第11组 Alpha（1/3）

目录

• 1.过去完成的工作
  • 1.1 数据处理
  • 1.2 代码
  • 1.3 模型的训练
• 2. 每个人的工作
• 3.燃尽图
• 4. 剩余任务
• 5. 疑惑
• 6.站立会议

1.过去完成的工作

1.1 数据处理

1.2 代码

部分代码：

class Encoder(nn.Module):
    def __init__(self, input_channels):
        super(Encoder_SSH, self).__init__()

        self.enco1 = nn.Sequential(
            nn.Conv2d(input_channels, 64, kernel_size=3, stride=1, padding=1),
            nn.BatchNorm2d(64, momentum=bn_momentum),
            nn.ReLU(),
            nn.Conv2d(64, 64, kernel_size=3, stride=1, padding=1),
            nn.BatchNorm2d(64, momentum=bn_momentum),
            nn.ReLU()
        )
        self.enco2 = nn.Sequential(
            nn.Conv2d(64, 128, kernel_size=3, stride=1, padding=1),
            nn.BatchNorm2d(128, momentum=bn_momentum),
            nn.ReLU(),
            nn.Conv2d(128, 128, kernel_size=3, stride=1, padding=1),
            nn.BatchNorm2d(128, momentum=bn_momentum),
            nn.ReLU()
        )
        self.enco3 = nn.Sequential(
            nn.Conv2d(128, 256, kernel_size=3, stride=1, padding=1),
            nn.BatchNorm2d(256, momentum=bn_momentum),
            nn.ReLU(),
            nn.Conv2d(256, 256, kernel_size=3, stride=1, padding=1),
            nn.BatchNorm2d(256, momentum=bn_momentum),
            nn.ReLU(),
            nn.Conv2d(256, 256, kernel_size=3, stride=1, padding=1),
            nn.BatchNorm2d(256, momentum=bn_momentum),
            nn.ReLU()
        )
        self.enco4 = nn.Sequential(
            nn.Conv2d(256, 512, kernel_size=3, stride=1, padding=1),
            nn.BatchNorm2d(512, momentum=bn_momentum),
            nn.ReLU(),
            nn.Conv2d(512, 512, kernel_size=3, stride=1, padding=1),
            nn.BatchNorm2d(512, momentum=bn_momentum),
            nn.ReLU(),
            nn.Conv2d(512, 512, kernel_size=3, stride=1, padding=1),
            nn.BatchNorm2d(512, momentum=bn_momentum),
            nn.ReLU()
        )
        self.enco5 = nn.Sequential(
            nn.Conv2d(512, 512, kernel_size=3, stride=1, padding=1),
            nn.BatchNorm2d(512, momentum=bn_momentum),
            nn.ReLU(),
            nn.Conv2d(512, 512, kernel_size=3, stride=1, padding=1),
            nn.BatchNorm2d(512, momentum=bn_momentum),
            nn.ReLU(),
            nn.Conv2d(512, 512, kernel_size=3, stride=1, padding=1),
            nn.BatchNorm2d(512, momentum=bn_momentum),
            nn.ReLU()
        )

    def forward(self, x):
        id = []

        rx1 = self.enco1(x)
        x1, id1 = F.max_pool2d(rx1, kernel_size=2, stride=2, return_indices=True)  # 保留最大值的位置
        id.append(id1)
        rx2 = self.enco2(x1)
        x2, id2 = F.max_pool2d(rx2, kernel_size=2, stride=2, return_indices=True)
        id.append(id2)
        rx3 = self.enco3(x2)
        x3, id3 = F.max_pool2d(rx3, kernel_size=2, stride=2, return_indices=True)
        id.append(id3)
        rx4 = self.enco4(x3)
        x4, id4 = F.max_pool2d(rx4, kernel_size=2, stride=2, return_indices=True)
        id.append(id4)
        rx5 = self.enco5(x4)
        x5, id5 = F.max_pool2d(rx5, kernel_size=2, stride=2, return_indices=True)
        id.append(id5)
        return x1,x3,x5,id

1.3 模型的训练

---

2. 每个人的工作

熊万富：模型代码的完成

刘翰文：数据处理及代码训练

姜逸飞：博客编辑、资料整理、代码讨论

张嘉保：博客编辑、资料整理、代码讨论

---

3.燃尽图

4. 剩余任务

1. 模型的优化
2. 代码整理及其他开发等

5. 疑惑

暂无

6.站立会议