## 1*3 3*1 卷积实现过程中，步长变化与3*3卷积相似，值得注意的是，padding部分需要作出相应的调整，否则输出的特征图大小无法实现整倍数变化
import torch.nn as nn
import torch
class Block(nn.Module):
expansion = 1
def __init__(self, in_channel=3, out_channel=16):
super(Block, self).__init__()
self.conv1 = nn.Conv2d(in_channels=in_channel, out_channels=out_channel,
kernel_size=3, stride=2, padding=1, bias=False)
self.bn1 = nn.BatchNorm2d(out_channel)
self.relu = nn.ReLU()
self.conv2 = nn.Conv2d(in_channels=out_channel, out_channels=out_channel,
kernel_size=(1, 3), stride=2, padding=(0, 1), bias=False)
self.bn2 = nn.BatchNorm2d(out_channel)
self.conv3 = nn.Conv2d(in_channels=out_channel, out_channels=out_channel,
kernel_size=(3, 1), stride=2, padding=(1, 0), bias=False)
self.bn3 = nn.BatchNorm2d(out_channel)
def forward(self, x):
out = self.conv1(x)
out = self.bn1(out)
out = self.relu(out)
out = self.conv2(out)
out = self.bn2(out)
out = self.relu(out)
out = self.conv3(out)
out = self.bn3(out)
out = self.relu(out)
return out
from torchsummary import summary
# 自己查看模型的架构
model = Block(in_channel=3, out_channel=16).cuda()
# 输出前channel在前面
summary(model, (3, 224, 224))
