服装图像分类模型训练 - 使用 TensorFlow 和 Keras - 常规

import os
import numpy as np
import matplotlib.pyplot as plt
import tensorflow as tf
from tensorflow import keras
from tensorflow.keras import layers
from keras.layers.core import Dropout
# 定义训练集和测试集路径
train_dir = 'C:/Users/28938/Desktop/shengdu/Data/train'
test_dir = 'C:/Users/28938/Desktop/shengdu/Data/test'
# 定义类别标签
class_names = ['MEN_Coats', 'MEN_Hood', 'MEN_Suits', 'WOMEN_Dress', 'WOMEN_Hood']
# 定义图像尺寸和批次大小
img_height = 224
img_width = 224
batch_size = 32
# 从目录中读取训练集和测试集
train_ds = keras.preprocessing.image_dataset_from_directory(
    train_dir,
    validation_split=0.2,
    subset='training',
    seed=42,
    image_size=(img_height, img_width),
    batch_size=batch_size
)
val_ds = keras.preprocessing.image_dataset_from_directory(
    train_dir,
    validation_split=0.2,
    subset='validation',
    seed=42,
    image_size=(img_height, img_width),
    batch_size=batch_size
)
test_ds = keras.preprocessing.image_dataset_from_directory(
    test_dir,
    seed=42,
    image_size=(img_height, img_width),
    batch_size=batch_size
)
# 定义函数以展示图像
def plotImages(img_arr):
    fig, axes = plt.subplots(1, 5, figsize=(20,20))
    axes = axes.flatten()
    for img, ax in zip( img_arr, axes):
        ax.imshow(img)
        ax.axis('off')
    plt.tight_layout()
    plt.show()
# 定义数据增强器
data_augmentation = keras.Sequential(
    [
        layers.experimental.preprocessing.RandomFlip('horizontal', input_shape=(img_height, img_width, 3)),
        layers.experimental.preprocessing.RandomRotation(0.1),
        layers.experimental.preprocessing.RandomZoom(0.1),
    ]
)
# 定义模型输入
input_shape = (img_height, img_width, 3)
# 建立模型
model = keras.Sequential([
    data_augmentation,
    layers.experimental.preprocessing.Rescaling(1./255),
    layers.Conv2D(32, 3, padding='same', activation='relu', input_shape=input_shape),
    layers.MaxPooling2D(),
    layers.Dropout(0.2),
    layers.Conv2D(64, 3, padding='same', activation='relu'),
    layers.MaxPooling2D(),
    layers.Dropout(0.2),
    layers.Conv2D(128, 3, padding='same', activation='relu'),
    layers.MaxPooling2D(),
    layers.Dropout(0.2),
    layers.Flatten(),
    layers.Dense(256, activation='relu'),
    layers.Dense(len(class_names))
])
# 编译模型
model.compile(optimizer='adam',
              loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),
              metrics=['accuracy'])
# 打印模型结构
model.summary()
# 显示前5张图像
plt.figure(figsize=(10, 10))
for images, labels in train_ds.take(1):
    for i in range(5):
        ax = plt.subplot(5, 5, i + 1)
        plt.imshow(images[i].numpy().astype('uint8'))
        plt.title(class_names[labels[i]])
        plt.axis('off')
# 设定训练参数
epochs = 15
# 开始训练模型

#history = model.fit(train_ds, validation_data=val_ds, epochs=epochs)
batch_size = 16
history = model.fit(train_ds, validation_data=val_ds, epochs=epochs, batch_size=batch_size)
# 对模型进行评估
test_loss, test_acc = model.evaluate(test_ds)
print('Test accuracy:', test_acc)
# 对模型进行预测
predictions = model.predict(test_ds)
# 绘制模型准确度和损失随时间变化的曲线
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=(8, 8))
plt.subplot(2, 1, 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(2, 1, 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()

详细分析每句代码内容：

导入必要的库

import os
import numpy as np
import matplotlib.pyplot as plt
import tensorflow as tf
from tensorflow import keras
from tensorflow.keras import layers
from keras.layers.core import Dropout

定义训练集和测试集路径，以及类别标签

train_dir = 'C:/Users/28938/Desktop/shengdu/Data/train'
test_dir = 'C:/Users/28938/Desktop/shengdu/Data/test'
class_names = ['MEN_Coats', 'MEN_Hood', 'MEN_Suits', 'WOMEN_Dress', 'WOMEN_Hood']

定义图像尺寸和批次大小，从目录中读取训练集和测试集

img_height = 224
img_width = 224
batch_size = 32

train_ds = keras.preprocessing.image_dataset_from_directory(
    train_dir,
    validation_split=0.2,
    subset='training',
    seed=42,
    image_size=(img_height, img_width),
    batch_size=batch_size
)
val_ds = keras.preprocessing.image_dataset_from_directory(
    train_dir,
    validation_split=0.2,
    subset='validation',
    seed=42,
    image_size=(img_height, img_width),
    batch_size=batch_size
)
test_ds = keras.preprocessing.image_dataset_from_directory(
    test_dir,
    seed=42,
    image_size=(img_height, img_width),
    batch_size=batch_size
)

定义函数以展示图像

def plotImages(img_arr):
    fig, axes = plt.subplots(1, 5, figsize=(20,20))
    axes = axes.flatten()
    for img, ax in zip( img_arr, axes):
        ax.imshow(img)
        ax.axis('off')
    plt.tight_layout()
    plt.show()

定义数据增强器

data_augmentation = keras.Sequential(
    [
        layers.experimental.preprocessing.RandomFlip('horizontal', input_shape=(img_height, img_width, 3)),
        layers.experimental.preprocessing.RandomRotation(0.1),
        layers.experimental.preprocessing.RandomZoom(0.1),
    ]
)

定义模型输入，建立模型

input_shape = (img_height, img_width, 3)

model = keras.Sequential([
    data_augmentation,
    layers.experimental.preprocessing.Rescaling(1./255),
    layers.Conv2D(32, 3, padding='same', activation='relu', input_shape=input_shape),
    layers.MaxPooling2D(),
    layers.Dropout(0.2),
    layers.Conv2D(64, 3, padding='same', activation='relu'),
    layers.MaxPooling2D(),
    layers.Dropout(0.2),
    layers.Conv2D(128, 3, padding='same', activation='relu'),
    layers.MaxPooling2D(),
    layers.Dropout(0.2),
    layers.Flatten(),
    layers.Dense(256, activation='relu'),
    layers.Dense(len(class_names))
])

编译模型

model.compile(optimizer='adam',
              loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),
              metrics=['accuracy'])

打印模型结构，显示前5张图像

model.summary()

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

设定训练参数，开始训练模型

epochs = 15

batch_size = 16
history = model.fit(train_ds, validation_data=val_ds, epochs=epochs, batch_size=batch_size)

对模型进行评估，输出测试集准确度

test_loss, test_acc = model.evaluate(test_ds)
print('Test accuracy:', test_acc)

对模型进行预测，绘制模型准确度和损失随时间变化的曲线

predictions = model.predict(test_ds)

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=(8, 8))
plt.subplot(2, 1, 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(2, 1, 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()