第T5周：运动鞋品牌识别

>- **🍨 本文为[🔗365天深度学习训练营](https://mp.weixin.qq.com/s/0dvHCaOoFnW8SCp3JpzKxg) 中的学习记录博客**
>- **🍖 原作者：[K同学啊](https://mtyjkh.blog.csdn.net/)**

一、前期工作

1. 设置GPU

如果使用的是CPU可以忽略这步

from tensorflow       import keras
from tensorflow.keras import layers,models
import os, PIL, pathlib
import matplotlib.pyplot as plt
import tensorflow        as tf

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

if gpus:
    gpu0 = gpus[0]                                        #如果有多个GPU，仅使用第0个GPU
    tf.config.experimental.set_memory_growth(gpu0, True)  #设置GPU显存用量按需使用
    tf.config.set_visible_devices([gpu0],"GPU")
    
gpus

[PhysicalDevice(name='/physical_device:GPU:0', device_type='GPU')] 

 2. 导入数据

data_dir = "./data/"

data_dir = pathlib.Path(data_dir)

3. 查看数据

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

print("图片总数为：",image_count)

图片总数为： 578 

 二、数据预处理

batch_size = 32
img_height = 224
img_width = 224
 
"""
关于image_dataset_from_directory()的详细介绍可以参考文章：https://mtyjkh.blog.csdn.net/article/details/117018789
"""
train_ds = tf.keras.preprocessing.image_dataset_from_directory(
    "./data/train/",
    seed=123,
    image_size=(img_height, img_width),
    batch_size=batch_size)
 
"""
关于image_dataset_from_directory()的详细介绍可以参考文章：https://mtyjkh.blog.csdn.net/article/details/117018789
"""
val_ds = tf.keras.preprocessing.image_dataset_from_directory(
    "./data/test/",
    seed=123,
    image_size=(img_height, img_width),
    batch_size=batch_size)

class_names = train_ds.class_names
print(class_names)

['adidas', 'nike'] 

 可视化数据：

plt.figure(figsize=(20, 10))
 
for images, labels in train_ds.take(1):
    for i in range(20):
        ax = plt.subplot(5, 10, i + 1)
 
        plt.imshow(images[i].numpy().astype("uint8"))
        plt.title(class_names[labels[i]])
        
        plt.axis("off")
plt.show()

 

三、构建CNN网络

"""
关于卷积核的计算不懂的可以参考文章：https://blog.csdn.net/qq_38251616/article/details/114278995

layers.Dropout(0.4) 作用是防止过拟合，提高模型的泛化能力。
关于Dropout层的更多介绍可以参考文章：https://mtyjkh.blog.csdn.net/article/details/115826689
"""

model = models.Sequential([
    layers.experimental.preprocessing.Rescaling(1./255, input_shape=(img_height, img_width, 3)),
    
    layers.Conv2D(16, (3, 3), activation='relu', input_shape=(img_height, img_width, 3)), # 卷积层1，卷积核3*3  
    layers.AveragePooling2D((2, 2)),               # 池化层1，2*2采样
    layers.Conv2D(32, (3, 3), activation='relu'),  # 卷积层2，卷积核3*3
    layers.AveragePooling2D((2, 2)),               # 池化层2，2*2采样
    layers.Dropout(0.3),  
    layers.Conv2D(64, (3, 3), activation='relu'),  # 卷积层3，卷积核3*3
    layers.Dropout(0.3),  
    
    layers.Flatten(),                       # Flatten层，连接卷积层与全连接层
    layers.Dense(128, activation='relu'),   # 全连接层，特征进一步提取
    layers.Dense(len(class_names))               # 输出层，输出预期结果
])

model.summary()  # 打印网络结构

Model: "sequential"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
rescaling (Rescaling)        (None, 224, 224, 3)       0         
_________________________________________________________________
conv2d (Conv2D)              (None, 222, 222, 16)      448       
_________________________________________________________________
average_pooling2d (AveragePo (None, 111, 111, 16)      0         
_________________________________________________________________
conv2d_1 (Conv2D)            (None, 109, 109, 32)      4640      
_________________________________________________________________
average_pooling2d_1 (Average (None, 54, 54, 32)        0         
_________________________________________________________________
dropout (Dropout)            (None, 54, 54, 32)        0         
_________________________________________________________________
conv2d_2 (Conv2D)            (None, 52, 52, 64)        18496     
_________________________________________________________________
dropout_1 (Dropout)          (None, 52, 52, 64)        0         
_________________________________________________________________
flatten (Flatten)            (None, 173056)            0         
_________________________________________________________________
dense (Dense)                (None, 128)               22151296  
_________________________________________________________________
dense_1 (Dense)              (None, 2)                 258       
=================================================================
Total params: 22,175,138
Trainable params: 22,175,138
Non-trainable params: 0
_________________________________________________________________ 

四、训练模型

设置动态学习率

# 设置初始学习率
initial_learning_rate = 0.001
 
lr_schedule = tf.keras.optimizers.schedules.ExponentialDecay(
        initial_learning_rate, 
        decay_steps=10,      # 敲黑板！！！这里是指 steps，不是指epochs
        decay_rate=0.92,     # lr经过一次衰减就会变成 decay_rate*lr
        staircase=True)
 
# 将指数衰减学习率送入优化器
optimizer = tf.keras.optimizers.Adam(learning_rate=lr_schedule)
 
model.compile(optimizer=optimizer,
              loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),
              metrics=['accuracy'])

早停与保存最佳模型参数 

from tensorflow.keras.callbacks import ModelCheckpoint, EarlyStopping

epochs = 50

# 保存最佳模型参数
checkpointer = ModelCheckpoint('best_model.h5',
                                monitor='val_accuracy',
                                verbose=1,
                                save_best_only=True,
                                save_weights_only=True)

# 设置早停
earlystopper = EarlyStopping(monitor='val_accuracy', 
                             min_delta=0.001,
                             patience=20, 
                             verbose=1)

 训练模型

history = model.fit(train_ds,
                    validation_data=val_ds,
                    epochs=epochs,
                    callbacks=[checkpointer, earlystopper])

 

Epoch 1/50
16/16 [==============================] - 4s 31ms/step - loss: 3.5439 - accuracy: 0.4721 - val_loss: 0.6931 - val_accuracy: 0.5789

Epoch 00001: val_accuracy improved from -inf to 0.57895, saving model to best_model.h5
Epoch 2/50
16/16 [==============================] - 0s 12ms/step - loss: 0.6929 - accuracy: 0.5279 - val_loss: 0.6891 - val_accuracy: 0.6447

......

Epoch 00040: val_accuracy did not improve from 0.89474
Epoch 41/50
16/16 [==============================] - 0s 12ms/step - loss: 0.0931 - accuracy: 0.9841 - val_loss: 0.3837 - val_accuracy: 0.8816

Epoch 00041: val_accuracy did not improve from 0.89474
Epoch 42/50
16/16 [==============================] - 0s 12ms/step - loss: 0.0871 - accuracy: 0.9801 - val_loss: 0.3834 - val_accuracy: 0.8816

Epoch 00042: val_accuracy did not improve from 0.89474
Epoch 00042: early stopping

五、模型评估

1. Loss与Accuracy图

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

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

epochs_range = range(len(loss))

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()

  

2. 指定图片进行预测

# 加载效果最好的模型权重
model.load_weights('best_model.h5')

from PIL import Image
import numpy as np


img = Image.open("./data/test/nike/1.jpg")  #这里选择你需要预测的图片
image = tf.image.resize(img, [img_height, img_width])

img_array = tf.expand_dims(image, 0) #/255.0  # 记得做归一化处理（与训练集处理方式保持一致）

predictions = model.predict(img_array) # 这里选用你已经训练好的模型
print("预测结果为：",class_names[np.argmax(predictions)])

预测结果为： nike