文章目录
- 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：
- 2. 导入数据
- 3. 查看数据
- 二、数据预处理
- 1. 加载数据
- 2. 可视化数据
- 3. 再次检查数据
- 4. 配置数据集
- 三、残差网络（ResNet）介绍
- 1. 残差网络解决了什么
- 2. ResNet-50介绍
- 四、构建ResNet-50网络模型
- 五、编译
- 六、训练模型
- 六、模型评估
- 八、保存and加载模型
- 九、预测
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")
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
data_dir = "weather_photos/"
data_dir = pathlib.Path(data_dir)
3. 查看数据
image_count = len(list(data_dir.glob('*/*')))
print("图片总数为：",image_count)
二、数据预处理
| 文件夹 | 数量 |
|---|---|
| Bananaquit | 166 张 |
| Black Throated Bushtiti | 111 张 |
| Black skimmer | 122 张 |
| Cockatoo | 166张 |
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
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)
Found 565 files belonging to 4 classes.
Using 452 files for training.
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)
Found 565 files belonging to 4 classes.
Using 113 files for validation.
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)
['Bananaquit', 'Black Skimmer', 'Black Throated Bushtiti', 'Cockatoo']
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")
plt.imshow(images[1].numpy().astype("uint8"))
3. 再次检查数据
for image_batch, labels_batch in train_ds:
print(image_batch.shape)
print(labels_batch.shape)
break
(8, 224, 224, 3)
(8,)
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)
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()
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
__________________________________________________________________________________________________
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'])
六、训练模型
epochs = 10
history = model.fit(
train_ds,
validation_data=val_ds,
epochs=epochs
)
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
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()
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')
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")
