kwargs=None):
super(_AttentionHead, self).__init__()
self.num_scale = num_scale
self.up_kwargs = {'height': height, 'width': width}
with self.name_scope():
self.seg_head = FCNHead(nclass,
norm_layer=norm_layer,
norm_kwargs=norm_kwargs)
self.conv3x3 = ConvBlock(512,
3,
1,
1,
norm_layer=norm_layer,
norm_kwargs=norm_kwargs,
activation='relu')
self.conv1x1 = nn.Conv2D(num_scale, 1, in_channels=512)
def __init__(self,
nclass,
input_height,
input_width,
capacity=128,
norm_layer=nn.BatchNorm,
norm_kwargs=None):
super(_SwiftNetHead, self).__init__()
with self.name_scope():
self.conv1x1 = ConvBlock(capacity,
1,
norm_layer=norm_layer,
norm_kwargs=norm_kwargs)
self.fusion_32x = _LateralFusion(capacity,
input_height // 32,
input_width // 32,
norm_layer=norm_layer,
norm_kwargs=norm_kwargs)
self.fusion_16x = _LateralFusion(capacity,
input_height // 16,
input_width // 16,
norm_layer=norm_layer,
norm_kwargs=norm_kwargs)
self.fusion_8x = _LateralFusion(capacity,
input_height // 8,
input_width // 8,
norm_layer=norm_layer,
norm_kwargs=norm_kwargs)
self.final = _LateralFusion(capacity,
input_height // 4,
input_width // 4,
True,
norm_layer=norm_layer,
norm_kwargs=norm_kwargs)
self.seg_head = FCNHead(nclass,
capacity,
norm_layer,
norm_kwargs,
reduction=1)
def __init__(self,
nclass,
in_channels,
norm_layer=nn.BatchNorm,
norm_kwargs=None,
height=60,
width=60):
super(_DeepLabHead, self).__init__()
self.up_kwargs = {'height': height, 'width': width}
with self.name_scope():
self.aspp = ASPP(256,
in_channels,
norm_layer,
norm_kwargs,
height // 2,
width // 2,
atrous_rates=(12, 24, 36))
self.conv_c1 = ConvBlock(48,
3,
1,
1,
norm_layer=norm_layer,
norm_kwargs=norm_kwargs)
self.conv3x3 = ConvBlock(256,
3,
1,
1,
in_channels=304,
norm_layer=norm_layer,
norm_kwargs=norm_kwargs)
self.drop = nn.Dropout(0.5)
self.head = FCNHead(nclass,
256,
norm_layer,
norm_kwargs,
reduction=1)
def __init__(self, nclass, height, width, norm_layer=nn.BatchNorm, norm_kwargs=None, activation='relu'):
super(_PyramidHead, self).__init__()
with self.name_scope():
self.pool = PyramidPooling(2048, height, width, norm_layer, norm_kwargs, activation, reduction=4)
self.seg_head = FCNHead(nclass, 4096, norm_layer, norm_kwargs)
def __init__(self, nclass, in_channels, norm_layer=nn.BatchNorm, norm_kwargs=None):
super(_DenseASPPHead, self).__init__()
with self.name_scope():
self.dense_aspp = DenseASPPBlock(256, in_channels, norm_layer, norm_kwargs)
self.head = FCNHead(nclass, 256, norm_layer, norm_kwargs, reduction=1)
def __init__(self, nclass, in_channels, norm_layer=nn.BatchNorm,
norm_kwargs=None, height=60, width=60):
super(_DeepLabHead, self).__init__()
with self.name_scope():
self.aspp = ASPP(256, in_channels, norm_layer, norm_kwargs, height, width)
self.head = FCNHead(nclass, 256, norm_layer, norm_kwargs, reduction=1)