VGG19实现fer2013人脸表情识别代码

以下是使用VGG19模型实现FER2013人脸表情识别的Python代码示例：

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import tensorflow as tf
from tensorflow.keras.preprocessing.image import ImageDataGenerator
from tensorflow.keras.applications.vgg19 import VGG19

# 加载数据
df = pd.read_csv('fer2013.csv')

# 将数据拆分为训练、验证和测试集
train_data = df[df['Usage'] == 'Training']
val_data = df[df['Usage'] == 'PublicTest']
test_data = df[df['Usage'] == 'PrivateTest']

# 创建数据生成器
train_datagen = ImageDataGenerator(rescale=1./255,
                                   rotation_range=10,
                                   width_shift_range=0.1,
                                   height_shift_range=0.1,
                                   shear_range=0.1,
                                   zoom_range=0.1,
                                   horizontal_flip=True,
                                   fill_mode='nearest')

val_datagen = ImageDataGenerator(rescale=1./255)

test_datagen = ImageDataGenerator(rescale=1./255)

# 定义批量大小和图像大小
batch_size = 32
img_size = (48, 48)

# 创建数据流
train_generator = train_datagen.flow_from_dataframe(train_data,
                                                    x_col='pixels',
                                                    y_col='emotion',
                                                    target_size=img_size,
                                                    batch_size=batch_size,
                                                    class_mode='categorical')

val_generator = val_datagen.flow_from_dataframe(val_data,
                                                x_col='pixels',
                                                y_col='emotion',
                                                target_size=img_size,
                                                batch_size=batch_size,
                                                class_mode='categorical')

test_generator = test_datagen.flow_from_dataframe(test_data,
                                                  x_col='pixels',
                                                  y_col='emotion',
                                                  target_size=img_size,
                                                  batch_size=batch_size,
                                                  class_mode='categorical')

# 加载预训练的VGG19模型
base_model = VGG19(input_shape=(48, 48, 3), include_top=False, weights='imagenet')

# 冻结预训练模型的所有层
for layer in base_model.layers:
    layer.trainable = False

# 添加自定义的分类层
x = base_model.output
x = tf.keras.layers.Flatten()(x)
x = tf.keras.layers.Dense(256, activation='relu')(x)
x = tf.keras.layers.Dropout(0.5)(x)
predictions = tf.keras.layers.Dense(7, activation='softmax')(x)

# 构建新模型
model = tf.keras.models.Model(inputs=base_model.input, outputs=predictions)

# 编译模型
model.compile(optimizer='adam',
              loss='categorical_crossentropy',
              metrics=['accuracy'])

# 训练模型
history = model.fit(train_generator,
                    steps_per_epoch=len(train_data) // batch_size,
                    epochs=50,
                    validation_data=val_generator,
                    validation_steps=len(val_data) // batch_size)

# 评估模型
test_loss, test_acc = model.evaluate(test_generator, verbose=2)
print('\nTest accuracy:', test_acc)

在这个示例中，我们使用了VGG19模型作为基础模型，并添加了自定义的分类层。我们还使用了预训练的权重，并冻结了所有预训练模型的层，只训练自定义分类层。这个模型可以用于FER2013数据集的人脸表情识别任务，并达到较高的准确率。