基于卷积神经网络的人脸表情识别代码

以下是基于卷积神经网络的人脸表情识别的代码：

import numpy as np
import pandas as pd
import keras
from keras.layers import Conv2D, MaxPooling2D, Dense, Flatten, Dropout
from keras.models import Sequential

# 加载数据
data = pd.read_csv("fer2013.csv")

# 将像素值转换为图像
def string_to_image(string):
    return np.array([int(x) for x in string.split()]).reshape(48, 48, 1) / 255.0

# 将数据转换为图像数据
data['image'] = data['pixels'].apply(string_to_image)

# 将标签转换为独热编码
def emotion_to_vector(emotion):
    if emotion == 0:
        return [1, 0, 0, 0, 0, 0, 0]
    elif emotion == 1:
        return [0, 1, 0, 0, 0, 0, 0]
    elif emotion == 2:
        return [0, 0, 1, 0, 0, 0, 0]
    elif emotion == 3:
        return [0, 0, 0, 1, 0, 0, 0]
    elif emotion == 4:
        return [0, 0, 0, 0, 1, 0, 0]
    elif emotion == 5:
        return [0, 0, 0, 0, 0, 1, 0]
    elif emotion == 6:
        return [0, 0, 0, 0, 0, 0, 1]

data['emotion_vec'] = data['emotion'].apply(emotion_to_vector)

# 划分训练集、验证集和测试集
train_data = data[data['Usage'] == 'Training'][['image', 'emotion_vec']]
val_data = data[data['Usage'] == 'PrivateTest'][['image', 'emotion_vec']]
test_data = data[data['Usage'] == 'PublicTest'][['image', 'emotion_vec']]

# 建立模型
model = Sequential()
model.add(Conv2D(32, kernel_size=(3, 3), activation='relu', input_shape=(48, 48, 1)))
model.add(Conv2D(64, kernel_size=(3, 3), activation='relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.25))
model.add(Conv2D(128, kernel_size=(3, 3), activation='relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Conv2D(128, kernel_size=(3, 3), activation='relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.25))
model.add(Flatten())
model.add(Dense(1024, activation='relu'))
model.add(Dropout(0.5))
model.add(Dense(7, activation='softmax'))

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

# 训练模型
history = model.fit(np.array([x for x in train_data['image']]), np.array([x for x in train_data['emotion_vec']]), 
                    validation_data=(np.array([x for x in val_data['image']]), np.array([x for x in val_data['emotion_vec']])), 
                    epochs=50, batch_size=128)

# 评估模型
score = model.evaluate(np.array([x for x in test_data['image']]), np.array([x for x in test_data['emotion_vec']]), verbose=0)
print('Test loss:', score[0])
print('Test accuracy:', score[1])

这段代码使用了Keras库来建立卷积神经网络模型，使用FER2013数据集进行训练和测试，实现人脸表情识别功能。