"""
    conv2d：卷积操作
"""

import torch
import torch.nn.functional as F

# 初始化输入数据：5*5
input = torch.tensor([[1, 2, 0, 3, 1],
                      [0, 1, 2, 3, 1],
                      [1, 2, 1, 0, 0],
                      [5, 2, 3, 1, 1],
                      [2, 1, 0, 1, 1]])

# 构建卷积核
kernel = torch.tensor([[1, 2, 1],
                       [0, 1, 0],
                       [2, 1, 0]])

# 因为卷积对输入的数据有要求，需要进一步 reshape
input = torch.reshape(input, (1, 1, 5, 5))  # 将5*5(H*W) 重组为 1*1*5*5(B*C*H*W)
kernel = torch.reshape(kernel, (1, 1, 3, 3))

# 开始卷积
'''
    stride: 卷积核滑动的距离    
'''
output = F.conv2d(input, kernel, stride=1)

print(input.shape)
print(kernel.shape)
print(output)

output2 = F.conv2d(input, kernel, stride=2)
print(output2)