NNDL作业XO识别

一、实现卷积-池化-激活

1. Numpy版本:手工实现 卷积-池化-激活
import numpy as np

# 初始化一张X图片矩阵
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]])
# 输出原矩阵
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)]
# 参考图卷积核1
Kernel[0] = np.array([[1, -1, -1],
                      [-1, 1, -1],
                      [-1, -1, 1]])
# 参考图卷积核2
Kernel[1] = np.array([[1, -1, 1],
                      [-1, 1, -1],
                      [1, -1, 1]])
# 参考图卷积核3
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))

import matplotlib.pyplot as plt
# 可视化特征图
plt.subplot(3, 1, 1)
plt.imshow(feature_map[0], cmap='gray')
plt.title('feature_map0')

plt.subplot(3, 1, 2)
plt.imshow(feature_map[1], cmap='gray')
plt.title('feature_map1')

plt.subplot(3, 1, 3)
plt.imshow(feature_map[2], cmap='gray')
plt.title('feature_map2')

plt.show()
# 可视化池化后的结果
plt.subplot(3, 1, 1)
plt.imshow(pooling[0], cmap='gray')
plt.title('pooling0')

plt.subplot(3, 1, 2)
plt.imshow(pooling[1], cmap='gray')
plt.title('pooling1')

plt.subplot(3, 1, 3)
plt.imshow(pooling[2], cmap='gray')
plt.title('pooling2')

plt.show()
# 可视化relu激活后的结果
plt.subplot(3, 1, 1)
plt.imshow(relu_map[0], cmap='gray')
plt.title('relu_map0')

plt.subplot(3, 1, 2)
plt.imshow(relu_map[1], cmap='gray')
plt.title('relu_map1')

plt.subplot(3, 1, 3)
plt.imshow(relu_map[2], cmap='gray')
plt.title('relu_map2')

plt.show()

2. Pytorch版本:调用函数实现 卷积-池化-激活
# https://blog.youkuaiyun.com/qq_26369907/article/details/88366147
# https://zhuanlan.zhihu.com/p/405242579
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)
# img = torch.tensor(feature_map3).data.squeeze().numpy()  # 将输出转换为图片的格式
# plt.imshow(img, cmap='gray')
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)
3. 可视化:了解数字与图像之间的关系

结果

numpy+++++++++++++++++++++
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]]
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]]]
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]]
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]]
torch++++++++++++++++++++++
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([[[[ 0.8099, -0.0790,  0.1432,  0.3655,  0.5877, -0.0790,  0.3655],
          [-0.0790,  1.0321, -0.0790,  0.3655, -0.0790,  0.1432, -0.0790],
          [ 0.1432, -0.0790,  1.0321, -0.3012,  0.1432, -0.0790,  0.5877],
          [ 0.3655,  0.3655, -0.3012,  0.5877, -0.3012,  0.3655,  0.3655],
          [ 0.5877, -0.0790,  0.1432, -0.3012,  1.0321, -0.0790,  0.1432],
          [-0.0790,  0.1432, -0.0790,  0.3655, -0.0790,  1.0321, -0.0790],
          [ 0.3655, -0.0790,  0.5877,  0.3655,  0.1432, -0.0790,  0.8099]]]],
       grad_fn=<DivBackward0>)
tensor([[[[ 0.3480, -0.5408,  0.1258, -0.0964,  0.1258, -0.5408,  0.3480],
          [-0.5408,  0.5703, -0.5408,  0.3480, -0.5408,  0.5703, -0.5408],
          [ 0.1258, -0.5408,  0.5703, -0.7631,  0.5703, -0.5408,  0.1258],
          [-0.0964,  0.3480, -0.7631,  1.0147, -0.7631,  0.3480, -0.0964],
          [ 0.1258, -0.5408,  0.5703, -0.7631,  0.5703, -0.5408,  0.1258],
          [-0.5408,  0.5703, -0.5408,  0.3480, -0.5408,  0.5703, -0.5408],
          [ 0.3480, -0.5408,  0.1258, -0.0964,  0.1258, -0.5408,  0.3480]]]],
       grad_fn=<DivBackward0>)
tensor([[[[ 0.2964, -0.1480,  0.5186,  0.2964,  0.0742, -0.1480,  0.7409],
          [-0.1480,  0.0742, -0.1480,  0.2964, -0.1480,  0.9631, -0.1480],
          [ 0.5186, -0.1480,  0.0742, -0.3702,  0.9631, -0.1480,  0.0742],
          [ 0.2964,  0.2964, -0.3702,  0.5186, -0.3702,  0.2964,  0.2964],
          [ 0.0742, -0.1480,  0.9631, -0.3702,  0.0742, -0.1480,  0.5186],
          [-0.1480,  0.9631, -0.1480,  0.2964, -0.1480,  0.0742, -0.1480],
          [ 0.7409, -0.1480,  0.0742,  0.2964,  0.5186, -0.1480,  0.2964]]]],
       grad_fn=<DivBackward0>)
--------------- 池化  ---------------
torch.Size([1, 1, 4, 4])
tensor([[[[1.0321, 0.3655, 0.5877, 0.3655],
          [0.3655, 1.0321, 0.3655, 0.5877],
          [0.5877, 0.3655, 1.0321, 0.1432],
          [0.3655, 0.5877, 0.1432, 0.8099]]]], grad_fn=<DivBackward0>)
tensor([[[[0.5703, 0.3480, 0.5703, 0.3480],
          [0.3480, 1.0147, 0.5703, 0.1258],
          [0.5703, 0.5703, 0.5703, 0.1258],
          [0.3480, 0.1258, 0.1258, 0.3480]]]], grad_fn=<DivBackward0>)
tensor([[[[0.2964, 0.5186, 0.9631, 0.7409],
          [0.5186, 0.5186, 0.9631, 0.2964],
          [0.9631, 0.9631, 0.0742, 0.5186],
          [0.7409, 0.2964, 0.5186, 0.2964]]]], grad_fn=<DivBackward0>)
--------------- 激活  ---------------
tensor([[[[0.8099, 0.0000, 0.1432, 0.3655, 0.5877, 0.0000, 0.3655],
          [0.0000, 1.0321, 0.0000, 0.3655, 0.0000, 0.1432, 0.0000],
          [0.1432, 0.0000, 1.0321, 0.0000, 0.1432, 0.0000, 0.5877],
          [0.3655, 0.3655, 0.0000, 0.5877, 0.0000, 0.3655, 0.3655],
          [0.5877, 0.0000, 0.1432, 0.0000, 1.0321, 0.0000, 0.1432],
          [0.0000, 0.1432, 0.0000, 0.3655, 0.0000, 1.0321, 0.0000],
          [0.3655, 0.0000, 0.5877, 0.3655, 0.1432, 0.0000, 0.8099]]]],
       grad_fn=<DivBackward0>)
tensor([[[[0.3480, 0.0000, 0.1258, 0.0000, 0.1258, 0.0000, 0.3480],
          [0.0000, 0.5703, 0.0000, 0.3480, 0.0000, 0.5703, 0.0000],
          [0.1258, 0.0000, 0.5703, 0.0000, 0.5703, 0.0000, 0.1258],
          [0.0000, 0.3480, 0.0000, 1.0147, 0.0000, 0.3480, 0.0000],
          [0.1258, 0.0000, 0.5703, 0.0000, 0.5703, 0.0000, 0.1258],
          [0.0000, 0.5703, 0.0000, 0.3480, 0.0000, 0.5703, 0.0000],
          [0.3480, 0.0000, 0.1258, 0.0000, 0.1258, 0.0000, 0.3480]]]],
       grad_fn=<DivBackward0>)
tensor([[[[0.2964, 0.0000, 0.5186, 0.2964, 0.0742, 0.0000, 0.7409],
          [0.0000, 0.0742, 0.0000, 0.2964, 0.0000, 0.9631, 0.0000],
          [0.5186, 0.0000, 0.0742, 0.0000, 0.9631, 0.0000, 0.0742],
          [0.2964, 0.2964, 0.0000, 0.5186, 0.0000, 0.2964, 0.2964],
          [0.0742, 0.0000, 0.9631, 0.0000, 0.0742, 0.0000, 0.5186],
          [0.0000, 0.9631, 0.0000, 0.2964, 0.0000, 0.0742, 0.0000],
          [0.7409, 0.0000, 0.0742, 0.2964, 0.5186, 0.0000, 0.2964]]]],
       grad_fn=<DivBackward0>)

在这里插入图片描述在这里插入图片描述在这里插入图片描述

二、 基于CNN的XO识别

数据集

在这里插入图片描述

构建模型
class Net(nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.conv1 = nn.Conv2d(1, 9, 3)  # in_channel , out_channel , kennel_size , stride
        self.maxpool = nn.MaxPool2d(2, 2)
        self.conv2 = nn.Conv2d(9, 5, 3)  # in_channel , out_channel , kennel_size , stride

        self.relu = nn.ReLU()
        self.fc1 = nn.Linear(27 * 27 * 5, 1200)  # full connect 1
        self.fc2 = nn.Linear(1200, 64)  # full connect 2
        self.fc3 = nn.Linear(64, 2)  # full connect 3

    def forward(self, x):
        x = self.maxpool(self.relu(self.conv1(x)))
        x = self.maxpool(self.relu(self.conv2(x)))
        x = x.view(-1, 27 * 27 * 5)
        x = self.relu(self.fc1(x))
        x = self.relu(self.fc2(x))
        x = self.fc3(x)
        return x
训练模型
model = Net()
criterion = torch.nn.CrossEntropyLoss()  # 损失函数 交叉熵损失函数
optimizer = optim.SGD(model.parameters(), lr=0.1)  # 优化函数:随机梯度下降

epochs = 5
for epoch in range(epochs):
    running_loss = 0.0
    for i, data in enumerate(data_loader):
        images, label = data
        out = model(images)
        loss = criterion(out, label)
        optimizer.zero_grad()
        loss.backward()
        optimizer.step()

        running_loss += loss.item()
        if (i + 1) % 10 == 0:
            print('[%d  %5d]   loss: %.3f' % (epoch + 1, i + 1, running_loss / 100))
            running_loss = 0.0

print('finished train')

# 保存模型 torch.save(model.state_dict(), model_path)
torch.save(model.state_dict(), 'model_name1.pth')  # 保存的是模型, 不止是w和b权重值
size of train_data: 1600
size of test_data: 400
torch.Size([64, 1, 116, 116])
torch.Size([64])
torch.Size([64, 1, 116, 116])
torch.Size([64])
[1     10]   loss: 0.069
[1     20]   loss: 0.068
[2     10]   loss: 0.064
[2     20]   loss: 0.053
[3     10]   loss: 0.069
[3     20]   loss: 0.068
[4     10]   loss: 0.059
[4     20]   loss: 0.065
[5     10]   loss: 0.050
[5     20]   loss: 0.015
odel = Net()
model.load_state_dict(torch.load('model_name1.pth', map_location='cpu'))  # 导入网络的参数

# model_load = torch.load('model_name1.pth')
# https://blog.youkuaiyun.com/qq_41360787/article/details/104332706

correct = 0
total = 0
with torch.no_grad():  # 进行评测的时候网络不更新梯度
    for data in data_loader_test:  # 读取测试集
        images, labels = data
        outputs = model(images)
        _, predicted = torch.max(outputs.data, 1)  # 取出 最大值的索引 作为 分类结果
        total += labels.size(0)  # labels 的长度
        correct += (predicted == labels).sum().item()  # 预测正确的数目
print('Accuracy of the network on the  test images: %f %%' % (100. * correct / total))

计算模型的准确率
finished train
Accuracy of the network on the  test images: 97.500000 %
查看训练好的模型的特征图
model_load = torch.load('model_name.pth')
# 读取一张图片 images[0],测试
print("label[0] truth:\t", label[0])
x = images[0]
x = x.reshape([1, 1, 116, 116])
predicted = torch.max(model_load(x), 1)
print("label[0] predict:\t", predicted.indices)

img = images[0].data.squeeze().numpy()  # 将输出转换为图片的格式
plt.imshow(img, cmap='gray')
plt.show()

# 看看每层的 卷积核 长相,特征图 长相
# 获取网络结构的特征矩阵并可视化
import torch
import matplotlib.pyplot as plt
import numpy as np
from PIL import Image
from torchvision import transforms, datasets
import torch.nn as nn
from torch.utils.data import DataLoader

#  定义图像预处理过程(要与网络模型训练过程中的预处理过程一致)

transforms = transforms.Compose([
    transforms.ToTensor(),  # 把图片进行归一化,并把数据转换成Tensor类型
    transforms.Grayscale(1)  # 把图片 转为灰度图
])
path = r'training_data_sm\train_data'
data_train = datasets.ImageFolder(path, transform=transforms)
data_loader = DataLoader(data_train, batch_size=64, shuffle=True)
for i, data in enumerate(data_loader):
    images, labels = data
    print(images.shape)
    print(labels.shape)
    break


class Net(nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.conv1 = nn.Conv2d(1, 9, 3)  # in_channel , out_channel , kennel_size , stride
        self.maxpool = nn.MaxPool2d(2, 2)
        self.conv2 = nn.Conv2d(9, 5, 3)  # in_channel , out_channel , kennel_size , stride

        self.relu = nn.ReLU()
        self.fc1 = nn.Linear(27 * 27 * 5, 1200)  # full connect 1
        self.fc2 = nn.Linear(1200, 64)  # full connect 2
        self.fc3 = nn.Linear(64, 2)  # full connect 3

    def forward(self, x):
        outputs = []
        x = self.conv1(x)
        outputs.append(x)
        x = self.relu(x)
        outputs.append(x)
        x = self.maxpool(x)
        outputs.append(x)
        x = self.conv2(x)

        x = self.relu(x)

        x = self.maxpool(x)

        x = x.view(-1, 27 * 27 * 5)
        x = self.relu(self.fc1(x))
        x = self.relu(self.fc2(x))
        x = self.fc3(x)
        return outputs


# create model
model1 = Net()

# load model weights加载预训练权重
# model_weight_path ="./AlexNet.pth"
model_weight_path = "model_name.pth"
model1 = torch.load(model_weight_path)

# 打印出模型的结构

print(model1)

x = images[0]
x = x.reshape([1, 1, 116, 116])
# forward正向传播过程
out_put = model1(x)
for feature_map in out_put:
    # [N, C, H, W] -> [C, H, W]    维度变换
    im = np.squeeze(feature_map.detach().numpy())
    # [C, H, W] -> [H, W, C]
    im = np.transpose(im, [1, 2, 0])
    print(im.shape)

    # show 9 feature maps
    plt.figure()
    for i in range(9):
        ax = plt.subplot(3, 3, i + 1)  # 参数意义:3:图片绘制行数,5:绘制图片列数,i+1:图的索引
        # [H, W, C]
        # 特征矩阵每一个channel对应的是一个二维的特征矩阵,就像灰度图像一样,channel=1
        # plt.imshow(im[:, :, i])
        plt.imshow(im[:, :, i], cmap='gray')
    plt.show()

在这里插入图片描述在这里插入图片描述在这里插入图片描述在这里插入图片描述

查看训练好的模型的卷积核
# 看看每层的 卷积核 长相,特征图 长相
# 获取网络结构的特征矩阵并可视化
import torch
import matplotlib.pyplot as plt
import numpy as np
from PIL import Image
from torchvision import transforms, datasets
import torch.nn as nn
from torch.utils.data import DataLoader

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
#  定义图像预处理过程(要与网络模型训练过程中的预处理过程一致)
transforms = transforms.Compose([
    transforms.ToTensor(),  # 把图片进行归一化,并把数据转换成Tensor类型
    transforms.Grayscale(1)  # 把图片 转为灰度图
])
path = r'training_data_sm\train_data'
data_train = datasets.ImageFolder(path, transform=transforms)
data_loader = DataLoader(data_train, batch_size=64, shuffle=True)
for i, data in enumerate(data_loader):
    images, labels = data
    # print(images.shape)
    # print(labels.shape)
    break


class Net(nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.conv1 = nn.Conv2d(1, 9, 3)  # in_channel , out_channel , kennel_size , stride
        self.maxpool = nn.MaxPool2d(2, 2)
        self.conv2 = nn.Conv2d(9, 5, 3)  # in_channel , out_channel , kennel_size , stride

        self.relu = nn.ReLU()
        self.fc1 = nn.Linear(27 * 27 * 5, 1200)  # full connect 1
        self.fc2 = nn.Linear(1200, 64)  # full connect 2
        self.fc3 = nn.Linear(64, 2)  # full connect 3

    def forward(self, x):
        outputs = []
        x = self.maxpool(self.relu(self.conv1(x)))
        # outputs.append(x)
        x = self.maxpool(self.relu(self.conv2(x)))
        outputs.append(x)
        x = x.view(-1, 27 * 27 * 5)
        x = self.relu(self.fc1(x))
        x = self.relu(self.fc2(x))
        x = self.fc3(x)
        return outputs


# create model
model1 = Net()

# load model weights加载预训练权重
model_weight_path = "model_name.pth"
model1 = torch.load(model_weight_path)

x = images[0]
x = x.reshape([1, 1, 116, 116])
# forward正向传播过程
out_put = model1(x)

weights_keys = model1.state_dict().keys()
for key in weights_keys:
    print("key :", key)
    # 卷积核通道排列顺序 [kernel_number, kernel_channel, kernel_height, kernel_width]
    if key == "conv1.weight":
        weight_t = model1.state_dict()[key].numpy()
        print("weight_t.shape", weight_t.shape)
        k = weight_t[:, 0, :, :]  # 获取第一个卷积核的信息参数
        # show 9 kernel ,1 channel
        plt.figure()

        for i in range(9):
            ax = plt.subplot(3, 3, i + 1)  # 参数意义:3:图片绘制行数,5:绘制图片列数,i+1:图的索引
            plt.imshow(k[i, :, :], cmap='gray')
            title_name = 'kernel' + str(i) + ',channel1'
            plt.title(title_name)
        plt.show()

    if key == "conv2.weight":
        weight_t = model1.state_dict()[key].numpy()
        print("weight_t.shape", weight_t.shape)
        k = weight_t[:, :, :, :]  # 获取第一个卷积核的信息参数
        print(k.shape)
        print(k)

        plt.figure()
        for c in range(9):
            channel = k[:, c, :, :]
            for i in range(5):
                ax = plt.subplot(2, 3, i + 1)  # 参数意义:3:图片绘制行数,5:绘制图片列数,i+1:图的索引
                plt.imshow(channel[i, :, :], cmap='gray')
                title_name = 'kernel' + str(i) + ',channel' + str(c)
                plt.title(title_name)
            plt.show()


第一层:
在这里插入图片描述
第二层:
在这里插入图片描述

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