pytorch(三)Alexnet网络

た入场券 2023-01-06 05:28 191阅读 0赞

网络结构

![format_png][]

各网络层特征图计算：

Alexnet:输入227*227
    conv1:    In-28*28 k:5  pd:2  s:1 -----out:(28+2*2-5)/1+1=28   6*28*28
    maxpool1: In-28*28 po:2 s:2       -----out:(28-2)/2+1=14   16*14*14
    conv2:    In-14*14 k:5    s:1 -----out:(14-5)/1+1=10   16*10*10
    maxpool2: In-10*10 po:2 s:2       -----out:(10-2)/2+1=5   16*5*5
    fc1:      In:5*5                 -----out:120
    fc2:      In:120                 -----out:84
    fc3:      In:84                 -----out:10

代码：

class Alexnet(nn.Module):
        def __init__(self):
            super().__init__()
            self.conv1=nn.Sequential(
                nn.Conv2d(3,96,11,4),
                nn.ReLU(),
                nn.MaxPool2d(3,2)
            )
            self.conv2 = nn.Sequential(
                nn.Conv2d(96, 256, 5,1,2),
                nn.ReLU(),
                nn.MaxPool2d(3, 2)
            )
            self.conv3 = nn.Sequential(
                nn.Conv2d(256, 384,3, 1, 1),
                nn.ReLU(),
            )
            self.conv4 = nn.Sequential(
                nn.Conv2d(384, 384, 3, 1, 1),
                nn.ReLU(),
            )
            self.conv5 = nn.Sequential(
                nn.Conv2d(384, 256, 3, 1, 1),
                nn.ReLU(),
                nn.MaxPool2d(3, 2)
            )
    
            self.fc1=nn.Sequential(
                nn.Linear(256*6*6,4096),
                nn.ReLU()
            )
            self.fc2 = nn.Sequential(
                nn.Linear(4096, 4096),
                nn.ReLU()
            )
            self.fc3=nn.Linear(4096,2)
    
        def forward(self,x):
            x=self.conv1(x)
            x=self.conv2(x)
            x = self.conv3(x)
            x = self.conv4(x)
            x = self.conv5(x)
            print(x.shape)#torch.Size([32, 256, 6, 6])
            x=x.view(x.size()[0],-1)
            print(x.shape)#torch.Size([32, 9216])
            x=self.fc1(x)
            x=self.fc2(x)
            x=self.fc3(x)
    
            return x

[训练测试同上篇。][Link 1]

[format_png]: /images/20221119/630a61cbc4d04d5c9c4a1d46892bbb9b.png
[Link 1]: https://blog.csdn.net/m0_37264397/article/details/112486401