File size: 14,088 Bytes
3c8ff2e |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 |
import torch
import torch.nn as nn
import torch.nn.functional as F
from functools import partial
import math
from abc import abstractmethod
class EmbedBlock(nn.Module):
"""
Any module where forward() takes embeddings as a second argument.
"""
@abstractmethod
def forward(self, x, emb):
"""
Apply the module to `x` given `emb` embeddings.
"""
class EmbedSequential(nn.Sequential, EmbedBlock):
"""
A sequential module that passes embeddings to the children that
support it as an extra input.
"""
def forward(self, x, emb):
for layer in self:
if isinstance(layer, EmbedBlock):
x = layer(x, emb)
else:
x = layer(x)
return x
def gamma_embedding(gammas, dim, max_period=10000):
"""
Create sinusoidal timestep embeddings.
:param gammas: a 1-D Tensor of N indices, one per batch element.
These may be fractional.
:param dim: the dimension of the output.
:param max_period: controls the minimum frequency of the embeddings.
:return: an [N x dim] Tensor of positional embeddings.
"""
half = dim // 2
freqs = torch.exp(
-math.log(max_period) * torch.arange(start=0,
end=half, dtype=torch.float32) / half
).to(device=gammas.device)
args = gammas[:, None].float() * freqs[None]
embedding = torch.cat([torch.cos(args), torch.sin(args)], dim=-1)
if dim % 2:
embedding = torch.cat(
[embedding, torch.zeros_like(embedding[:, :1])], dim=-1)
return embedding
class LayerNormFunction(torch.autograd.Function):
@staticmethod
def forward(ctx, x, weight, bias, eps):
ctx.eps = eps
N, C, H, W = x.size()
mu = x.mean(1, keepdim=True)
var = (x - mu).pow(2).mean(1, keepdim=True)
y = (x - mu) / (var + eps).sqrt()
ctx.save_for_backward(y, var, weight)
y = weight.view(1, C, 1, 1) * y + bias.view(1, C, 1, 1)
return y
@staticmethod
def backward(ctx, grad_output):
eps = ctx.eps
N, C, H, W = grad_output.size()
y, var, weight = ctx.saved_variables
g = grad_output * weight.view(1, C, 1, 1)
mean_g = g.mean(dim=1, keepdim=True)
mean_gy = (g * y).mean(dim=1, keepdim=True)
gx = 1. / torch.sqrt(var + eps) * (g - y * mean_gy - mean_g)
return gx, (grad_output * y).sum(dim=3).sum(dim=2).sum(dim=0), grad_output.sum(dim=3).sum(dim=2).sum(
dim=0), None
class LayerNorm2d(nn.Module):
def __init__(self, channels, eps=1e-6):
super(LayerNorm2d, self).__init__()
self.register_parameter('weight', nn.Parameter(torch.ones(channels)))
self.register_parameter('bias', nn.Parameter(torch.zeros(channels)))
self.eps = eps
def forward(self, x):
return LayerNormFunction.apply(x, self.weight, self.bias, self.eps)
class SimpleGate(nn.Module):
def forward(self, x):
x1, x2 = x.chunk(2, dim=1)
return x1 * x2
class CondNAFBlock(nn.Module):
def __init__(self, c, DW_Expand=2, FFN_Expand=2, drop_out_rate=0.):
super().__init__()
dw_channel = c * DW_Expand
self.conv1 = nn.Conv2d(in_channels=c, out_channels=dw_channel,
kernel_size=1, padding=0, stride=1, groups=1, bias=True)
self.conv2 = nn.Conv2d(in_channels=dw_channel, out_channels=dw_channel, kernel_size=3, padding=1, stride=1, groups=dw_channel,
bias=True)
self.conv3 = nn.Conv2d(in_channels=dw_channel // 2, out_channels=c,
kernel_size=1, padding=0, stride=1, groups=1, bias=True)
# Simplified Channel Attention
# self.sca = nn.Sequential(
# nn.AdaptiveAvgPool2d(1),
# nn.Conv2d(in_channels=dw_channel // 2, out_channels=dw_channel // 2, kernel_size=1, padding=0, stride=1,
# groups=1, bias=True),
# )
self.sca_avg = nn.Sequential(
nn.AdaptiveAvgPool2d(1),
nn.Conv2d(in_channels=dw_channel // 4, out_channels=dw_channel // 4, kernel_size=1, padding=0, stride=1,
groups=1, bias=True),
)
self.sca_max = nn.Sequential(
nn.AdaptiveMaxPool2d(1),
nn.Conv2d(in_channels=dw_channel // 4, out_channels=dw_channel // 4, kernel_size=1, padding=0, stride=1,
groups=1, bias=True),
)
# SimpleGate
self.sg = SimpleGate()
ffn_channel = FFN_Expand * c
self.conv4 = nn.Conv2d(in_channels=c, out_channels=ffn_channel,
kernel_size=1, padding=0, stride=1, groups=1, bias=True)
self.conv5 = nn.Conv2d(in_channels=ffn_channel // 2, out_channels=c,
kernel_size=1, padding=0, stride=1, groups=1, bias=True)
self.norm1 = LayerNorm2d(c)
self.norm2 = LayerNorm2d(c)
self.dropout1 = nn.Dropout(
drop_out_rate) if drop_out_rate > 0. else nn.Identity()
self.dropout2 = nn.Dropout(
drop_out_rate) if drop_out_rate > 0. else nn.Identity()
self.beta = nn.Parameter(torch.zeros((1, c, 1, 1)), requires_grad=True)
self.gamma = nn.Parameter(torch.zeros(
(1, c, 1, 1)), requires_grad=True)
def forward(self, inp):
x = inp
x = self.norm1(x)
x = self.conv1(x)
x = self.conv2(x)
x = self.sg(x)
x_avg, x_max = x.chunk(2, dim=1)
x_avg = self.sca_avg(x_avg)*x_avg
x_max = self.sca_max(x_max)*x_max
x = torch.cat([x_avg, x_max], dim=1)
x = self.conv3(x)
x = self.dropout1(x)
y = inp + x * self.beta
x = self.conv4(self.norm2(y))
x = self.sg(x)
x = self.conv5(x)
x = self.dropout2(x)
return y + x * self.gamma
class NAFBlock(nn.Module):
def __init__(self, c, DW_Expand=2, FFN_Expand=2, drop_out_rate=0.):
super().__init__()
dw_channel = c * DW_Expand
self.conv1 = nn.Conv2d(in_channels=c, out_channels=dw_channel,
kernel_size=1, padding=0, stride=1, groups=1, bias=True)
self.conv2 = nn.Conv2d(in_channels=dw_channel, out_channels=dw_channel, kernel_size=3, padding=1, stride=1, groups=dw_channel,
bias=True)
self.conv3 = nn.Conv2d(in_channels=dw_channel // 2, out_channels=c,
kernel_size=1, padding=0, stride=1, groups=1, bias=True)
# Simplified Channel Attention
# self.sca = nn.Sequential(
# nn.AdaptiveAvgPool2d(1),
# nn.Conv2d(in_channels=dw_channel // 2, out_channels=dw_channel // 2, kernel_size=1, padding=0, stride=1,
# groups=1, bias=True),
# )
self.sca_avg = nn.Sequential(
nn.AdaptiveAvgPool2d(1),
nn.Conv2d(in_channels=dw_channel // 4, out_channels=dw_channel // 4, kernel_size=1, padding=0, stride=1,
groups=1, bias=True),
)
self.sca_max = nn.Sequential(
nn.AdaptiveMaxPool2d(1),
nn.Conv2d(in_channels=dw_channel // 4, out_channels=dw_channel // 4, kernel_size=1, padding=0, stride=1,
groups=1, bias=True),
)
# SimpleGate
self.sg = SimpleGate()
ffn_channel = FFN_Expand * c
self.conv4 = nn.Conv2d(in_channels=c, out_channels=ffn_channel,
kernel_size=1, padding=0, stride=1, groups=1, bias=True)
self.conv5 = nn.Conv2d(in_channels=ffn_channel // 2, out_channels=c,
kernel_size=1, padding=0, stride=1, groups=1, bias=True)
self.norm1 = LayerNorm2d(c)
self.norm2 = LayerNorm2d(c)
self.dropout1 = nn.Dropout(
drop_out_rate) if drop_out_rate > 0. else nn.Identity()
self.dropout2 = nn.Dropout(
drop_out_rate) if drop_out_rate > 0. else nn.Identity()
self.beta = nn.Parameter(torch.zeros((1, c, 1, 1)), requires_grad=True)
self.gamma = nn.Parameter(torch.zeros(
(1, c, 1, 1)), requires_grad=True)
# self.time_emb = nn.Sequential(
# nn.SiLU(),
# nn.Linear(256, c),
# )
def forward(self, inp):
x = inp
x = self.norm1(x)
x = self.conv1(x)
x = self.conv2(x)
x = self.sg(x)
x_avg, x_max = x.chunk(2, dim=1)
x_avg = self.sca_avg(x_avg)*x_avg
x_max = self.sca_max(x_max)*x_max
x = torch.cat([x_avg, x_max], dim=1)
x = self.conv3(x)
x = self.dropout1(x)
y = inp + x * self.beta
# y = y+self.time_emb(t)[..., None, None]
x = self.conv4(self.norm2(y))
x = self.sg(x)
x = self.conv5(x)
x = self.dropout2(x)
return y + x * self.gamma
class UNCRTAINTS(nn.Module):
def __init__(
self,
input_dim=15,
out_conv=[13],
width=64,
middle_blk_num=1,
enc_blk_nums=[1, 1, 1, 1],
dec_blk_nums=[1, 1, 1, 1],
encoder_widths=[128],
decoder_widths=[128,128,128,128,128],
out_nonlin_mean=False,
out_nonlin_var='relu',
agg_mode="att_group",
encoder_norm="group",
decoder_norm="batch",
n_head=16,
d_model=256,
d_k=4,
pad_value=0,
padding_mode="reflect",
positional_encoding=True,
covmode='diag',
scale_by=1,
separate_out=False,
use_v=False,
block_type='mbconv',
is_mono=False
):
super().__init__()
self.intro = nn.Conv2d(in_channels=input_dim, out_channels=width, kernel_size=3, padding=1, stride=1, groups=1,
bias=True)
# self.cond_intro = nn.Conv2d(in_channels=img_channel+2, out_channels=width, kernel_size=3, padding=1, stride=1, groups=1,
# bias=True)
self.ending = nn.Conv2d(in_channels=width, out_channels=out_conv[0], kernel_size=3, padding=1, stride=1, groups=1,
bias=True)
# self.inp_ending = nn.Conv2d(in_channels=img_channel, out_channels=3, kernel_size=3, padding=1, stride=1, groups=1,
# bias=True)
self.encoders = nn.ModuleList()
self.cond_encoders = nn.ModuleList()
self.decoders = nn.ModuleList()
self.middle_blks = nn.ModuleList()
self.ups = nn.ModuleList()
self.downs = nn.ModuleList()
self.cond_downs = nn.ModuleList()
chan = width
for num in enc_blk_nums:
self.encoders.append(
nn.Sequential(
*[NAFBlock(chan) for _ in range(num)]
)
)
self.cond_encoders.append(
nn.Sequential(
*[CondNAFBlock(chan) for _ in range(num)]
)
)
self.downs.append(
nn.Conv2d(chan, 2*chan, 2, 2)
)
self.cond_downs.append(
nn.Conv2d(chan, 2*chan, 2, 2)
)
chan = chan * 2
self.middle_blks = \
nn.Sequential(
*[NAFBlock(chan) for _ in range(middle_blk_num)]
)
for num in dec_blk_nums:
self.ups.append(
nn.Sequential(
nn.Conv2d(chan, chan * 2, 1, bias=False),
nn.PixelShuffle(2)
)
)
chan = chan // 2
self.decoders.append(
nn.Sequential(
*[NAFBlock(chan) for _ in range(num)]
)
)
self.padder_size = 2 ** len(self.encoders)
self.map = nn.Sequential(
nn.Linear(64, 256),
nn.SiLU(),
nn.Linear(256, 256),
)
def forward(self, inp):
inp = self.check_image_size(inp)
x = self.intro(inp)
encs = []
for encoder, down in zip(self.encoders, self.downs):
x = encoder(x)
# b, c, h, w = cond.shape
# tmp_cond = cond.view(b//3, 3, c, h, w).sum(dim=1)
# tmp_cond = cond
# x = x + tmp_cond
encs.append(x)
x = down(x)
# cond = cond_down(cond)
x = self.middle_blks(x)
for decoder, up, enc_skip in zip(self.decoders, self.ups, encs[::-1]):
x = up(x)
x = x + enc_skip
x = decoder(x)
x = self.ending(x)
# x = x + self.inp_ending(inp)
return x
def check_image_size(self, x):
_, _, h, w = x.size()
mod_pad_h = (self.padder_size - h %
self.padder_size) % self.padder_size
mod_pad_w = (self.padder_size - w %
self.padder_size) % self.padder_size
x = F.pad(x, (0, mod_pad_w, 0, mod_pad_h))
return x
if __name__ == '__main__':
# unit test for ground resolution
inp = torch.randn(1, 15, 256, 256)
net = UNCRTAINTS(
input_dim=15,
out_conv=[13],
width=64,
middle_blk_num=1,
enc_blk_nums=[1, 1, 1, 1],
dec_blk_nums=[1, 1, 1, 1],
)
out = net(inp)
assert out.shape == (1, 13, 256, 256)
# from thop import profile
# out_shape = (1, 12, 384, 384)
# input_shape = (1, 13, 384, 384)
# model = DiffCR(
# img_channel=13,
# width=32,
# middle_blk_num=1,
# enc_blk_nums=[1, 1, 1, 1],
# dec_blk_nums=[1, 1, 1, 1],
# )
# # 使用 thop 的 profile 函数来获取 FLOPs 和参数量
# flops, params = profile(model, inputs=(torch.randn(out_shape), torch.ones(1,), torch.randn(input_shape)))
# print(f"FLOPs: {flops / 1e9} G")
# print(f"Parameters: {params / 1e6} M")
|