【2021-2022 春学期】人工智能-作业5:卷积-池化-激活

for循环版本:

import numpy as np
 
x = np.array([[-1, -1, -1, -1, -1, -1, -1, -1, -1],
              [-1, 1, -1, -1, -1, -1, -1, 1, -1],
              [-1, -1, 1, -1, -1, -1, 1, -1, -1],
              [-1, -1, -1, 1, -1, 1, -1, -1, -1],
              [-1, -1, -1, -1, 1, -1, -1, -1, -1],
              [-1, -1, -1, 1, -1, 1, -1, -1, -1],
              [-1, -1, 1, -1, -1, -1, 1, -1, -1],
              [-1, 1, -1, -1, -1, -1, -1, 1, -1],
              [-1, -1, -1, -1, -1, -1, -1, -1, -1]])
print("x=\n", x)
# 初始化 三个 卷积核
Kernel = [[0 for i in range(0, 3)] for j in range(0, 3)]
Kernel[0] = np.array([[1, -1, -1],
                      [-1, 1, -1],
                      [-1, -1, 1]])
Kernel[1] = np.array([[1, -1, 1],
                      [-1, 1, -1],
                      [1, -1, 1]])
Kernel[2] = np.array([[-1, -1, 1],
                      [-1, 1, -1],
                      [1, -1, -1]])
 
# --------------- 卷积  ---------------
stride = 1  # 步长
feature_map_h = 7  # 特征图的高
feature_map_w = 7  # 特征图的宽
feature_map = [0 for i in range(0, 3)]  # 初始化3个特征图
for i in range(0, 3):
    feature_map[i] = np.zeros((feature_map_h, feature_map_w))  # 初始化特征图
for h in range(feature_map_h):  # 向下滑动,得到卷积后的固定行
    for w in range(feature_map_w):  # 向右滑动,得到卷积后的固定行的列
        v_start = h * stride  # 滑动窗口的起始行(高)
        v_end = v_start + 3  # 滑动窗口的结束行(高)
        h_start = w * stride  # 滑动窗口的起始列(宽)
        h_end = h_start + 3  # 滑动窗口的结束列(宽)
        window = x[v_start:v_end, h_start:h_end]  # 从图切出一个滑动窗口
        for i in range(0, 3):
            feature_map[i][h, w] = np.divide(np.sum(np.multiply(window, Kernel[i][:, :])), 9)
print("feature_map:\n", np.around(feature_map, decimals=2))
 
# --------------- 池化  ---------------
pooling_stride = 2  # 步长
pooling_h = 4  # 特征图的高
pooling_w = 4  # 特征图的宽
feature_map_pad_0 = [[0 for i in range(0, 8)] for j in range(0, 8)]
for i in range(0, 3):  # 特征图 补 0 ,行 列 都要加 1 (因为上一层是奇数,池化窗口用的偶数)
    feature_map_pad_0[i] = np.pad(feature_map[i], ((0, 1), (0, 1)), 'constant', constant_values=(0, 0))
# print("feature_map_pad_0 0:\n", np.around(feature_map_pad_0[0], decimals=2))
 
pooling = [0 for i in range(0, 3)]
for i in range(0, 3):
    pooling[i] = np.zeros((pooling_h, pooling_w))  # 初始化特征图
for h in range(pooling_h):  # 向下滑动,得到卷积后的固定行
    for w in range(pooling_w):  # 向右滑动,得到卷积后的固定行的列
        v_start = h * pooling_stride  # 滑动窗口的起始行(高)
        v_end = v_start + 2  # 滑动窗口的结束行(高)
        h_start = w * pooling_stride  # 滑动窗口的起始列(宽)
        h_end = h_start + 2  # 滑动窗口的结束列(宽)
        for i in range(0, 3):
            pooling[i][h, w] = np.max(feature_map_pad_0[i][v_start:v_end, h_start:h_end])
print("pooling:\n", np.around(pooling[0], decimals=2))
print("pooling:\n", np.around(pooling[1], decimals=2))
print("pooling:\n", np.around(pooling[2], decimals=2))
 
 
# --------------- 激活  ---------------
def relu(x):
    return (abs(x) + x) / 2
 
 
relu_map_h = 7  # 特征图的高
relu_map_w = 7  # 特征图的宽
relu_map = [0 for i in range(0, 3)]  # 初始化3个特征图
for i in range(0, 3):
    relu_map[i] = np.zeros((relu_map_h, relu_map_w))  # 初始化特征图
 
for i in range(0, 3):
    relu_map[i] = relu(feature_map[i])
 
print("relu map :\n",np.around(relu_map[0], decimals=2))
print("relu map :\n",np.around(relu_map[1], decimals=2))
print("relu map :\n",np.around(relu_map[2], decimals=2))

输入:

x=
 [[-1 -1 -1 -1 -1 -1 -1 -1 -1]
 [-1  1 -1 -1 -1 -1 -1  1 -1]
 [-1 -1  1 -1 -1 -1  1 -1 -1]
 [-1 -1 -1  1 -1  1 -1 -1 -1]
 [-1 -1 -1 -1  1 -1 -1 -1 -1]
 [-1 -1 -1  1 -1  1 -1 -1 -1]
 [-1 -1  1 -1 -1 -1  1 -1 -1]
 [-1  1 -1 -1 -1 -1 -1  1 -1]
 [-1 -1 -1 -1 -1 -1 -1 -1 -1]]

输出

#卷积结果
feature_map:
 [[[ 0.78 -0.11  0.11  0.33  0.56 -0.11  0.33]
  [-0.11  1.   -0.11  0.33 -0.11  0.11 -0.11]
  [ 0.11 -0.11  1.   -0.33  0.11 -0.11  0.56]
  [ 0.33  0.33 -0.33  0.56 -0.33  0.33  0.33]
  [ 0.56 -0.11  0.11 -0.33  1.   -0.11  0.11]
  [-0.11  0.11 -0.11  0.33 -0.11  1.   -0.11]
  [ 0.33 -0.11  0.56  0.33  0.11 -0.11  0.78]]

 [[ 0.33 -0.56  0.11 -0.11  0.11 -0.56  0.33]
  [-0.56  0.56 -0.56  0.33 -0.56  0.56 -0.56]
  [ 0.11 -0.56  0.56 -0.78  0.56 -0.56  0.11]
  [-0.11  0.33 -0.78  1.   -0.78  0.33 -0.11]
  [ 0.11 -0.56  0.56 -0.78  0.56 -0.56  0.11]
  [-0.56  0.56 -0.56  0.33 -0.56  0.56 -0.56]
  [ 0.33 -0.56  0.11 -0.11  0.11 -0.56  0.33]]

 [[ 0.33 -0.11  0.56  0.33  0.11 -0.11  0.78]
  [-0.11  0.11 -0.11  0.33 -0.11  1.   -0.11]
  [ 0.56 -0.11  0.11 -0.33  1.   -0.11  0.11]
  [ 0.33  0.33 -0.33  0.56 -0.33  0.33  0.33]
  [ 0.11 -0.11  1.   -0.33  0.11 -0.11  0.56]
  [-0.11  1.   -0.11  0.33 -0.11  0.11 -0.11]
  [ 0.78 -0.11  0.11  0.33  0.56 -0.11  0.33]]]

#池化结果
pooling:
 [[1.   0.33 0.56 0.33]
 [0.33 1.   0.33 0.56]
 [0.56 0.33 1.   0.11]
 [0.33 0.56 0.11 0.78]]
pooling:
 [[0.56 0.33 0.56 0.33]
 [0.33 1.   0.56 0.11]
 [0.56 0.56 0.56 0.11]
 [0.33 0.11 0.11 0.33]]
pooling:
 [[0.33 0.56 1.   0.78]
 [0.56 0.56 1.   0.33]
 [1.   1.   0.11 0.56]
 [0.78 0.33 0.56 0.33]]


#激活结果
relu map :
 [[0.78 0.   0.11 0.33 0.56 0.   0.33]
 [0.   1.   0.   0.33 0.   0.11 0.  ]
 [0.11 0.   1.   0.   0.11 0.   0.56]
 [0.33 0.33 0.   0.56 0.   0.33 0.33]
 [0.56 0.   0.11 0.   1.   0.   0.11]
 [0.   0.11 0.   0.33 0.   1.   0.  ]
 [0.33 0.   0.56 0.33 0.11 0.   0.78]]
relu map :
 [[0.33 0.   0.11 0.   0.11 0.   0.33]
 [0.   0.56 0.   0.33 0.   0.56 0.  ]
 [0.11 0.   0.56 0.   0.56 0.   0.11]
 [0.   0.33 0.   1.   0.   0.33 0.  ]
 [0.11 0.   0.56 0.   0.56 0.   0.11]
 [0.   0.56 0.   0.33 0.   0.56 0.  ]
 [0.33 0.   0.11 0.   0.11 0.   0.33]]
relu map :
 [[0.33 0.   0.56 0.33 0.11 0.   0.78]
 [0.   0.11 0.   0.33 0.   1.   0.  ]
 [0.56 0.   0.11 0.   1.   0.   0.11]
 [0.33 0.33 0.   0.56 0.   0.33 0.33]
 [0.11 0.   1.   0.   0.11 0.   0.56]
 [0.   1.   0.   0.33 0.   0.11 0.  ]
 [0.78 0.   0.11 0.33 0.56 0.   0.33]]


Pytorch版本

import numpy as np
import torch
import torch.nn as nn
 
x = torch.tensor([[[[-1, -1, -1, -1, -1, -1, -1, -1, -1],
                    [-1, 1, -1, -1, -1, -1, -1, 1, -1],
                    [-1, -1, 1, -1, -1, -1, 1, -1, -1],
                    [-1, -1, -1, 1, -1, 1, -1, -1, -1],
                    [-1, -1, -1, -1, 1, -1, -1, -1, -1],
                    [-1, -1, -1, 1, -1, 1, -1, -1, -1],
                    [-1, -1, 1, -1, -1, -1, 1, -1, -1],
                    [-1, 1, -1, -1, -1, -1, -1, 1, -1],
                    [-1, -1, -1, -1, -1, -1, -1, -1, -1]]]], dtype=torch.float)
print(x.shape)
print(x)
 
print("--------------- 卷积  ---------------")
conv1 = nn.Conv2d(1, 1, (3, 3), 1)  # in_channel , out_channel , kennel_size , stride
conv1.weight.data = torch.Tensor([[[[1, -1, -1],
                                    [-1, 1, -1],
                                    [-1, -1, 1]]
                                   ]])
conv2 = nn.Conv2d(1, 1, (3, 3), 1)  # in_channel , out_channel , kennel_size , stride
conv2.weight.data = torch.Tensor([[[[1, -1, 1],
                                    [-1, 1, -1],
                                    [1, -1, 1]]
                                   ]])
conv3 = nn.Conv2d(1, 1, (3, 3), 1)  # in_channel , out_channel , kennel_size , stride
conv3.weight.data = torch.Tensor([[[[-1, -1, 1],
                                    [-1, 1, -1],
                                    [1, -1, -1]]
                                   ]])
 
feature_map1 = conv1(x)
feature_map2 = conv2(x)
feature_map3 = conv3(x)
 
print(feature_map1 / 9)
print(feature_map2 / 9)
print(feature_map3 / 9)
 
print("--------------- 池化  ---------------")
max_pool = nn.MaxPool2d(2, padding=0, stride=2)  # Pooling
zeroPad = nn.ZeroPad2d(padding=(0, 1, 0, 1))  # pad 0 , Left Right Up Down
 
feature_map_pad_0_1 = zeroPad(feature_map1)
feature_pool_1 = max_pool(feature_map_pad_0_1)
feature_map_pad_0_2 = zeroPad(feature_map2)
feature_pool_2 = max_pool(feature_map_pad_0_2)
feature_map_pad_0_3 = zeroPad(feature_map3)
feature_pool_3 = max_pool(feature_map_pad_0_3)
 
print(feature_pool_1.size())
print(feature_pool_1 / 9)
print(feature_pool_2 / 9)
print(feature_pool_3 / 9)
 
print("--------------- 激活  ---------------")
activation_function = nn.ReLU()
 
feature_relu1 = activation_function(feature_map1)
feature_relu2 = activation_function(feature_map2)
feature_relu3 = activation_function(feature_map3)
print(feature_relu1 / 9)
print(feature_relu2 / 9)
print(feature_relu3 / 9)

torch.Size([1, 1, 9, 9])
tensor([[[[-1., -1., -1., -1., -1., -1., -1., -1., -1.],
          [-1.,  1., -1., -1., -1., -1., -1.,  1., -1.],
          [-1., -1.,  1., -1., -1., -1.,  1., -1., -1.],
          [-1., -1., -1.,  1., -1.,  1., -1., -1., -1.],
          [-1., -1., -1., -1.,  1., -1., -1., -1., -1.],
          [-1., -1., -1.,  1., -1.,  1., -1., -1., -1.],
          [-1., -1.,  1., -1., -1., -1.,  1., -1., -1.],
          [-1.,  1., -1., -1., -1., -1., -1.,  1., -1.],
          [-1., -1., -1., -1., -1., -1., -1., -1., -1.]]]])
--------------- 卷积  ---------------
tensor([[[[ 0.7503, -0.1385,  0.0837,  0.3059,  0.5281, -0.1385,  0.3059],
          [-0.1385,  0.9726, -0.1385,  0.3059, -0.1385,  0.0837, -0.1385],
          [ 0.0837, -0.1385,  0.9726, -0.3608,  0.0837, -0.1385,  0.5281],
          [ 0.3059,  0.3059, -0.3608,  0.5281, -0.3608,  0.3059,  0.3059],
          [ 0.5281, -0.1385,  0.0837, -0.3608,  0.9726, -0.1385,  0.0837],
          [-0.1385,  0.0837, -0.1385,  0.3059, -0.1385,  0.9726, -0.1385],
          [ 0.3059, -0.1385,  0.5281,  0.3059,  0.0837, -0.1385,  0.7503]]]],
       grad_fn=<DivBackward0>)
tensor([[[[ 0.3379, -0.5510,  0.1157, -0.1066,  0.1157, -0.5510,  0.3379],
          [-0.5510,  0.5601, -0.5510,  0.3379, -0.5510,  0.5601, -0.5510],
          [ 0.1157, -0.5510,  0.5601, -0.7732,  0.5601, -0.5510,  0.1157],
          [-0.1066,  0.3379, -0.7732,  1.0045, -0.7732,  0.3379, -0.1066],
          [ 0.1157, -0.5510,  0.5601, -0.7732,  0.5601, -0.5510,  0.1157],
          [-0.5510,  0.5601, -0.5510,  0.3379, -0.5510,  0.5601, -0.5510],
          [ 0.3379, -0.5510,  0.1157, -0.1066,  0.1157, -0.5510,  0.3379]]]],
       grad_fn=<DivBackward0>)
tensor([[[[ 0.3643, -0.0802,  0.5865,  0.3643,  0.1421, -0.0802,  0.8087],
          [-0.0802,  0.1421, -0.0802,  0.3643, -0.0802,  1.0309, -0.0802],
          [ 0.5865, -0.0802,  0.1421, -0.3024,  1.0309, -0.0802,  0.1421],
          [ 0.3643,  0.3643, -0.3024,  0.5865, -0.3024,  0.3643,  0.3643],
          [ 0.1421, -0.0802,  1.0309, -0.3024,  0.1421, -0.0802,  0.5865],
          [-0.0802,  1.0309, -0.0802,  0.3643, -0.0802,  0.1421, -0.0802],
          [ 0.8087, -0.0802,  0.1421,  0.3643,  0.5865, -0.0802,  0.3643]]]],
       grad_fn=<DivBackward0>)
--------------- 池化  ---------------
torch.Size([1, 1, 4, 4])
tensor([[[[0.9726, 0.3059, 0.5281, 0.3059],
          [0.3059, 0.9726, 0.3059, 0.5281],
          [0.5281, 0.3059, 0.9726, 0.0837],
          [0.3059, 0.5281, 0.0837, 0.7503]]]], grad_fn=<DivBackward0>)
tensor([[[[0.5601, 0.3379, 0.5601, 0.3379],
          [0.3379, 1.0045, 0.5601, 0.1157],
          [0.5601, 0.5601, 0.5601, 0.1157],
          [0.3379, 0.1157, 0.1157, 0.3379]]]], grad_fn=<DivBackward0>)
tensor([[[[0.3643, 0.5865, 1.0309, 0.8087],
          [0.5865, 0.5865, 1.0309, 0.3643],
          [1.0309, 1.0309, 0.1421, 0.5865],
          [0.8087, 0.3643, 0.5865, 0.3643]]]], grad_fn=<DivBackward0>)
--------------- 激活  ---------------
tensor([[[[0.7503, 0.0000, 0.0837, 0.3059, 0.5281, 0.0000, 0.3059],
          [0.0000, 0.9726, 0.0000, 0.3059, 0.0000, 0.0837, 0.0000],
          [0.0837, 0.0000, 0.9726, 0.0000, 0.0837, 0.0000, 0.5281],
          [0.3059, 0.3059, 0.0000, 0.5281, 0.0000, 0.3059, 0.3059],
          [0.5281, 0.0000, 0.0837, 0.0000, 0.9726, 0.0000, 0.0837],
          [0.0000, 0.0837, 0.0000, 0.3059, 0.0000, 0.9726, 0.0000],
          [0.3059, 0.0000, 0.5281, 0.3059, 0.0837, 0.0000, 0.7503]]]],
       grad_fn=<DivBackward0>)
tensor([[[[0.3379, 0.0000, 0.1157, 0.0000, 0.1157, 0.0000, 0.3379],
          [0.0000, 0.5601, 0.0000, 0.3379, 0.0000, 0.5601, 0.0000],
          [0.1157, 0.0000, 0.5601, 0.0000, 0.5601, 0.0000, 0.1157],
          [0.0000, 0.3379, 0.0000, 1.0045, 0.0000, 0.3379, 0.0000],
          [0.1157, 0.0000, 0.5601, 0.0000, 0.5601, 0.0000, 0.1157],
          [0.0000, 0.5601, 0.0000, 0.3379, 0.0000, 0.5601, 0.0000],
          [0.3379, 0.0000, 0.1157, 0.0000, 0.1157, 0.0000, 0.3379]]]],
       grad_fn=<DivBackward0>)
tensor([[[[0.3643, 0.0000, 0.5865, 0.3643, 0.1421, 0.0000, 0.8087],
          [0.0000, 0.1421, 0.0000, 0.3643, 0.0000, 1.0309, 0.0000],
          [0.5865, 0.0000, 0.1421, 0.0000, 1.0309, 0.0000, 0.1421],
          [0.3643, 0.3643, 0.0000, 0.5865, 0.0000, 0.3643, 0.3643],
          [0.1421, 0.0000, 1.0309, 0.0000, 0.1421, 0.0000, 0.5865],
          [0.0000, 1.0309, 0.0000, 0.3643, 0.0000, 0.1421, 0.0000],
          [0.8087, 0.0000, 0.1421, 0.3643, 0.5865, 0.0000, 0.3643]]]],
       grad_fn=<DivBackward0>)

可视化:了解数字与图像之间的关系

import torch
import torch.nn as nn
import matplotlib.pyplot as plt
plt.rcParams['font.sans-serif']=['SimHei'] #用来正常显示中文标签
plt.rcParams['axes.unicode_minus']=False #用来正常显示负号 #有中文出现的情况,需要u'内容
x = torch.tensor([[[[-1, -1, -1, -1, -1, -1, -1, -1, -1],
                    [-1, 1, -1, -1, -1, -1, -1, 1, -1],
                    [-1, -1, 1, -1, -1, -1, 1, -1, -1],
                    [-1, -1, -1, 1, -1, 1, -1, -1, -1],
                    [-1, -1, -1, -1, 1, -1, -1, -1, -1],
                    [-1, -1, -1, 1, -1, 1, -1, -1, -1],
                    [-1, -1, 1, -1, -1, -1, 1, -1, -1],
                    [-1, 1, -1, -1, -1, -1, -1, 1, -1],
                    [-1, -1, -1, -1, -1, -1, -1, -1, -1]]]], dtype=torch.float)
print(x.shape)
print(x)
img = x.data.squeeze().numpy()  # 将输出转换为图片的格式
plt.imshow(img, cmap='gray')
plt.title('原图')
plt.show()
 
print("--------------- 卷积  ---------------")
conv1 = nn.Conv2d(1, 1, (3, 3), 1)  # in_channel , out_channel , kennel_size , stride
conv1.weight.data = torch.Tensor([[[[1, -1, -1],
                                    [-1, 1, -1],
                                    [-1, -1, 1]]
                                   ]])
img = conv1.weight.data.squeeze().numpy()  # 将输出转换为图片的格式
plt.imshow(img, cmap='gray')
plt.title('Kernel 1')
plt.show()
conv2 = nn.Conv2d(1, 1, (3, 3), 1)  # in_channel , out_channel , kennel_size , stride
conv2.weight.data = torch.Tensor([[[[1, -1, 1],
                                    [-1, 1, -1],
                                    [1, -1, 1]]
                                   ]])
img = conv2.weight.data.squeeze().numpy()  # 将输出转换为图片的格式
plt.imshow(img, cmap='gray')
plt.title('Kernel 2')
plt.show()
conv3 = nn.Conv2d(1, 1, (3, 3), 1)  # in_channel , out_channel , kennel_size , stride
conv3.weight.data = torch.Tensor([[[[-1, -1, 1],
                                    [-1, 1, -1],
                                    [1, -1, -1]]
                                   ]])
img = conv3.weight.data.squeeze().numpy()  # 将输出转换为图片的格式
plt.imshow(img, cmap='gray')
plt.title('Kernel 3')
plt.show()
 
feature_map1 = conv1(x)
feature_map2 = conv2(x)
feature_map3 = conv3(x)
 
print(feature_map1 / 9)
print(feature_map2 / 9)
print(feature_map3 / 9)
 
img = feature_map1.data.squeeze().numpy()  # 将输出转换为图片的格式
plt.imshow(img, cmap='gray')
plt.title('卷积后的特征图1')
plt.show()
 
print("--------------- 池化  ---------------")
max_pool = nn.MaxPool2d(2, padding=0, stride=2)  # Pooling
zeroPad = nn.ZeroPad2d(padding=(0, 1, 0, 1))  # pad 0 , Left Right Up Down
 
feature_map_pad_0_1 = zeroPad(feature_map1)
feature_pool_1 = max_pool(feature_map_pad_0_1)
feature_map_pad_0_2 = zeroPad(feature_map2)
feature_pool_2 = max_pool(feature_map_pad_0_2)
feature_map_pad_0_3 = zeroPad(feature_map3)
feature_pool_3 = max_pool(feature_map_pad_0_3)
 
print(feature_pool_1.size())
print(feature_pool_1 / 9)
print(feature_pool_2 / 9)
print(feature_pool_3 / 9)
img = feature_pool_1.data.squeeze().numpy()  # 将输出转换为图片的格式
plt.imshow(img, cmap='gray')
plt.title('卷积池化后的特征图1')
plt.show()
 
print("--------------- 激活  ---------------")
activation_function = nn.ReLU()
 
feature_relu1 = activation_function(feature_map1)
feature_relu2 = activation_function(feature_map2)
feature_relu3 = activation_function(feature_map3)
print(feature_relu1 / 9)
print(feature_relu2 / 9)
print(feature_relu3 / 9)
img = feature_relu1.data.squeeze().numpy()  # 将输出转换为图片的格式
plt.imshow(img, cmap='gray')
plt.title('卷积 + relu 后的特征图1')
plt.show()

运行结果

torch.Size([1, 1, 9, 9])
tensor([[[[-1., -1., -1., -1., -1., -1., -1., -1., -1.],
          [-1.,  1., -1., -1., -1., -1., -1.,  1., -1.],
          [-1., -1.,  1., -1., -1., -1.,  1., -1., -1.],
          [-1., -1., -1.,  1., -1.,  1., -1., -1., -1.],
          [-1., -1., -1., -1.,  1., -1., -1., -1., -1.],
          [-1., -1., -1.,  1., -1.,  1., -1., -1., -1.],
          [-1., -1.,  1., -1., -1., -1.,  1., -1., -1.],
          [-1.,  1., -1., -1., -1., -1., -1.,  1., -1.],
          [-1., -1., -1., -1., -1., -1., -1., -1., -1.]]]])
--------------- 卷积  ---------------
tensor([[[[ 0.7500, -0.1389,  0.0833,  0.3055,  0.5278, -0.1389,  0.3055],
          [-0.1389,  0.9722, -0.1389,  0.3055, -0.1389,  0.0833, -0.1389],
          [ 0.0833, -0.1389,  0.9722, -0.3611,  0.0833, -0.1389,  0.5278],
          [ 0.3055,  0.3055, -0.3611,  0.5278, -0.3611,  0.3055,  0.3055],
          [ 0.5278, -0.1389,  0.0833, -0.3611,  0.9722, -0.1389,  0.0833],
          [-0.1389,  0.0833, -0.1389,  0.3055, -0.1389,  0.9722, -0.1389],
          [ 0.3055, -0.1389,  0.5278,  0.3055,  0.0833, -0.1389,  0.7500]]]],
       grad_fn=<DivBackward0>)
tensor([[[[ 0.3406, -0.5483,  0.1184, -0.1038,  0.1184, -0.5483,  0.3406],
          [-0.5483,  0.5629, -0.5483,  0.3406, -0.5483,  0.5629, -0.5483],
          [ 0.1184, -0.5483,  0.5629, -0.7705,  0.5629, -0.5483,  0.1184],
          [-0.1038,  0.3406, -0.7705,  1.0073, -0.7705,  0.3406, -0.1038],
          [ 0.1184, -0.5483,  0.5629, -0.7705,  0.5629, -0.5483,  0.1184],
          [-0.5483,  0.5629, -0.5483,  0.3406, -0.5483,  0.5629, -0.5483],
          [ 0.3406, -0.5483,  0.1184, -0.1038,  0.1184, -0.5483,  0.3406]]]],
       grad_fn=<DivBackward0>)
tensor([[[[ 0.3218, -0.1227,  0.5440,  0.3218,  0.0995, -0.1227,  0.7662],
          [-0.1227,  0.0995, -0.1227,  0.3218, -0.1227,  0.9884, -0.1227],
          [ 0.5440, -0.1227,  0.0995, -0.3449,  0.9884, -0.1227,  0.0995],
          [ 0.3218,  0.3218, -0.3449,  0.5440, -0.3449,  0.3218,  0.3218],
          [ 0.0995, -0.1227,  0.9884, -0.3449,  0.0995, -0.1227,  0.5440],
          [-0.1227,  0.9884, -0.1227,  0.3218, -0.1227,  0.0995, -0.1227],
          [ 0.7662, -0.1227,  0.0995,  0.3218,  0.5440, -0.1227,  0.3218]]]],
       grad_fn=<DivBackward0>)
--------------- 池化  ---------------
torch.Size([1, 1, 4, 4])
tensor([[[[0.9722, 0.3055, 0.5278, 0.3055],
          [0.3055, 0.9722, 0.3055, 0.5278],
          [0.5278, 0.3055, 0.9722, 0.0833],
          [0.3055, 0.5278, 0.0833, 0.7500]]]], grad_fn=<DivBackward0>)
tensor([[[[0.5629, 0.3406, 0.5629, 0.3406],
          [0.3406, 1.0073, 0.5629, 0.1184],
          [0.5629, 0.5629, 0.5629, 0.1184],
          [0.3406, 0.1184, 0.1184, 0.3406]]]], grad_fn=<DivBackward0>)
tensor([[[[0.3218, 0.5440, 0.9884, 0.7662],
          [0.5440, 0.5440, 0.9884, 0.3218],
          [0.9884, 0.9884, 0.0995, 0.5440],
          [0.7662, 0.3218, 0.5440, 0.3218]]]], grad_fn=<DivBackward0>)
--------------- 激活  ---------------
tensor([[[[0.7500, 0.0000, 0.0833, 0.3055, 0.5278, 0.0000, 0.3055],
          [0.0000, 0.9722, 0.0000, 0.3055, 0.0000, 0.0833, 0.0000],
          [0.0833, 0.0000, 0.9722, 0.0000, 0.0833, 0.0000, 0.5278],
          [0.3055, 0.3055, 0.0000, 0.5278, 0.0000, 0.3055, 0.3055],
          [0.5278, 0.0000, 0.0833, 0.0000, 0.9722, 0.0000, 0.0833],
          [0.0000, 0.0833, 0.0000, 0.3055, 0.0000, 0.9722, 0.0000],
          [0.3055, 0.0000, 0.5278, 0.3055, 0.0833, 0.0000, 0.7500]]]],
       grad_fn=<DivBackward0>)
tensor([[[[0.3406, 0.0000, 0.1184, 0.0000, 0.1184, 0.0000, 0.3406],
          [0.0000, 0.5629, 0.0000, 0.3406, 0.0000, 0.5629, 0.0000],
          [0.1184, 0.0000, 0.5629, 0.0000, 0.5629, 0.0000, 0.1184],
          [0.0000, 0.3406, 0.0000, 1.0073, 0.0000, 0.3406, 0.0000],
          [0.1184, 0.0000, 0.5629, 0.0000, 0.5629, 0.0000, 0.1184],
          [0.0000, 0.5629, 0.0000, 0.3406, 0.0000, 0.5629, 0.0000],
          [0.3406, 0.0000, 0.1184, 0.0000, 0.1184, 0.0000, 0.3406]]]],
       grad_fn=<DivBackward0>)
tensor([[[[0.3218, 0.0000, 0.5440, 0.3218, 0.0995, 0.0000, 0.7662],
          [0.0000, 0.0995, 0.0000, 0.3218, 0.0000, 0.9884, 0.0000],
          [0.5440, 0.0000, 0.0995, 0.0000, 0.9884, 0.0000, 0.0995],
          [0.3218, 0.3218, 0.0000, 0.5440, 0.0000, 0.3218, 0.3218],
          [0.0995, 0.0000, 0.9884, 0.0000, 0.0995, 0.0000, 0.5440],
          [0.0000, 0.9884, 0.0000, 0.3218, 0.0000, 0.0995, 0.0000],
          [0.7662, 0.0000, 0.0995, 0.3218, 0.5440, 0.0000, 0.3218]]]],
       grad_fn=<DivBackward0>)
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值