FPN(feature pyramid networks)特征金字塔网络
说明:这部分代码及截图是从psenet中截取出来的
自底向上+自顶向下+横向连接
自底向上
上图左,用resnet提取多层特征
自顶向下
上图右,将高层特征图上采样后与低层特征图融合
横向连接
从上图左到上图右,需要将特征图经过卷积减少通道数
add与cat的区别
https://blog.csdn.net/weixin_42926076/article/details/100660188
class ResNet(nn.Module):
def __init__(self, block, layers, num_classes=7, scale=1): # block: BasicBlock、Bottleneck
self.inplanes = 64
super(ResNet, self).__init__()
# 去掉了全连接层的resnet
self.conv1 = nn.Conv2d(3, 64, kernel_size=7, stride=2, padding=3,bias=False) # 第一个卷积层将输入3通道图变为64通道
self.bn1 = nn.BatchNorm2d(64)
self.relu1 = nn.ReLU(inplace=True)
self.maxpool = nn.MaxPool2d(kernel_size=3, stride=2, padding=1)
self.layer1 = self._make_layer(block, 64, layers[0])
self.layer2 = self._make_layer(block, 128, layers[1], stride=2)
self.layer3 = self._make_layer(block, 256, layers[2], stride=2)
self.layer4 = self._make_layer(block, 512, layers[3], stride=2)
# self.avgpool = nn.AvgPool2d(7, stride=1)
# self.fc = nn.Linear(512 * block.expansion, num_classes)
# Top layer
# 用于获得P5
self.toplayer = nn.Conv2d(2048, 256, kernel_size=1, stride=1, padding=0) # Reduce channels
self.toplayer_bn = nn.BatchNorm2d(256)
self.toplayer_relu = nn.ReLU(inplace=True)
# Smooth layers
# 可以看出这三个smooth层都是一样的,自顶向下上采样相加后融合特征,用于获得P4\P3\P2
self.smooth1 = nn.Conv2d(256, 256, kernel_size=3, stride=1, padding=1)
self.smooth1_bn = nn.BatchNorm2d(256)
self.smooth1_relu = nn.ReLU(inplace=True)
self.smooth2 = nn.Conv2d(256, 256, kernel_size=3, stride=1, padding=1)
self.smooth2_bn = nn.BatchNorm2d(256)
self.smooth2_relu = nn.ReLU(inplace=True)
self.smooth3 = nn.Conv2d(256, 256, kernel_size=3, stride=1, padding=1)
self.smooth3_bn = nn.BatchNorm2d(256)
self.smooth3_relu = nn.ReLU(inplace=True)
# Lateral layers
# 横向连接,将resnet输出的特征图通过卷积减少通道数,统一处理为256维
self.latlayer1 = nn.Conv2d(1024, 256, kernel_size=1, stride=1, padding=0)
self.latlayer1_bn = nn.BatchNorm2d(256)
self.latlayer1_relu = nn.ReLU(inplace=True)
self.latlayer2 = nn.Conv2d(512, 256, kernel_size=1, stride=1, padding=0)
self.latlayer2_bn = nn.BatchNorm2d(256)
self.latlayer2_relu = nn.ReLU(inplace=True)
self.latlayer3 = nn.Conv2d(256, 256, kernel_size=1, stride=1, padding=0)
self.latlayer3_bn = nn.BatchNorm2d(256)
self.latlayer3_relu = nn.ReLU(inplace=True)
# 将P5\P4\P3\P2 4个256通道的特征图上采样得到的1024通道特征图降低通道数量为256
self.conv2 = nn.Conv2d(1024, 256, kernel_size=3, stride=1, padding=1)
self.bn2 = nn.BatchNorm2d(256)
self.relu2 = nn.ReLU(inplace=True)
self.conv3 = nn.Conv2d(256, num_classes, kernel_size=1, stride=1, padding=0)
self.scale = scale
for m in self.modules():
if isinstance(m, nn.Conv2d):
n = m.kernel_size[0] * m.kernel_size[1] * m.out_channels
m.weight.data.normal_(0, math.sqrt(2. / n))
elif isinstance(m, nn.BatchNorm2d):
m.weight.data.fill_(1)
m.bias.data.zero_()
def _make_layer(self, block, planes, blocks, stride=1):
downsample = None
if stride != 1 or self.inplanes != planes * block.expansion:
downsample = nn.Sequential(
nn.Conv2d(self.inplanes, planes * block.expansion,
kernel_size=1, stride=stride, bias=False),
nn.BatchNorm2d(planes * block.expansion),
)
layers = []
layers.append(block(self.inplanes, planes, stride, downsample))
self.inplanes = planes * block.expansion
for i in range(1, blocks):
layers.append(block(self.inplanes, planes))
return nn.Sequential(*layers)
def _upsample(self, x, y, scale=1):
_, _, H, W = y.size()
return F.upsample(x, size=(H // scale, W // scale), mode='bilinear')
def _upsample_add(self, x, y):
_, _, H, W = y.size()
return F.upsample(x, size=(H, W), mode='bilinear') + y
def forward(self, x):
# resnet down-Top
h = x
h = self.conv1(h)
h = self.bn1(h)
h = self.relu1(h)
h = self.maxpool(h)
h = self.layer1(h)
c2 = h
h = self.layer2(h)
c3 = h
h = self.layer3(h)
c4 = h
h = self.layer4(h)
c5 = h
# Top-down
p5 = self.toplayer(c5) # 2048->256
p5 = self.toplayer_relu(self.toplayer_bn(p5))
c4 = self.latlayer1(c4) # 1024->256
c4 = self.latlayer1_relu(self.latlayer1_bn(c4))
p4 = self._upsample_add(p5, c4)
p4 = self.smooth1(p4) # 通过卷积进行特征融合
p4 = self.smooth1_relu(self.smooth1_bn(p4))
c3 = self.latlayer2(c3)
c3 = self.latlayer2_relu(self.latlayer2_bn(c3))
p3 = self._upsample_add(p4, c3)
p3 = self.smooth2(p3)
p3 = self.smooth2_relu(self.smooth2_bn(p3))
c2 = self.latlayer3(c2)
c2 = self.latlayer3_relu(self.latlayer3_bn(c2))
p2 = self._upsample_add(p3, c2)
p2 = self.smooth3(p2)
p2 = self.smooth3_relu(self.smooth3_bn(p2))
p3 = self._upsample(p3, p2)
p4 = self._upsample(p4, p2)
p5 = self._upsample(p5, p2)
out = torch.cat((p2, p3, p4, p5), 1) # 256*4 ->1024
out = self.conv2(out) # 1024 ->256
out = self.relu2(self.bn2(out))
out = self.conv3(out)
out = self._upsample(out, x, scale=self.scale) # 处理为和输入同样的尺寸
return out