File size: 20,168 Bytes
f348660 | 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 | # Copyright (C) 2019 Jin Han Lee
#
# This file is a part of BTS.
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program. If not, see <http://www.gnu.org/licenses/>
import torch
import torch.nn as nn
import torch.nn.functional as torch_nn_func
import math
def bn_init_as_tf(m):
if isinstance(m, nn.BatchNorm2d):
m.track_running_stats = True # These two lines enable using stats (moving mean and var) loaded from pretrained model
m.eval() # or zero mean and variance of one if the batch norm layer has no pretrained values
m.affine = True
m.requires_grad = True
def weights_init_xavier(m):
if isinstance(m, nn.Conv2d):
torch.nn.init.xavier_uniform_(m.weight)
if m.bias is not None:
torch.nn.init.zeros_(m.bias)
class atrous_conv(nn.Sequential):
def __init__(self, in_channels, out_channels, dilation, apply_bn_first=True):
super(atrous_conv, self).__init__()
self.atrous_conv = torch.nn.Sequential()
if apply_bn_first:
self.atrous_conv.add_module('first_bn', nn.BatchNorm2d(in_channels, momentum=0.01, affine=True, track_running_stats=True, eps=1.1e-5))
self.atrous_conv.add_module('aconv_sequence', nn.Sequential(nn.ReLU(),
nn.Conv2d(in_channels=in_channels, out_channels=out_channels*2, bias=False, kernel_size=1, stride=1, padding=0),
nn.BatchNorm2d(out_channels*2, momentum=0.01, affine=True, track_running_stats=True),
nn.ReLU(),
nn.Conv2d(in_channels=out_channels * 2, out_channels=out_channels, bias=False, kernel_size=3, stride=1,
padding=(dilation, dilation), dilation=dilation)))
def forward(self, x):
return self.atrous_conv.forward(x)
class upconv(nn.Module):
def __init__(self, in_channels, out_channels, ratio=2):
super(upconv, self).__init__()
self.elu = nn.ELU()
self.conv = nn.Conv2d(in_channels=in_channels, out_channels=out_channels, bias=False, kernel_size=3, stride=1, padding=1)
self.ratio = ratio
def forward(self, x):
up_x = torch_nn_func.interpolate(x, scale_factor=self.ratio, mode='nearest')
out = self.conv(up_x)
out = self.elu(out)
return out
class reduction_1x1(nn.Sequential):
def __init__(self, num_in_filters, num_out_filters, max_depth, is_final=False):
super(reduction_1x1, self).__init__()
self.max_depth = max_depth
self.is_final = is_final
self.sigmoid = nn.Sigmoid()
self.reduc = torch.nn.Sequential()
while num_out_filters >= 4:
if num_out_filters < 8:
if self.is_final:
self.reduc.add_module('final', torch.nn.Sequential(nn.Conv2d(num_in_filters, out_channels=1, bias=False,
kernel_size=1, stride=1, padding=0),
nn.Sigmoid()))
else:
self.reduc.add_module('plane_params', torch.nn.Conv2d(num_in_filters, out_channels=3, bias=False,
kernel_size=1, stride=1, padding=0))
break
else:
self.reduc.add_module('inter_{}_{}'.format(num_in_filters, num_out_filters),
torch.nn.Sequential(nn.Conv2d(in_channels=num_in_filters, out_channels=num_out_filters,
bias=False, kernel_size=1, stride=1, padding=0),
nn.ELU()))
num_in_filters = num_out_filters
num_out_filters = num_out_filters // 2
def forward(self, net):
net = self.reduc.forward(net)
if not self.is_final:
theta = self.sigmoid(net[:, 0, :, :]) * math.pi / 3
phi = self.sigmoid(net[:, 1, :, :]) * math.pi * 2
dist = self.sigmoid(net[:, 2, :, :]) * self.max_depth
n1 = torch.mul(torch.sin(theta), torch.cos(phi)).unsqueeze(1)
n2 = torch.mul(torch.sin(theta), torch.sin(phi)).unsqueeze(1)
n3 = torch.cos(theta).unsqueeze(1)
n4 = dist.unsqueeze(1)
net = torch.cat([n1, n2, n3, n4], dim=1)
return net
class local_planar_guidance(nn.Module):
def __init__(self, upratio):
super(local_planar_guidance, self).__init__()
self.upratio = upratio
self.u = torch.arange(self.upratio).reshape([1, 1, self.upratio]).float()
self.v = torch.arange(int(self.upratio)).reshape([1, self.upratio, 1]).float()
self.upratio = float(upratio)
def forward(self, plane_eq, focal):
plane_eq_expanded = torch.repeat_interleave(plane_eq, int(self.upratio), 2)
plane_eq_expanded = torch.repeat_interleave(plane_eq_expanded, int(self.upratio), 3)
n1 = plane_eq_expanded[:, 0, :, :]
n2 = plane_eq_expanded[:, 1, :, :]
n3 = plane_eq_expanded[:, 2, :, :]
n4 = plane_eq_expanded[:, 3, :, :]
u = self.u.repeat(plane_eq.size(0), plane_eq.size(2) * int(self.upratio), plane_eq.size(3)).cuda()
u = (u - (self.upratio - 1) * 0.5) / self.upratio
v = self.v.repeat(plane_eq.size(0), plane_eq.size(2), plane_eq.size(3) * int(self.upratio)).cuda()
v = (v - (self.upratio - 1) * 0.5) / self.upratio
return n4 / (n1 * u + n2 * v + n3)
class bts_gated_fuse(nn.Module):
def __init__(self, params, feat_out_channels, feat_out_channels_rad, num_features=512):
super(bts_gated_fuse, self).__init__()
self.params = params
self.weight5 = torch.nn.Sequential(nn.Conv2d(feat_out_channels_rad[4], feat_out_channels[4], 1, 1, bias=False),
nn.Sigmoid())
self.project5 = torch.nn.Sequential(nn.Conv2d(feat_out_channels_rad[4], feat_out_channels[4], 1, 1, bias=False),
nn.ReLU())
self.upconv5 = upconv(feat_out_channels[4], num_features)
self.bn5 = nn.BatchNorm2d(num_features, momentum=0.01, affine=True, eps=1.1e-5)
self.conv5 = torch.nn.Sequential(nn.Conv2d(num_features + feat_out_channels[3], num_features, 3, 1, 1, bias=False),
nn.ELU())
self.weight4 = torch.nn.Sequential(nn.Conv2d(feat_out_channels_rad[3], num_features, 1, 1, bias=False),
nn.Sigmoid())
self.project4 = torch.nn.Sequential(nn.Conv2d(feat_out_channels_rad[3], num_features, 1, 1, bias=False),
nn.ReLU())
self.upconv4 = upconv(num_features, num_features // 2)
self.bn4 = nn.BatchNorm2d(num_features // 2, momentum=0.01, affine=True, eps=1.1e-5)
self.conv4 = torch.nn.Sequential(nn.Conv2d(num_features // 2 + feat_out_channels[2], num_features // 2, 3, 1, 1, bias=False),
nn.ELU())
self.bn4_2 = nn.BatchNorm2d(num_features // 2, momentum=0.01, affine=True, eps=1.1e-5)
self.daspp_3 = atrous_conv(num_features // 2, num_features // 4, 3, apply_bn_first=False)
self.daspp_6 = atrous_conv(num_features // 2 + num_features // 4 + feat_out_channels[2], num_features // 4, 6)
self.daspp_12 = atrous_conv(num_features + feat_out_channels[2], num_features // 4, 12)
self.daspp_18 = atrous_conv(num_features + num_features // 4 + feat_out_channels[2], num_features // 4, 18)
self.daspp_24 = atrous_conv(num_features + num_features // 2 + feat_out_channels[2], num_features // 4, 24)
self.daspp_conv = torch.nn.Sequential(nn.Conv2d(num_features + num_features // 2 + num_features // 4, num_features // 4, 3, 1, 1, bias=False),
nn.ELU())
self.reduc8x8 = reduction_1x1(num_features // 4, num_features // 4, self.params.max_depth)
self.lpg8x8 = local_planar_guidance(8)
self.weight3 = torch.nn.Sequential(nn.Conv2d(feat_out_channels_rad[2], num_features // 4, 1, 1, bias=False),
nn.Sigmoid())
self.project3 = torch.nn.Sequential(nn.Conv2d(feat_out_channels_rad[2], num_features // 4, 1, 1, bias=False),
nn.ReLU())
self.upconv3 = upconv(num_features // 4, num_features // 4)
self.bn3 = nn.BatchNorm2d(num_features // 4, momentum=0.01, affine=True, eps=1.1e-5)
self.conv3 = torch.nn.Sequential(nn.Conv2d(num_features // 4 + feat_out_channels[1] + 1, num_features // 4, 3, 1, 1, bias=False),
nn.ELU())
self.reduc4x4 = reduction_1x1(num_features // 4, num_features // 8, self.params.max_depth)
self.lpg4x4 = local_planar_guidance(4)
self.weight2 = torch.nn.Sequential(nn.Conv2d(feat_out_channels_rad[1], num_features // 4, 1, 1, bias=False),
nn.Sigmoid())
self.project2 = torch.nn.Sequential(nn.Conv2d(feat_out_channels_rad[1], num_features // 4, 1, 1, bias=False),
nn.ReLU())
self.upconv2 = upconv(num_features // 4, num_features // 8)
self.bn2 = nn.BatchNorm2d(num_features // 8, momentum=0.01, affine=True, eps=1.1e-5)
self.conv2 = torch.nn.Sequential(nn.Conv2d(num_features // 8 + feat_out_channels[0] + 1, num_features // 8, 3, 1, 1, bias=False),
nn.ELU())
self.reduc2x2 = reduction_1x1(num_features // 8, num_features // 16, self.params.max_depth)
self.lpg2x2 = local_planar_guidance(2)
self.weight1 = torch.nn.Sequential(nn.Conv2d(feat_out_channels_rad[0], num_features // 8, 1, 1, bias=False),
nn.Sigmoid())
self.project1 = torch.nn.Sequential(nn.Conv2d(feat_out_channels_rad[0], num_features // 8, 1, 1, bias=False),
nn.ReLU())
self.upconv1 = upconv(num_features // 8, num_features // 16)
self.reduc1x1 = reduction_1x1(num_features // 16, num_features // 32, self.params.max_depth, is_final=True)
self.conv1 = torch.nn.Sequential(nn.Conv2d(num_features // 16 + 4, num_features // 16, 3, 1, 1, bias=False),
nn.ELU())
self.get_depth = torch.nn.Sequential(nn.Conv2d(num_features // 16, 1, 3, 1, 1, bias=False),
nn.Sigmoid())
self.pool5 = torch.nn.AvgPool2d(32, 32)
self.pool4 = torch.nn.AvgPool2d(16, 16)
self.pool3 = torch.nn.AvgPool2d(8, 8)
self.pool2 = torch.nn.AvgPool2d(4, 4)
self.pool1 = torch.nn.AvgPool2d(2, 2)
def forward(self, img_features, rad_features, focal, radar_confidence):
skip0, skip1, skip2, skip3 = img_features[0], img_features[1], img_features[2], img_features[3]
rad_skip0, rad_skip1, rad_skip2, rad_skip3 = rad_features[0], rad_features[1], rad_features[2], rad_features[3]
# prepare radar confidence
radar_confidence5 = self.pool5(radar_confidence)
radar_confidence4 = self.pool4(radar_confidence)
radar_confidence3 = self.pool3(radar_confidence)
radar_confidence2 = self.pool2(radar_confidence)
radar_confidence1 = self.pool1(radar_confidence)
rad_weight5 = self.weight5(rad_features[4])
rad_project5 = self.project5(rad_features[4])
dense_features = torch.nn.ReLU()(img_features[4])
dense_features = dense_features + rad_weight5*rad_project5*radar_confidence5
upconv5 = self.upconv5(dense_features) # H/16
upconv5 = self.bn5(upconv5)
concat5 = torch.cat([upconv5, skip3], dim=1)
iconv5 = self.conv5(concat5)
rad_weight4 = self.weight4(rad_skip3)
rad_project4 = self.project4(rad_skip3)
iconv5 = iconv5 + rad_weight4*rad_project4*radar_confidence4
upconv4 = self.upconv4(iconv5) # H/8
upconv4 = self.bn4(upconv4)
concat4 = torch.cat([upconv4, skip2], dim=1)
iconv4 = self.conv4(concat4)
iconv4 = self.bn4_2(iconv4)
daspp_3 = self.daspp_3(iconv4)
concat4_2 = torch.cat([concat4, daspp_3], dim=1)
daspp_6 = self.daspp_6(concat4_2)
concat4_3 = torch.cat([concat4_2, daspp_6], dim=1)
daspp_12 = self.daspp_12(concat4_3)
concat4_4 = torch.cat([concat4_3, daspp_12], dim=1)
daspp_18 = self.daspp_18(concat4_4)
concat4_5 = torch.cat([concat4_4, daspp_18], dim=1)
daspp_24 = self.daspp_24(concat4_5)
concat4_daspp = torch.cat([iconv4, daspp_3, daspp_6, daspp_12, daspp_18, daspp_24], dim=1)
daspp_feat = self.daspp_conv(concat4_daspp)
rad_weight3 = self.weight3(rad_skip2)
rad_project3 = self.project3(rad_skip2)
daspp_feat = daspp_feat + rad_weight3*rad_project3*radar_confidence3
reduc8x8 = self.reduc8x8(daspp_feat)
plane_normal_8x8 = reduc8x8[:, :3, :, :]
plane_normal_8x8 = torch_nn_func.normalize(plane_normal_8x8, 2, 1)
plane_dist_8x8 = reduc8x8[:, 3, :, :]
plane_eq_8x8 = torch.cat([plane_normal_8x8, plane_dist_8x8.unsqueeze(1)], 1)
depth_8x8 = self.lpg8x8(plane_eq_8x8, focal)
depth_8x8_scaled = depth_8x8.unsqueeze(1) / self.params.max_depth
depth_8x8_scaled_ds = torch_nn_func.interpolate(depth_8x8_scaled, scale_factor=0.25, mode='nearest')
upconv3 = self.upconv3(daspp_feat) # H/4
upconv3 = self.bn3(upconv3)
concat3 = torch.cat([upconv3, skip1, depth_8x8_scaled_ds], dim=1)
iconv3 = self.conv3(concat3)
rad_weight2 = self.weight2(rad_skip1)
rad_project2 = self.project2(rad_skip1)
iconv3 = iconv3 + rad_weight2*rad_project2*radar_confidence2
reduc4x4 = self.reduc4x4(iconv3)
plane_normal_4x4 = reduc4x4[:, :3, :, :]
plane_normal_4x4 = torch_nn_func.normalize(plane_normal_4x4, 2, 1)
plane_dist_4x4 = reduc4x4[:, 3, :, :]
plane_eq_4x4 = torch.cat([plane_normal_4x4, plane_dist_4x4.unsqueeze(1)], 1)
depth_4x4 = self.lpg4x4(plane_eq_4x4, focal)
depth_4x4_scaled = depth_4x4.unsqueeze(1) / self.params.max_depth
depth_4x4_scaled_ds = torch_nn_func.interpolate(depth_4x4_scaled, scale_factor=0.5, mode='nearest')
upconv2 = self.upconv2(iconv3) # H/2
upconv2 = self.bn2(upconv2)
concat2 = torch.cat([upconv2, skip0, depth_4x4_scaled_ds], dim=1)
iconv2 = self.conv2(concat2)
rad_weight1 = self.weight1(rad_skip0)
rad_project1 = self.project1(rad_skip0)
iconv2 = iconv2 + rad_weight1*rad_project1*radar_confidence1
reduc2x2 = self.reduc2x2(iconv2)
plane_normal_2x2 = reduc2x2[:, :3, :, :]
plane_normal_2x2 = torch_nn_func.normalize(plane_normal_2x2, 2, 1)
plane_dist_2x2 = reduc2x2[:, 3, :, :]
plane_eq_2x2 = torch.cat([plane_normal_2x2, plane_dist_2x2.unsqueeze(1)], 1)
depth_2x2 = self.lpg2x2(plane_eq_2x2, focal)
depth_2x2_scaled = depth_2x2.unsqueeze(1) / self.params.max_depth
rad_weight1 = self.weight1(rad_skip0)
rad_project1 = self.project1(rad_skip0)
upconv1 = self.upconv1(iconv2)
reduc1x1 = self.reduc1x1(upconv1)
concat1 = torch.cat([upconv1, reduc1x1, depth_2x2_scaled, depth_4x4_scaled, depth_8x8_scaled], dim=1)
iconv1 = self.conv1(concat1)
final_depth = self.params.max_depth * self.get_depth(iconv1)
return depth_8x8_scaled, depth_4x4_scaled, depth_2x2_scaled, reduc1x1, final_depth
class encoder_image(nn.Module):
def __init__(self, params):
super(encoder_image, self).__init__()
self.params = params
import torchvision.models as models
if params.encoder == 'densenet121_bts':
self.base_model = models.densenet121(pretrained=False).features
self.feat_names = ['relu0', 'pool0', 'transition1', 'transition2', 'norm5']
self.feat_out_channels = [64, 64, 128, 256, 1024]
elif params.encoder == 'densenet161_bts':
self.base_model = models.densenet161(pretrained=False).features
self.feat_names = ['relu0', 'pool0', 'transition1', 'transition2', 'norm5']
self.feat_out_channels = [96, 96, 192, 384, 2208]
elif params.encoder == 'resnet50_bts':
self.base_model = models.resnet50(pretrained=False)
self.feat_names = ['relu', 'layer1', 'layer2', 'layer3', 'layer4']
self.feat_out_channels = [64, 256, 512, 1024, 2048]
elif params.encoder == 'resnet34_bts':
self.base_model = models.resnet34(pretrained=False)
self.feat_names = ['relu', 'layer1', 'layer2', 'layer3', 'layer4']
self.feat_out_channels = [64, 64, 128, 256, 512]
elif params.encoder == 'resnet18_bts':
self.base_model = models.resnet18(pretrained=False)
self.feat_names = ['relu', 'layer1', 'layer2', 'layer3', 'layer4']
self.feat_out_channels = [64, 64, 128, 256, 512]
elif params.encoder == 'resnet101_bts':
self.base_model = models.resnet101(pretrained=False)
self.feat_names = ['relu', 'layer1', 'layer2', 'layer3', 'layer4']
self.feat_out_channels = [64, 256, 512, 1024, 2048]
elif params.encoder == 'resnext50_bts':
self.base_model = models.resnext50_32x4d(pretrained=False)
self.feat_names = ['relu', 'layer1', 'layer2', 'layer3', 'layer4']
self.feat_out_channels = [64, 256, 512, 1024, 2048]
elif params.encoder == 'resnext101_bts':
self.base_model = models.resnext101_32x8d(pretrained=False)
self.feat_names = ['relu', 'layer1', 'layer2', 'layer3', 'layer4']
self.feat_out_channels = [64, 256, 512, 1024, 2048]
elif params.encoder == 'mobilenetv2_bts':
self.base_model = models.mobilenet_v2(pretrained=False).features
self.feat_inds = [2, 4, 7, 11, 19]
self.feat_out_channels = [16, 24, 32, 64, 1280]
self.feat_names = []
else:
print('Not supported encoder: {}'.format(params.encoder))
def forward(self, x):
feature = x
skip_feat = []
i = 1
for k, v in self.base_model._modules.items():
if 'fc' in k or 'avgpool' in k:
continue
feature = v(feature)
if self.params.encoder == 'mobilenetv2_bts':
if i == 2 or i == 4 or i == 7 or i == 11 or i == 19:
skip_feat.append(feature)
else:
if any(x in k for x in self.feat_names):
skip_feat.append(feature)
i = i + 1
return skip_feat
|