File size: 11,977 Bytes
edcf5ee |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 |
import torch
import torch.nn as nn
import torch.nn.functional as F
from timm.models.layers import trunc_normal_, DropPath
import pdb
__all__ = [
'ConvNormAct',
'SingleConv',
'BasicBlock',
'Bottleneck',
'DepthwiseSeparableConv',
'SEBlock',
'DropPath',
'MBConv',
'FusedMBConv',
'ConvNeXtBlock',
'LayerNorm'
]
class ConvNormAct(nn.Module):
"""
Layer grouping a convolution, normalization and activation funtion
normalization includes BN and IN
"""
def __init__(self, in_ch, out_ch, kernel_size=3, stride=1, padding=0,
groups=1, dilation=1, bias=False, norm=nn.BatchNorm2d, act=nn.ReLU, preact=False):
super().__init__()
assert norm in [nn.BatchNorm2d, nn.InstanceNorm2d, True, False]
assert act in [nn.ReLU, nn.ReLU6, nn.GELU, nn.SiLU, True, False]
self.conv = nn.Conv2d(
in_channels=in_ch,
out_channels=out_ch,
kernel_size=kernel_size,
stride=stride,
padding=padding,
groups=groups,
dilation=dilation,
bias=bias
)
if preact:
self.norm = norm(in_ch) if norm else nn.Identity()
else:
self.norm = norm(out_ch) if norm else nn.Identity()
self.act = act() if act else nn.Identity()
self.preact = preact
def forward(self, x):
if self.preact:
out = self.conv(self.act(self.norm(x))) # norm relu conv
else:
out = self.act(self.norm(self.conv(x))) # conv norm relu
return out
class SingleConv(nn.Module):
def __init__(self, in_ch, out_ch, stride=1, norm=nn.BatchNorm2d, act=nn.ReLU, preact=False):
super().__init__()
assert norm in [nn.BatchNorm2d, nn.InstanceNorm2d, LayerNorm, True, False]
assert act in [nn.ReLU, nn.ReLU6, nn.GELU, nn.SiLU, True, False]
self.conv = ConvNormAct(in_ch, out_ch, 3, stride=stride, padding=1, norm=norm, act=act, preact=preact)
def forward(self, x):
return self.conv(x)
class BasicBlock(nn.Module):
def __init__(self, in_ch, out_ch, stride=1, norm=nn.BatchNorm2d, act=nn.ReLU, preact=True):
super().__init__()
assert norm in [nn.BatchNorm2d, nn.InstanceNorm2d, True, False]
assert act in [nn.ReLU, nn.ReLU6, nn.GELU, nn.SiLU, True, False]
self.conv1 = ConvNormAct(in_ch, out_ch, 3, stride=stride, padding=1, norm=norm, act=act, preact=preact)
self.conv2 = ConvNormAct(out_ch, out_ch, 3, stride=1, padding=1, norm=norm, act=act, preact=preact)
self.shortcut = nn.Sequential()
if stride != 1 or in_ch != out_ch: # 如果不相等就通过一层conv将残差改变
self.shortcut = ConvNormAct(in_ch, out_ch, 3, stride=stride, padding=1, norm=norm, act=act, preact=preact)
def forward(self, x):
residual = x
out = self.conv1(x)
out = self.conv2(out)
out += self.shortcut(residual)
return out
class Bottleneck(nn.Module):
def __init__(self, in_ch, out_ch, stride=1, groups=1, dilation=1, norm=nn.BatchNorm2d, act=nn.ReLU, preact=True):
super().__init__()
assert norm in [nn.BatchNorm2d, nn.InstanceNorm2d, True, False]
assert act in [nn.ReLU, nn.ReLU6, nn.GELU, nn.SiLU, True, False]
self.expansion = 4
self.conv1 = ConvNormAct(in_ch, out_ch//self.expansion, 1, stride=1, padding=0, norm=norm, act=act, preact=preact)
self.conv2 = ConvNormAct(out_ch//self.expansion, out_ch//self.expansion, 3, stride=stride, padding=1, norm=norm, act=act, groups=groups, dilation=dilation, preact=preact)
self.conv3 = ConvNormAct(out_ch//self.expansion, out_ch, 1, stride=1, padding=0, norm=norm, act=act, preact=preact)
self.shortcut = nn.Sequential()
if stride != 1 or in_ch != out_ch:
self.shortcut = ConvNormAct(in_ch, out_ch, 3, stride=stride, padding=1, norm=norm, act=act, preact=preact)
def forward(self, x):
residual = x
out = self.conv1(x)
out = self.conv2(out)
out = self.conv3(out)
out += self.shortcut(residual)
return out
class DepthwiseSeparableConv(nn.Module):
def __init__(self, in_ch, out_ch, stride=1, kernel_size=3, padding=1, bias=False):
super().__init__()
self.depthwise = nn.Conv2d(
in_channels=in_ch,
out_channels=in_ch,
kernel_size=kernel_size,
stride=stride,
padding=padding,
groups=in_ch,
bias=bias
)
self.pointwise = nn.Conv2d(
in_channels=in_ch,
out_channels=out_ch,
kernel_size=1,
stride=1,
padding=0,
groups=1,
bias=bias
)
def forward(self, x):
out = self.depthwise(x)
out = self.pointwise(out)
return out
class SEBlock(nn.Module):
def __init__(self, in_ch, ratio=4, act=nn.ReLU):
super().__init__()
self.squeeze = nn.AdaptiveAvgPool2d(1)
self.excitation = nn.Sequential(
nn.Conv2d(in_ch, in_ch//ratio, kernel_size=1),
act(),
nn.Conv2d(in_ch//ratio, in_ch, kernel_size=1),
nn.Sigmoid()
)
def forward(self, x):
out = self.squeeze(x)
out = self.excitation(out)
return x * out
class DropPath(nn.Module):
"""
Drop connection with pobability p
"""
def __init__(self, p=0):
super().__init__()
self.p = p
def forward(self, x):
if (not self.p) or (not self.training):
return x
batch_size = x.shape[0]
random_tensor = torch.rand(batch_size, 1, 1, 1).to(x.device)
binary_mask = self.p < random_tensor
x = x.div(1 - self.p)
x = x * binary_mask
return x
class MBConv(nn.Module):
"""
MBConv with an expansion factor of N, and squeeze-and-excitation module
"""
def __init__(self, in_ch, out_ch, expansion=4, kernel_size=3, stride=1, ratio=4, p=0, se=True, norm=nn.BatchNorm2d, act=nn.ReLU):
super().__init__()
padding = (kernel_size - 1) // 2
expanded = expansion * in_ch
self.se = se
self.expand_proj = nn.Identity() if (expansion==1) else ConvNormAct(in_ch, expanded, kernel_size=1, norm=norm, act=act, preact=True)
self.depthwise = ConvNormAct(expanded, expanded, kernel_size=kernel_size, stride=stride, padding=padding, groups=expanded, act=act, norm=norm, preact=True)
if self.se:
self.se_block = SEBlock(expanded, ratio=ratio)
self.pointwise = ConvNormAct(expanded, out_ch, kernel_size=1, padding=0, norm=norm, act=False, preact=True)
self.drop_path = DropPath(p)
self.shortcut = nn.Sequential()
if in_ch != out_ch or stride != 1:
self.shortcut = nn.Sequential(ConvNormAct(in_ch, out_ch, kernel_size, stride=stride, padding=padding, norm=False, act=False))
def forward(self, x):
residual = x
x = self.expand_proj(x)
x = self.depthwise(x)
if self.se:
x = self.se_block(x)
x = self.pointwise(x)
x = self.drop_path(x)
x = x + self.shortcut(residual)
return x
class FusedMBConv(nn.Module):
"""
MBConv with an expansion factor of N, and squeeze-and-excitation module
"""
def __init__(self, in_ch, out_ch, expansion=4, kernel_size=3, stride=1, ratio=4, p=0, se=True, norm=nn.BatchNorm2d, act=nn.ReLU):
super().__init__()
padding = (kernel_size - 1) // 2
expanded = expansion * in_ch
self.stride = stride
self.se = se
self.conv3x3 = ConvNormAct(in_ch, expanded, kernel_size=kernel_size, stride=stride, padding=padding, groups=1, norm=norm, act=act, preact=True)
if self.se:
self.se_block = SEBlock(expanded, ratio=ratio)
self.pointwise = ConvNormAct(expanded, out_ch, kernel_size=1, padding=0, norm=norm, act=False, preact=True)
self.drop_path = DropPath(p)
self.shortcut = nn.Sequential()
if in_ch != out_ch or stride != 1:
self.shortcut = nn.Sequential(ConvNormAct(in_ch, out_ch, 3, stride=stride, padding=1, norm=False, act=False))
def forward(self, x):
residual = x
x = self.conv3x3(x)
if self.se:
x = self.se_block(x)
x = self.pointwise(x)
x = self.drop_path(x)
x = x + self.shortcut(residual)
return x
class ConvNeXtBlock(nn.Module):
r""" ConvNeXt Block. There are two equivalent implementations:
(1) DwConv -> LayerNorm (channels_first) -> 1x1 Conv -> GELU -> 1x1 Conv; all in (N, C, H, W)
(2) DwConv -> Permute to (N, H, W, C); LayerNorm (channels_last) -> Linear -> GELU -> Linear; Permute back
We use (2) as we find it slightly faster in PyTorch
Args:
dim (int): Number of input channels.
drop_path (float): Stochastic depth rate. Default: 0.0
layer_scale_init_value (float): Init value for Layer Scale. Default: 1e-6.
"""
def __init__(self, dim, out_ch, stride=1, kernel_size=7, norm=None, act=None, preact=None, drop_path=0., layer_scale_init_value=1e-6):
super().__init__()
padding = kernel_size // 2
self.dwconv = nn.Conv2d(dim, dim, kernel_size=kernel_size, padding=padding, groups=dim) # depthwise conv
self.norm = LayerNorm(dim, eps=1e-6)
self.pwconv1 = nn.Linear(dim, 4 * dim) # pointwise/1x1 convs, implemented with linear layers
self.act = nn.GELU()
self.pwconv2 = nn.Linear(4 * dim, dim)
self.gamma = nn.Parameter(layer_scale_init_value * torch.ones((dim)),
requires_grad=True) if layer_scale_init_value > 0 else None
self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity()
def forward(self, x):
input = x
x = self.dwconv(x)
x = x.permute(0, 2, 3, 1) # (N, C, H, W) -> (N, H, W, C)
x = self.norm(x)
x = self.pwconv1(x)
x = self.act(x)
x = self.pwconv2(x)
if self.gamma is not None:
x = self.gamma * x
x = x.permute(0, 3, 1, 2) # (N, H, W, C) -> (N, C, H, W)
x = input + self.drop_path(x)
return x
class LayerNorm(nn.Module):
r""" LayerNorm that supports two data formats: channels_last (default) or channels_first.
The ordering of the dimensions in the inputs. channels_last corresponds to inputs with
shape (batch_size, height, width, channels) while channels_first corresponds to inputs
with shape (batch_size, channels, height, width).
"""
def __init__(self, normalized_shape, eps=1e-6, data_format="channels_last"):
super().__init__()
self.weight = nn.Parameter(torch.ones(normalized_shape))
self.bias = nn.Parameter(torch.zeros(normalized_shape))
self.eps = eps
self.data_format = data_format
if self.data_format not in ["channels_last", "channels_first"]:
raise NotImplementedError
self.normalized_shape = (normalized_shape, )
def forward(self, x):
if self.data_format == "channels_last":
return F.layer_norm(x, self.normalized_shape, self.weight, self.bias, self.eps)
elif self.data_format == "channels_first":
u = x.mean(1, keepdim=True)
s = (x - u).pow(2).mean(1, keepdim=True)
x = (x - u) / torch.sqrt(s + self.eps)
x = self.weight[:, None, None] * x + self.bias[:, None, None]
return x
if __name__ == '__main__':
img = torch.randn(2, 3, 256, 256)
depth_conv = DepthwiseSeparableConv(3, 32)
out = depth_conv(img)
print(out.shape)
|