ResNet

概念


吴恩达老师讲的ResNet理论,通俗易懂(吴恩达老师yyds)
传送门

代码

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
import torch
from torch import nn
from torch.nn import functional as F
from d2l import torch as d2l


class Residual(nn.Module): # @save
def __init__(self, input_channels, num_channels,
use_1x1conv=False, strides=1):
super().__init__()
# 定义两层卷积
self.conv1 = nn.Conv2d(input_channels, num_channels,
kernel_size=3, padding=1, stride=strides)
self.conv2 = nn.Conv2d(num_channels, num_channels,
kernel_size=3, padding=1)
if use_1x1conv: # 如果要使用1x1卷积网络
self.conv3 = nn.Conv2d(input_channels, num_channels,
kernel_size=1, stride=strides)
else:
self.conv3 = None
self.bn1 = nn.BatchNorm2d(num_channels)
self.bn2 = nn.BatchNorm2d(num_channels)

def forward(self, X):
Y = F.relu(self.bn1(self.conv1(X)))
Y = self.bn2(self.conv2(Y))
# 在第二个层激活之前,加上第一层的初始数据,实现残差
if self.conv3:
X = self.conv3(X)
Y += X
return F.relu(Y)


ResNet实现

ResNet则使用4个由残差块组成的模块,每个模块使用若干个同样输出通道数的残差块。 第一个模块的通道数同输入通道数一致。之后的每个模块在第一个残差块里将上一个模块的通道数翻倍,并将高和宽减半。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
# 模块
b1 = nn.Sequential(nn.Conv2d(1, 64, kernel_size=7, stride=2, padding=3), nn.BatchNorm2d(
64), nn.ReLU(), nn.MaxPool2d(kernel_size=3, stride=2, padding=1))


# 残差块
def resnet_block(input_channels, num_channels, num_residuals, first_block=False):
blk = []
for i in range(num_residuals):
if i == 0 and not first_block: # 不是第一个
blk.append(Residual(input_channels, num_channels,
use_1x1conv=True, strides=2))
else:
blk.append(Residual(num_channels, num_channels))

return blk

1
2
3
4
5
b2 = nn.Sequential(*resnet_block(64, 64, 2, first_block=True))
b3 = nn.Sequential(*resnet_block(64, 128, 2))
b4 = nn.Sequential(*resnet_block(128, 256, 2))
b5 = nn.Sequential(*resnet_block(256, 512, 2))

封装一个整的网络

1
2
net = nn.Sequential(b1, b2, b3, b4, b5, nn.AdaptiveAvgPool2d(
(1, 1)), nn.Flatten(), nn.Linear(512, 10))

打印一下每一层的信息

1
2
3
4
X = torch.rand(size=(1, 1, 224, 224))
for layer in net:
X = layer(X)
print(layer.__class__.__name__,'output shape:\t', X.shape)
Sequential output shape:     torch.Size([1, 64, 56, 56])
Sequential output shape:     torch.Size([1, 64, 56, 56])
Sequential output shape:     torch.Size([1, 128, 28, 28])
Sequential output shape:     torch.Size([1, 256, 14, 14])
Sequential output shape:     torch.Size([1, 512, 7, 7])
AdaptiveAvgPool2d output shape:     torch.Size([1, 512, 1, 1])
Flatten output shape:     torch.Size([1, 512])
Linear output shape:     torch.Size([1, 10])