卷积神经网络（ResNet-50）鸟类识别

卷积神经网络（CNN）mnist手写数字分类识别的实现

卷积神经网络（CNN）多种图片分类的实现

卷积神经网络（CNN）衣服图像分类的实现

卷积神经网络（CNN）鲜花的识别

卷积神经网络（CNN）天气识别

卷积神经网络（VGG-16）海贼王人物识别

卷积神经网络（VGG-19）灵笼人物识别

前期工作

1. 设置GPU（如果使用的是CPU可以忽略这步）

我的环境：

• 语言环境：Python3.6.5
• 编译器：jupyter notebook
• 深度学习环境：TensorFlow2.4.1

import tensorflow as tf

gpus = tf.config.list_physical_devices("GPU")

if gpus:
    tf.config.experimental.set_memory_growth(gpus[0], True)  #设置GPU显存用量按需使用
    tf.config.set_visible_devices([gpus[0]],"GPU")
1
2
3
4
5
6
7

2. 导入数据

import matplotlib.pyplot as plt
# 支持中文
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

import os,PIL

# 设置随机种子尽可能使结果可以重现
import numpy as np
np.random.seed(1)

# 设置随机种子尽可能使结果可以重现
import tensorflow as tf
tf.random.set_seed(1)

from tensorflow import keras
from tensorflow.keras import layers,models

import pathlib
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19

data_dir = "weather_photos/"
data_dir = pathlib.Path(data_dir)
1
2

3. 查看数据

image_count = len(list(data_dir.glob('*/*')))

print("图片总数为：",image_count)
1
2
3

二、数据预处理

1. 加载数据

使用image_dataset_from_directory方法将磁盘中的数据加载到tf.data.Dataset中

batch_size = 32
img_height = 224
img_width = 224
1
2
3

train_ds = tf.keras.preprocessing.image_dataset_from_directory(
    data_dir,
    validation_split=0.2,
    subset="training",
    seed=123,
    image_size=(img_height, img_width),
    batch_size=batch_size)
1
2
3
4
5
6
7

Found 565 files belonging to 4 classes.
Using 452 files for training.
1
2

val_ds = tf.keras.preprocessing.image_dataset_from_directory(
    data_dir,
    validation_split=0.2,
    subset="validation",
    seed=123,
    image_size=(img_height, img_width),
    batch_size=batch_size)
1
2
3
4
5
6
7

Found 565 files belonging to 4 classes.
Using 113 files for validation.
1
2

我们可以通过class_names输出数据集的标签。标签将按字母顺序对应于目录名称。

class_names = train_ds.class_names
print(class_names)
1
2

['Bananaquit', 'Black Skimmer', 'Black Throated Bushtiti', 'Cockatoo']
1

2. 可视化数据

plt.figure(figsize=(10, 5))  # 图形的宽为10高为5

for images, labels in train_ds.take(1):
    for i in range(8):
        
        ax = plt.subplot(2, 4, i + 1)  

        plt.imshow(images[i].numpy().astype("uint8"))
        plt.title(class_names[labels[i]])
        
        plt.axis("off")
1
2
3
4
5
6
7
8
9
10
11

plt.imshow(images[1].numpy().astype("uint8"))
1

3. 再次检查数据

for image_batch, labels_batch in train_ds:
    print(image_batch.shape)
    print(labels_batch.shape)
    break
1
2
3
4

(8, 224, 224, 3)
(8,)
1
2

• Image_batch是形状的张量（8, 224, 224, 3)。这是一批形状240x240x3的8张图片（最后一维指的是彩色通道RGB）。
• Label_batch是形状（8，）的张量，这些标签对应8张图片

4. 配置数据集

AUTOTUNE = tf.data.AUTOTUNE

train_ds = train_ds.cache().shuffle(1000).prefetch(buffer_size=AUTOTUNE)
val_ds = val_ds.cache().prefetch(buffer_size=AUTOTUNE)
1
2
3
4

三、残差网络（ResNet）介绍

1. 残差网络解决了什么

残差网络是为了解决神经网络隐藏层过多时，而引起的网络退化问题。退化（degradation）问题是指：当网络隐藏层变多时，网络的准确度达到饱和然后急剧退化，而且这个退化不是由于过拟合引起的。

拓展： 深度神经网络的“两朵乌云”

• 梯度弥散/爆炸

简单来讲就是网络太深了，会导致模型训练难以收敛。这个问题可以被标准初始化和中间层正规化的方法有效控制。（现阶段知道这么一回事就好了）

• 网络退化

随着网络深度增加，网络的表现先是逐渐增加至饱和，然后迅速下降，这个退化不是由于过拟合引起的。

2. ResNet-50介绍

ResNet-50有两个基本的块，分别名为Conv Block和Identity Block

四、构建ResNet-50网络模型

下面是本文的重点，可以试着按照上面三张图自己构建一下ResNet-50

from keras import layers

from keras.layers import Input,Activation,BatchNormalization,Flatten
from keras.layers import Dense,Conv2D,MaxPooling2D,ZeroPadding2D,AveragePooling2D
from keras.models import Model

def identity_block(input_tensor, kernel_size, filters, stage, block):

    filters1, filters2, filters3 = filters

    name_base = str(stage) + block + '_identity_block_'

    x = Conv2D(filters1, (1, 1), name=name_base + 'conv1')(input_tensor)
    x = BatchNormalization(name=name_base + 'bn1')(x)
    x = Activation('relu', name=name_base + 'relu1')(x)

    x = Conv2D(filters2, kernel_size,padding='same', name=name_base + 'conv2')(x)
    x = BatchNormalization(name=name_base + 'bn2')(x)
    x = Activation('relu', name=name_base + 'relu2')(x)

    x = Conv2D(filters3, (1, 1), name=name_base + 'conv3')(x)
    x = BatchNormalization(name=name_base + 'bn3')(x)

    x = layers.add([x, input_tensor] ,name=name_base + 'add')
    x = Activation('relu', name=name_base + 'relu4')(x)
    return x


def conv_block(input_tensor, kernel_size, filters, stage, block, strides=(2, 2)):

    filters1, filters2, filters3 = filters

    res_name_base = str(stage) + block + '_conv_block_res_'
    name_base = str(stage) + block + '_conv_block_'

    x = Conv2D(filters1, (1, 1), strides=strides, name=name_base + 'conv1')(input_tensor)
    x = BatchNormalization(name=name_base + 'bn1')(x)
    x = Activation('relu', name=name_base + 'relu1')(x)

    x = Conv2D(filters2, kernel_size, padding='same', name=name_base + 'conv2')(x)
    x = BatchNormalization(name=name_base + 'bn2')(x)
    x = Activation('relu', name=name_base + 'relu2')(x)

    x = Conv2D(filters3, (1, 1), name=name_base + 'conv3')(x)
    x = BatchNormalization(name=name_base + 'bn3')(x)

    shortcut = Conv2D(filters3, (1, 1), strides=strides, name=res_name_base + 'conv')(input_tensor)
    shortcut = BatchNormalization(name=res_name_base + 'bn')(shortcut)

    x = layers.add([x, shortcut], name=name_base+'add')
    x = Activation('relu', name=name_base+'relu4')(x)
    return x

def ResNet50(input_shape=[224,224,3],classes=1000):

    img_input = Input(shape=input_shape)
    x = ZeroPadding2D((3, 3))(img_input)

    x = Conv2D(64, (7, 7), strides=(2, 2), name='conv1')(x)
    x = BatchNormalization(name='bn_conv1')(x)
    x = Activation('relu')(x)
    x = MaxPooling2D((3, 3), strides=(2, 2))(x)

    x =     conv_block(x, 3, [64, 64, 256], stage=2, block='a', strides=(1, 1))
    x = identity_block(x, 3, [64, 64, 256], stage=2, block='b')
    x = identity_block(x, 3, [64, 64, 256], stage=2, block='c')

    x =     conv_block(x, 3, [128, 128, 512], stage=3, block='a')
    x = identity_block(x, 3, [128, 128, 512], stage=3, block='b')
    x = identity_block(x, 3, [128, 128, 512], stage=3, block='c')
    x = identity_block(x, 3, [128, 128, 512], stage=3, block='d')

    x =     conv_block(x, 3, [256, 256, 1024], stage=4, block='a')
    x = identity_block(x, 3, [256, 256, 1024], stage=4, block='b')
    x = identity_block(x, 3, [256, 256, 1024], stage=4, block='c')
    x = identity_block(x, 3, [256, 256, 1024], stage=4, block='d')
    x = identity_block(x, 3, [256, 256, 1024], stage=4, block='e')
    x = identity_block(x, 3, [256, 256, 1024], stage=4, block='f')

    x =     conv_block(x, 3, [512, 512, 2048], stage=5, block='a')
    x = identity_block(x, 3, [512, 512, 2048], stage=5, block='b')
    x = identity_block(x, 3, [512, 512, 2048], stage=5, block='c')

    x = AveragePooling2D((7, 7), name='avg_pool')(x)

    x = Flatten()(x)
    x = Dense(classes, activation='softmax', name='fc1000')(x)

    model = Model(img_input, x, name='resnet50')
    
    # 加载预训练模型
    model.load_weights("resnet50_weights_tf_dim_ordering_tf_kernels.h5")

    return model

model = ResNet50()
model.summary()
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97

Model: "resnet50"
__________________________________________________________________________________________________
Layer (type)                    Output Shape         Param #     Connected to                     
==================================================================================================
input_1 (InputLayer)            [(None, 224, 224, 3) 0                                            
__________________________________________________________________________________________________
zero_padding2d (ZeroPadding2D)  (None, 230, 230, 3)  0           input_1[0][0]                    
__________________________________________________________________________________________________
conv1 (Conv2D)                  (None, 112, 112, 64) 9472        zero_padding2d[0][0]             
__________________________________________________________________________________________________
bn_conv1 (BatchNormalization)   (None, 112, 112, 64) 256         conv1[0][0]                      
__________________________________________________________________________________________________
activation (Activation)         (None, 112, 112, 64) 0           bn_conv1[0][0]                   
__________________________________________________________________________________________________
max_pooling2d (MaxPooling2D)    (None, 55, 55, 64)   0           activation[0][0]                 
__________________________________________________________________________________________________
2a_conv_block_conv1 (Conv2D)    (None, 55, 55, 64)   4160        max_pooling2d[0][0]              
__________________________________________________________________________________________________
2a_conv_block_bn1 (BatchNormali (None, 55, 55, 64)   256         2a_conv_block_conv1[0][0]        
__________________________________________________________________________________________________
2a_conv_block_relu1 (Activation (None, 55, 55, 64)   0           2a_conv_block_bn1[0][0]          
__________________________________________________________________________________________________
2a_conv_block_conv2 (Conv2D)    (None, 55, 55, 64)   36928       2a_conv_block_relu1[0][0]        
__________________________________________________________________________________________________
2a_conv_block_bn2 (BatchNormali (None, 55, 55, 64)   256         2a_conv_block_conv2[0][0]        
__________________________________________________________________________________________________
2a_conv_block_relu2 (Activation (None, 55, 55, 64)   0           2a_conv_block_bn2[0][0]          
__________________________________________________________________________________________________
2a_conv_block_conv3 (Conv2D)    (None, 55, 55, 256)  16640       2a_conv_block_relu2[0][0]        
__________________________________________________________________________________________________
2a_conv_block_res_conv (Conv2D) (None, 55, 55, 256)  16640       max_pooling2d[0][0]              
__________________________________________________________________________________________________
2a_conv_block_bn3 (BatchNormali (None, 55, 55, 256)  1024        2a_conv_block_conv3[0][0]        
__________________________________________________________________________________________________
2a_conv_block_res_bn (BatchNorm (None, 55, 55, 256)  1024        2a_conv_block_res_conv[0][0]     
__________________________________________________________________________________________________
2a_conv_block_add (Add)         (None, 55, 55, 256)  0           2a_conv_block_bn3[0][0]          
                                                                 2a_conv_block_res_bn[0][0]       
__________________________________________________________________________________________________
2a_conv_block_relu4 (Activation (None, 55, 55, 256)  0           2a_conv_block_add[0][0]          
__________________________________________________________________________________________________
2b_identity_block_conv1 (Conv2D (None, 55, 55, 64)   16448       2a_conv_block_relu4[0][0]        
__________________________________________________________________________________________________
2b_identity_block_bn1 (BatchNor (None, 55, 55, 64)   256         2b_identity_block_conv1[0][0]    
__________________________________________________________________________________________________
2b_identity_block_relu1 (Activa (None, 55, 55, 64)   0           2b_identity_block_bn1[0][0]      
__________________________________________________________________________________________________
2b_identity_block_conv2 (Conv2D (None, 55, 55, 64)   36928       2b_identity_block_relu1[0][0]    
__________________________________________________________________________________________________
2b_identity_block_bn2 (BatchNor (None, 55, 55, 64)   256         2b_identity_block_conv2[0][0]    
__________________________________________________________________________________________________
2b_identity_block_relu2 (Activa (None, 55, 55, 64)   0           2b_identity_block_bn2[0][0]      
__________________________________________________________________________________________________
2b_identity_block_conv3 (Conv2D (None, 55, 55, 256)  16640       2b_identity_block_relu2[0][0]    
__________________________________________________________________________________________________
2b_identity_block_bn3 (BatchNor (None, 55, 55, 256)  1024        2b_identity_block_conv3[0][0]    
__________________________________________________________________________________________________
2b_identity_block_add (Add)     (None, 55, 55, 256)  0           2b_identity_block_bn3[0][0]      
                                                                 2a_conv_block_relu4[0][0]        
__________________________________________________________________________________________________
2b_identity_block_relu4 (Activa (None, 55, 55, 256)  0           2b_identity_block_add[0][0]      
__________________________________________________________________________________________________
2c_identity_block_conv1 (Conv2D (None, 55, 55, 64)   16448       2b_identity_block_relu4[0][0]    
__________________________________________________________________________________________________
2c_identity_block_bn1 (BatchNor (None, 55, 55, 64)   256         2c_identity_block_conv1[0][0]    
__________________________________________________________________________________________________
2c_identity_block_relu1 (Activa (None, 55, 55, 64)   0           2c_identity_block_bn1[0][0]      
__________________________________________________________________________________________________
2c_identity_block_conv2 (Conv2D (None, 55, 55, 64)   36928       2c_identity_block_relu1[0][0]    
__________________________________________________________________________________________________
2c_identity_block_bn2 (BatchNor (None, 55, 55, 64)   256         2c_identity_block_conv2[0][0]    
__________________________________________________________________________________________________
2c_identity_block_relu2 (Activa (None, 55, 55, 64)   0           2c_identity_block_bn2[0][0]      
__________________________________________________________________________________________________
2c_identity_block_conv3 (Conv2D (None, 55, 55, 256)  16640       2c_identity_block_relu2[0][0]    
__________________________________________________________________________________________________
2c_identity_block_bn3 (BatchNor (None, 55, 55, 256)  1024        2c_identity_block_conv3[0][0]    
__________________________________________________________________________________________________
2c_identity_block_add (Add)     (None, 55, 55, 256)  0           2c_identity_block_bn3[0][0]      
                                                                 2b_identity_block_relu4[0][0]    
__________________________________________________________________________________________________
2c_identity_block_relu4 (Activa (None, 55, 55, 256)  0           2c_identity_block_add[0][0]      
__________________________________________________________________________________________________
3a_conv_block_conv1 (Conv2D)    (None, 28, 28, 128)  32896       2c_identity_block_relu4[0][0]    
__________________________________________________________________________________________________
3a_conv_block_bn1 (BatchNormali (None, 28, 28, 128)  512         3a_conv_block_conv1[0][0]        
__________________________________________________________________________________________________
3a_conv_block_relu1 (Activation (None, 28, 28, 128)  0           3a_conv_block_bn1[0][0]          
__________________________________________________________________________________________________
3a_conv_block_conv2 (Conv2D)    (None, 28, 28, 128)  147584      3a_conv_block_relu1[0][0]        
__________________________________________________________________________________________________
3a_conv_block_bn2 (BatchNormali (None, 28, 28, 128)  512         3a_conv_block_conv2[0][0]        
__________________________________________________________________________________________________
3a_conv_block_relu2 (Activation (None, 28, 28, 128)  0           3a_conv_block_bn2[0][0]          
__________________________________________________________________________________________________
3a_conv_block_conv3 (Conv2D)    (None, 28, 28, 512)  66048       3a_conv_block_relu2[0][0]        
__________________________________________________________________________________________________
3a_conv_block_res_conv (Conv2D) (None, 28, 28, 512)  131584      2c_identity_block_relu4[0][0]    
__________________________________________________________________________________________________
3a_conv_block_bn3 (BatchNormali (None, 28, 28, 512)  2048        3a_conv_block_conv3[0][0]        
__________________________________________________________________________________________________
3a_conv_block_res_bn (BatchNorm (None, 28, 28, 512)  2048        3a_conv_block_res_conv[0][0]     
__________________________________________________________________________________________________
3a_conv_block_add (Add)         (None, 28, 28, 512)  0           3a_conv_block_bn3[0][0]          
                                                                 3a_conv_block_res_bn[0][0]       
__________________________________________________________________________________________________
3a_conv_block_relu4 (Activation (None, 28, 28, 512)  0           3a_conv_block_add[0][0]          
__________________________________________________________________________________________________
3b_identity_block_conv1 (Conv2D (None, 28, 28, 128)  65664       3a_conv_block_relu4[0][0]        
__________________________________________________________________________________________________
3b_identity_block_bn1 (BatchNor (None, 28, 28, 128)  512         3b_identity_block_conv1[0][0]    
__________________________________________________________________________________________________
3b_identity_block_relu1 (Activa (None, 28, 28, 128)  0           3b_identity_block_bn1[0][0]      
__________________________________________________________________________________________________
3b_identity_block_conv2 (Conv2D (None, 28, 28, 128)  147584      3b_identity_block_relu1[0][0]    
__________________________________________________________________________________________________
3b_identity_block_bn2 (BatchNor (None, 28, 28, 128)  512         3b_identity_block_conv2[0][0]    
__________________________________________________________________________________________________
3b_identity_block_relu2 (Activa (None, 28, 28, 128)  0           3b_identity_block_bn2[0][0]      
__________________________________________________________________________________________________
3b_identity_block_conv3 (Conv2D (None, 28, 28, 512)  66048       3b_identity_block_relu2[0][0]    
__________________________________________________________________________________________________
3b_identity_block_bn3 (BatchNor (None, 28, 28, 512)  2048        3b_identity_block_conv3[0][0]    
__________________________________________________________________________________________________
3b_identity_block_add (Add)     (None, 28, 28, 512)  0           3b_identity_block_bn3[0][0]      
                                                                 3a_conv_block_relu4[0][0]        
__________________________________________________________________________________________________
3b_identity_block_relu4 (Activa (None, 28, 28, 512)  0           3b_identity_block_add[0][0]      
__________________________________________________________________________________________________
3c_identity_block_conv1 (Conv2D (None, 28, 28, 128)  65664       3b_identity_block_relu4[0][0]    
__________________________________________________________________________________________________
3c_identity_block_bn1 (BatchNor (None, 28, 28, 128)  512         3c_identity_block_conv1[0][0]    
__________________________________________________________________________________________________
3c_identity_block_relu1 (Activa (None, 28, 28, 128)  0           3c_identity_block_bn1[0][0]      
__________________________________________________________________________________________________
3c_identity_block_conv2 (Conv2D (None, 28, 28, 128)  147584      3c_identity_block_relu1[0][0]    
__________________________________________________________________________________________________
3c_identity_block_bn2 (BatchNor (None, 28, 28, 128)  512         3c_identity_block_conv2[0][0]    
__________________________________________________________________________________________________
3c_identity_block_relu2 (Activa (None, 28, 28, 128)  0           3c_identity_block_bn2[0][0]      
__________________________________________________________________________________________________
3c_identity_block_conv3 (Conv2D (None, 28, 28, 512)  66048       3c_identity_block_relu2[0][0]    
__________________________________________________________________________________________________
3c_identity_block_bn3 (BatchNor (None, 28, 28, 512)  2048        3c_identity_block_conv3[0][0]    
__________________________________________________________________________________________________
3c_identity_block_add (Add)     (None, 28, 28, 512)  0           3c_identity_block_bn3[0][0]      
                                                                 3b_identity_block_relu4[0][0]    
__________________________________________________________________________________________________
3c_identity_block_relu4 (Activa (None, 28, 28, 512)  0           3c_identity_block_add[0][0]      
__________________________________________________________________________________________________
3d_identity_block_conv1 (Conv2D (None, 28, 28, 128)  65664       3c_identity_block_relu4[0][0]    
__________________________________________________________________________________________________
3d_identity_block_bn1 (BatchNor (None, 28, 28, 128)  512         3d_identity_block_conv1[0][0]    
__________________________________________________________________________________________________
3d_identity_block_relu1 (Activa (None, 28, 28, 128)  0           3d_identity_block_bn1[0][0]      
__________________________________________________________________________________________________
3d_identity_block_conv2 (Conv2D (None, 28, 28, 128)  147584      3d_identity_block_relu1[0][0]    
__________________________________________________________________________________________________
3d_identity_block_bn2 (BatchNor (None, 28, 28, 128)  512         3d_identity_block_conv2[0][0]    
__________________________________________________________________________________________________
3d_identity_block_relu2 (Activa (None, 28, 28, 128)  0           3d_identity_block_bn2[0][0]      
__________________________________________________________________________________________________
3d_identity_block_conv3 (Conv2D (None, 28, 28, 512)  66048       3d_identity_block_relu2[0][0]    
__________________________________________________________________________________________________
3d_identity_block_bn3 (BatchNor (None, 28, 28, 512)  2048        3d_identity_block_conv3[0][0]    
__________________________________________________________________________________________________
3d_identity_block_add (Add)     (None, 28, 28, 512)  0           3d_identity_block_bn3[0][0]      
                                                                 3c_identity_block_relu4[0][0]    
__________________________________________________________________________________________________
3d_identity_block_relu4 (Activa (None, 28, 28, 512)  0           3d_identity_block_add[0][0]      
__________________________________________________________________________________________________
4a_conv_block_conv1 (Conv2D)    (None, 14, 14, 256)  131328      3d_identity_block_relu4[0][0]    
__________________________________________________________________________________________________
4a_conv_block_bn1 (BatchNormali (None, 14, 14, 256)  1024        4a_conv_block_conv1[0][0]        
__________________________________________________________________________________________________
4a_conv_block_relu1 (Activation (None, 14, 14, 256)  0           4a_conv_block_bn1[0][0]          
__________________________________________________________________________________________________
4a_conv_block_conv2 (Conv2D)    (None, 14, 14, 256)  590080      4a_conv_block_relu1[0][0]        
__________________________________________________________________________________________________
4a_conv_block_bn2 (BatchNormali (None, 14, 14, 256)  1024        4a_conv_block_conv2[0][0]        
__________________________________________________________________________________________________
4a_conv_block_relu2 (Activation (None, 14, 14, 256)  0           4a_conv_block_bn2[0][0]          
__________________________________________________________________________________________________
4a_conv_block_conv3 (Conv2D)    (None, 14, 14, 1024) 263168      4a_conv_block_relu2[0][0]        
__________________________________________________________________________________________________
4a_conv_block_res_conv (Conv2D) (None, 14, 14, 1024) 525312      3d_identity_block_relu4[0][0]    
__________________________________________________________________________________________________
4a_conv_block_bn3 (BatchNormali (None, 14, 14, 1024) 4096        4a_conv_block_conv3[0][0]        
__________________________________________________________________________________________________
4a_conv_block_res_bn (BatchNorm (None, 14, 14, 1024) 4096        4a_conv_block_res_conv[0][0]     
__________________________________________________________________________________________________
4a_conv_block_add (Add)         (None, 14, 14, 1024) 0           4a_conv_block_bn3[0][0]          
                                                                 4a_conv_block_res_bn[0][0]       
__________________________________________________________________________________________________
4a_conv_block_relu4 (Activation (None, 14, 14, 1024) 0           4a_conv_block_add[0][0]          
__________________________________________________________________________________________________
4b_identity_block_conv1 (Conv2D (None, 14, 14, 256)  262400      4a_conv_block_relu4[0][0]        
__________________________________________________________________________________________________
4b_identity_block_bn1 (BatchNor (None, 14, 14, 256)  1024        4b_identity_block_conv1[0][0]    
__________________________________________________________________________________________________
4b_identity_block_relu1 (Activa (None, 14, 14, 256)  0           4b_identity_block_bn1[0][0]      
__________________________________________________________________________________________________
4b_identity_block_conv2 (Conv2D (None, 14, 14, 256)  590080      4b_identity_block_relu1[0][0]    
__________________________________________________________________________________________________
4b_identity_block_bn2 (BatchNor (None, 14, 14, 256)  1024        4b_identity_block_conv2[0][0]    
__________________________________________________________________________________________________
4b_identity_block_relu2 (Activa (None, 14, 14, 256)  0           4b_identity_block_bn2[0][0]      
__________________________________________________________________________________________________
4b_identity_block_conv3 (Conv2D (None, 14, 14, 1024) 263168      4b_identity_block_relu2[0][0]    
__________________________________________________________________________________________________
4b_identity_block_bn3 (BatchNor (None, 14, 14, 1024) 4096        4b_identity_block_conv3[0][0]    
__________________________________________________________________________________________________
4b_identity_block_add (Add)     (None, 14, 14, 1024) 0           4b_identity_block_bn3[0][0]      
                                                                 4a_conv_block_relu4[0][0]        
__________________________________________________________________________________________________
4b_identity_block_relu4 (Activa (None, 14, 14, 1024) 0           4b_identity_block_add[0][0]      
__________________________________________________________________________________________________
4c_identity_block_conv1 (Conv2D (None, 14, 14, 256)  262400      4b_identity_block_relu4[0][0]    
__________________________________________________________________________________________________
4c_identity_block_bn1 (BatchNor (None, 14, 14, 256)  1024        4c_identity_block_conv1[0][0]    
__________________________________________________________________________________________________
4c_identity_block_relu1 (Activa (None, 14, 14, 256)  0           4c_identity_block_bn1[0][0]      
__________________________________________________________________________________________________
4c_identity_block_conv2 (Conv2D (None, 14, 14, 256)  590080      4c_identity_block_relu1[0][0]    
__________________________________________________________________________________________________
4c_identity_block_bn2 (BatchNor (None, 14, 14, 256)  1024        4c_identity_block_conv2[0][0]    
__________________________________________________________________________________________________
4c_identity_block_relu2 (Activa (None, 14, 14, 256)  0           4c_identity_block_bn2[0][0]      
__________________________________________________________________________________________________
4c_identity_block_conv3 (Conv2D (None, 14, 14, 1024) 263168      4c_identity_block_relu2[0][0]    
__________________________________________________________________________________________________
4c_identity_block_bn3 (BatchNor (None, 14, 14, 1024) 4096        4c_identity_block_conv3[0][0]    
__________________________________________________________________________________________________
4c_identity_block_add (Add)     (None, 14, 14, 1024) 0           4c_identity_block_bn3[0][0]      
                                                                 4b_identity_block_relu4[0][0]    
__________________________________________________________________________________________________
4c_identity_block_relu4 (Activa (None, 14, 14, 1024) 0           4c_identity_block_add[0][0]      
__________________________________________________________________________________________________
4d_identity_block_conv1 (Conv2D (None, 14, 14, 256)  262400      4c_identity_block_relu4[0][0]    
__________________________________________________________________________________________________
4d_identity_block_bn1 (BatchNor (None, 14, 14, 256)  1024        4d_identity_block_conv1[0][0]    
__________________________________________________________________________________________________
4d_identity_block_relu1 (Activa (None, 14, 14, 256)  0           4d_identity_block_bn1[0][0]      
__________________________________________________________________________________________________
4d_identity_block_conv2 (Conv2D (None, 14, 14, 256)  590080      4d_identity_block_relu1[0][0]    
__________________________________________________________________________________________________
4d_identity_block_bn2 (BatchNor (None, 14, 14, 256)  1024        4d_identity_block_conv2[0][0]    
__________________________________________________________________________________________________
4d_identity_block_relu2 (Activa (None, 14, 14, 256)  0           4d_identity_block_bn2[0][0]      
__________________________________________________________________________________________________
4d_identity_block_conv3 (Conv2D (None, 14, 14, 1024) 263168      4d_identity_block_relu2[0][0]    
__________________________________________________________________________________________________
4d_identity_block_bn3 (BatchNor (None, 14, 14, 1024) 4096        4d_identity_block_conv3[0][0]    
__________________________________________________________________________________________________
4d_identity_block_add (Add)     (None, 14, 14, 1024) 0           4d_identity_block_bn3[0][0]      
                                                                 4c_identity_block_relu4[0][0]    
__________________________________________________________________________________________________
4d_identity_block_relu4 (Activa (None, 14, 14, 1024) 0           4d_identity_block_add[0][0]      
__________________________________________________________________________________________________
4e_identity_block_conv1 (Conv2D (None, 14, 14, 256)  262400      4d_identity_block_relu4[0][0]    
__________________________________________________________________________________________________
4e_identity_block_bn1 (BatchNor (None, 14, 14, 256)  1024        4e_identity_block_conv1[0][0]    
__________________________________________________________________________________________________
4e_identity_block_relu1 (Activa (None, 14, 14, 256)  0           4e_identity_block_bn1[0][0]      
__________________________________________________________________________________________________
4e_identity_block_conv2 (Conv2D (None, 14, 14, 256)  590080      4e_identity_block_relu1[0][0]    
__________________________________________________________________________________________________
4e_identity_block_bn2 (BatchNor (None, 14, 14, 256)  1024        4e_identity_block_conv2[0][0]    
__________________________________________________________________________________________________
4e_identity_block_relu2 (Activa (None, 14, 14, 256)  0           4e_identity_block_bn2[0][0]      
__________________________________________________________________________________________________
4e_identity_block_conv3 (Conv2D (None, 14, 14, 1024) 263168      4e_identity_block_relu2[0][0]    
__________________________________________________________________________________________________
4e_identity_block_bn3 (BatchNor (None, 14, 14, 1024) 4096        4e_identity_block_conv3[0][0]    
__________________________________________________________________________________________________
4e_identity_block_add (Add)     (None, 14, 14, 1024) 0           4e_identity_block_bn3[0][0]      
                                                                 4d_identity_block_relu4[0][0]    
__________________________________________________________________________________________________
4e_identity_block_relu4 (Activa (None, 14, 14, 1024) 0           4e_identity_block_add[0][0]      
__________________________________________________________________________________________________
4f_identity_block_conv1 (Conv2D (None, 14, 14, 256)  262400      4e_identity_block_relu4[0][0]    
__________________________________________________________________________________________________
4f_identity_block_bn1 (BatchNor (None, 14, 14, 256)  1024        4f_identity_block_conv1[0][0]    
__________________________________________________________________________________________________
4f_identity_block_relu1 (Activa (None, 14, 14, 256)  0           4f_identity_block_bn1[0][0]      
__________________________________________________________________________________________________
4f_identity_block_conv2 (Conv2D (None, 14, 14, 256)  590080      4f_identity_block_relu1[0][0]    
__________________________________________________________________________________________________
4f_identity_block_bn2 (BatchNor (None, 14, 14, 256)  1024        4f_identity_block_conv2[0][0]    
__________________________________________________________________________________________________
4f_identity_block_relu2 (Activa (None, 14, 14, 256)  0           4f_identity_block_bn2[0][0]      
__________________________________________________________________________________________________
4f_identity_block_conv3 (Conv2D (None, 14, 14, 1024) 263168      4f_identity_block_relu2[0][0]    
__________________________________________________________________________________________________
4f_identity_block_bn3 (BatchNor (None, 14, 14, 1024) 4096        4f_identity_block_conv3[0][0]    
__________________________________________________________________________________________________
4f_identity_block_add (Add)     (None, 14, 14, 1024) 0           4f_identity_block_bn3[0][0]      
                                                                 4e_identity_block_relu4[0][0]    
__________________________________________________________________________________________________
4f_identity_block_relu4 (Activa (None, 14, 14, 1024) 0           4f_identity_block_add[0][0]      
__________________________________________________________________________________________________
5a_conv_block_conv1 (Conv2D)    (None, 7, 7, 512)    524800      4f_identity_block_relu4[0][0]    
__________________________________________________________________________________________________
5a_conv_block_bn1 (BatchNormali (None, 7, 7, 512)    2048        5a_conv_block_conv1[0][0]        
__________________________________________________________________________________________________
5a_conv_block_relu1 (Activation (None, 7, 7, 512)    0           5a_conv_block_bn1[0][0]          
__________________________________________________________________________________________________
5a_conv_block_conv2 (Conv2D)    (None, 7, 7, 512)    2359808     5a_conv_block_relu1[0][0]        
__________________________________________________________________________________________________
5a_conv_block_bn2 (BatchNormali (None, 7, 7, 512)    2048        5a_conv_block_conv2[0][0]        
__________________________________________________________________________________________________
5a_conv_block_relu2 (Activation (None, 7, 7, 512)    0           5a_conv_block_bn2[0][0]          
__________________________________________________________________________________________________
5a_conv_block_conv3 (Conv2D)    (None, 7, 7, 2048)   1050624     5a_conv_block_relu2[0][0]        
__________________________________________________________________________________________________
5a_conv_block_res_conv (Conv2D) (None, 7, 7, 2048)   2099200     4f_identity_block_relu4[0][0]    
__________________________________________________________________________________________________
5a_conv_block_bn3 (BatchNormali (None, 7, 7, 2048)   8192        5a_conv_block_conv3[0][0]        
__________________________________________________________________________________________________
5a_conv_block_res_bn (BatchNorm (None, 7, 7, 2048)   8192        5a_conv_block_res_conv[0][0]     
__________________________________________________________________________________________________
5a_conv_block_add (Add)         (None, 7, 7, 2048)   0           5a_conv_block_bn3[0][0]          
                                                                 5a_conv_block_res_bn[0][0]       
__________________________________________________________________________________________________
5a_conv_block_relu4 (Activation (None, 7, 7, 2048)   0           5a_conv_block_add[0][0]          
__________________________________________________________________________________________________
5b_identity_block_conv1 (Conv2D (None, 7, 7, 512)    1049088     5a_conv_block_relu4[0][0]        
__________________________________________________________________________________________________
5b_identity_block_bn1 (BatchNor (None, 7, 7, 512)    2048        5b_identity_block_conv1[0][0]    
__________________________________________________________________________________________________
5b_identity_block_relu1 (Activa (None, 7, 7, 512)    0           5b_identity_block_bn1[0][0]      
__________________________________________________________________________________________________
5b_identity_block_conv2 (Conv2D (None, 7, 7, 512)    2359808     5b_identity_block_relu1[0][0]    
__________________________________________________________________________________________________
5b_identity_block_bn2 (BatchNor (None, 7, 7, 512)    2048        5b_identity_block_conv2[0][0]    
__________________________________________________________________________________________________
5b_identity_block_relu2 (Activa (None, 7, 7, 512)    0           5b_identity_block_bn2[0][0]      
__________________________________________________________________________________________________
5b_identity_block_conv3 (Conv2D (None, 7, 7, 2048)   1050624     5b_identity_block_relu2[0][0]    
__________________________________________________________________________________________________
5b_identity_block_bn3 (BatchNor (None, 7, 7, 2048)   8192        5b_identity_block_conv3[0][0]    
__________________________________________________________________________________________________
5b_identity_block_add (Add)     (None, 7, 7, 2048)   0           5b_identity_block_bn3[0][0]      
                                                                 5a_conv_block_relu4[0][0]        
__________________________________________________________________________________________________
5b_identity_block_relu4 (Activa (None, 7, 7, 2048)   0           5b_identity_block_add[0][0]      
__________________________________________________________________________________________________
5c_identity_block_conv1 (Conv2D (None, 7, 7, 512)    1049088     5b_identity_block_relu4[0][0]    
__________________________________________________________________________________________________
5c_identity_block_bn1 (BatchNor (None, 7, 7, 512)    2048        5c_identity_block_conv1[0][0]    
__________________________________________________________________________________________________
5c_identity_block_relu1 (Activa (None, 7, 7, 512)    0           5c_identity_block_bn1[0][0]      
__________________________________________________________________________________________________
5c_identity_block_conv2 (Conv2D (None, 7, 7, 512)    2359808     5c_identity_block_relu1[0][0]    
__________________________________________________________________________________________________
5c_identity_block_bn2 (BatchNor (None, 7, 7, 512)    2048        5c_identity_block_conv2[0][0]    
__________________________________________________________________________________________________
5c_identity_block_relu2 (Activa (None, 7, 7, 512)    0           5c_identity_block_bn2[0][0]      
__________________________________________________________________________________________________
5c_identity_block_conv3 (Conv2D (None, 7, 7, 2048)   1050624     5c_identity_block_relu2[0][0]    
__________________________________________________________________________________________________
5c_identity_block_bn3 (BatchNor (None, 7, 7, 2048)   8192        5c_identity_block_conv3[0][0]    
__________________________________________________________________________________________________
5c_identity_block_add (Add)     (None, 7, 7, 2048)   0           5c_identity_block_bn3[0][0]      
                                                                 5b_identity_block_relu4[0][0]    
__________________________________________________________________________________________________
5c_identity_block_relu4 (Activa (None, 7, 7, 2048)   0           5c_identity_block_add[0][0]      
__________________________________________________________________________________________________
avg_pool (AveragePooling2D)     (None, 1, 1, 2048)   0           5c_identity_block_relu4[0][0]    
__________________________________________________________________________________________________
flatten (Flatten)               (None, 2048)         0           avg_pool[0][0]                   
__________________________________________________________________________________________________
fc1000 (Dense)                  (None, 1000)         2049000     flatten[0][0]                    
==================================================================================================
Total params: 25,636,712
Trainable params: 25,583,592
Non-trainable params: 53,120
__________________________________________________________________________________________________
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378

五、编译

在准备对模型进行训练之前，还需要再对其进行一些设置。以下内容是在模型的编译步骤中添加的：

• 损失函数（loss）：用于衡量模型在训练期间的准确率。
• 优化器（optimizer）：决定模型如何根据其看到的数据和自身的损失函数进行更新。
• 指标（metrics）：用于监控训练和测试步骤。以下示例使用了准确率，即被正确分类的图像的比率。

---

# 设置优化器，我这里改变了学习率。
opt = tf.keras.optimizers.Adam(learning_rate=1e-7)

model.compile(optimizer="adam",
              loss='sparse_categorical_crossentropy',
              metrics=['accuracy'])
1
2
3
4
5
6

六、训练模型

epochs = 10

history = model.fit(
    train_ds,
    validation_data=val_ds,
    epochs=epochs
)
1
2
3
4
5
6
7

Epoch 1/10
57/57 [==============================] - 12s 87ms/step - loss: 2.4394 - accuracy: 0.6620 - val_loss: 484.4415 - val_accuracy: 0.1858
Epoch 2/10
57/57 [==============================] - 3s 55ms/step - loss: 0.3268 - accuracy: 0.8904 - val_loss: 5.5874 - val_accuracy: 0.4513
Epoch 3/10
57/57 [==============================] - 3s 52ms/step - loss: 0.1593 - accuracy: 0.9558 - val_loss: 2.0023 - val_accuracy: 0.6726
Epoch 4/10
57/57 [==============================] - 3s 52ms/step - loss: 0.0546 - accuracy: 0.9869 - val_loss: 1.4410 - val_accuracy: 0.7788
Epoch 5/10
57/57 [==============================] - 3s 51ms/step - loss: 0.1656 - accuracy: 0.9567 - val_loss: 2.1653 - val_accuracy: 0.5487
Epoch 6/10
57/57 [==============================] - 3s 51ms/step - loss: 0.2403 - accuracy: 0.9221 - val_loss: 1.0641 - val_accuracy: 0.7257
Epoch 7/10
57/57 [==============================] - 3s 52ms/step - loss: 0.1774 - accuracy: 0.9377 - val_loss: 0.4922 - val_accuracy: 0.8673
Epoch 8/10
57/57 [==============================] - 3s 51ms/step - loss: 0.0733 - accuracy: 0.9760 - val_loss: 0.2036 - val_accuracy: 0.9381
Epoch 9/10
57/57 [==============================] - 3s 51ms/step - loss: 0.0185 - accuracy: 0.9992 - val_loss: 0.1091 - val_accuracy: 0.9735
Epoch 10/10
57/57 [==============================] - 3s 51ms/step - loss: 0.0086 - accuracy: 1.0000 - val_loss: 0.1246 - val_accuracy: 0.9735
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20

六、模型评估

acc = history.history['accuracy']
val_acc = history.history['val_accuracy']

loss = history.history['loss']
val_loss = history.history['val_loss']

epochs_range = range(epochs)

plt.figure(figsize=(12, 4))
plt.subplot(1, 2, 1)
plt.plot(epochs_range, acc, label='Training Accuracy')
plt.plot(epochs_range, val_acc, label='Validation Accuracy')
plt.legend(loc='lower right')
plt.title('Training and Validation Accuracy')

plt.subplot(1, 2, 2)
plt.plot(epochs_range, loss, label='Training Loss')
plt.plot(epochs_range, val_loss, label='Validation Loss')
plt.legend(loc='upper right')
plt.title('Training and Validation Loss')
plt.show()
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21

八、保存and加载模型

# 保存模型
model.save('model/my_model.h5')
# 加载模型
new_model = keras.models.load_model('model/my_model.h5')
1
2
3
4

九、预测

# 采用加载的模型（new_model）来看预测结果

plt.figure(figsize=(10, 5))  # 图形的宽为10高为5

for images, labels in val_ds.take(1):
    for i in range(8):
        ax = plt.subplot(2, 4, i + 1)  
        
        # 显示图片
        plt.imshow(images[i])
        
        # 需要给图片增加一个维度
        img_array = tf.expand_dims(images[i], 0) 
        
        # 使用模型预测图片中的人物
        predictions = new_model.predict(img_array)
        plt.title(class_names[np.argmax(predictions)])

        plt.axis("off")
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19