J8 - Inception v1算法

• 🍨 本文为🔗365天深度学习训练营 中的学习记录博客
• 🍖 原作者：K同学啊

---

---

理论知识

GoogLeNet首次出现就在2014年的ILSVRC比赛中获得冠军，最初的版本为InceptionV1。共有22层深，参数量5M。
可以达到同时期VGGNet的性能，但是参数量更少。

Inception

GoogleLeNet的核心模块是Inception模块。主要的思路是将提取特征的卷积并行，在同一层进行多种尺度的卷积计算。

初始版本（上图左侧所示）就是沿着上述的思路设计的，后面在初始版本的基础上，借鉴了Network-in-Network的思想，使用了1x1卷积实现降维操作，减小网络的参数量和计算量。如上图右侧所示。

卷积计算

卷积的参数量计算可以通俗的来讲为（原图通道数*目标通道数*卷积核宽*卷积核高 + 目标通道数）
例如：对100x100x128做通道为256的5x5卷积（填充方式为same，不改变图像尺寸）。参数量为128x256x5x5+256 = 819456

如果加入了一个通道为32的1x1卷积后，再做通道为256的5x5卷积。参数量就是1283211 + 32 + 3255256 + 256 = 209184 ，由此可见，加入1x1卷积参数量减少到原来的1/4左右

模型结构

InceptionV1的完整模块和结构图如下

模型实现

inception 结构

class inception_block(nn.Module):
    def __init__(self, in_channels, ch1x1, ch3x3red, ch3x3, ch5x5red, ch5x5, pool_proj):
        super().__init__()

        # 1x1 分支
        self.branch1 = nn.Sequential(
            nn.Conv2d(in_channels, ch1x1, kernel_size=1),
            nn.BatchNorm2d(ch1x1),
            nn.ReLU(inplace=True)
        )
        # 1x1 -> 3x3 分支
        self.branch2 = nn.Sequential(
            nn.Conv2d(in_channels, ch3x3red, kernel_size=1),
            nn.BatchNorm2d(ch3x3red),
            nn.ReLU(inplace=True),
            nn.Conv2d(ch3x3red, ch3x3, kernel_size=3, padding=1),
            nn.BatchNorm2d(ch3x3),
            nn.ReLU(inplace=True)
        )
        # 1x1 -> 5x5 分支
        self.branch3 = nn.Sequential(
            nn.Conv2d(in_channels, ch5x5red, kernel_size=1),
            nn.BatchNorm2d(ch5x5red),
            nn.ReLU(inplace=True),
            nn.Conv2d(ch5x5red, ch5x5, kernel_size=5, padding=2),
            nn.BatchNorm2d(ch5x5),
            nn.ReLU(inplace=True)
        )
        # 3x3 -> 1x3 分支
        self.branch4 = nn.Sequential(
            nn.MaxPool2d(kernel_size=3, stride=1, padding=1),
            nn.Conv2d(in_channels, pool_proj, kernel_size=1),
            nn.BatchNorm2d(pool_proj),
            nn.ReLU(inplace=True)
        )

    def forward(self, x):
        branch1_output = self.branch1(x)
        branch2_output = self.branch2(x)
        branch3_output = self.branch3(x)
        branch4_output = self.branch4(x)

        outputs = [branch1_output, branch2_output, branch3_output, branch4_output]
        return torch.cat(outputs, 1)

GoogLeNet模型

class InceptionV1(nn.Module):
    def __init__(self, num_classes=1000):
        super().__init__()

        self.conv1 = nn.Conv2d(3, 64, kernel_size=7, stride=2, padding=3)
        self.maxpool1 = nn.MaxPool2d(kernel_size=3, stride=2, padding=1)
        self.conv2 = nn.Conv2d(64, 64, kernel_size=1, stride=1, padding=0)
        self.conv3 = nn.Conv2d(64, 192, kernel_size=3, stride=1, padding=1)
        self.maxpool2 = nn.MaxPool2d(kernel_size=3, stride=2, padding=1)

        self.inception3a = inception_block(192, 64, 96, 128, 16, 32, 32)
        self.inception3b = inception_block(256, 128, 128, 192, 32, 96, 64)
        self.maxpool3 = nn.MaxPool2d(kernel_size=3, stride=2, padding=1)

        self.inception4a = inception_block(480, 192, 96, 208, 16, 48, 64)
        self.inception4b = inception_block(512, 160, 112, 224, 24, 64, 64)
        self.inception4c = inception_block(512, 128, 128, 256, 24, 64, 64)
        self.inception4d = inception_block(512, 112, 144, 288, 32, 64, 64)
        self.inception4e = inception_block(528, 256, 160, 320, 32, 128, 128)
        self.maxpool4 = nn.MaxPool2d(kernel_size=3, stride=2, padding=1)

        self.inception5a = inception_block(832, 256, 160, 320, 32, 128, 128)

        self.inception5b = nn.Sequential(
            inception_block(832, 384, 192, 384, 48, 128, 128),
            nn.AvgPool2d(kernel_size=7, stride=1, padding=0),
            nn.Dropout(0.4)
        )

        # 全连接层
        self.classifier = nn.Sequential(
            nn.Linear(in_features=1024, out_features=1024),
            nn.ReLU(),
            nn.Linear(in_features=1024, out_features=num_classes),
            nn.S