基于Keras的Python实践 15 chapter 卷积神经网络分类cifar10

最新推荐文章于 2024-01-11 19:21:40 发布

wzw12315

最新推荐文章于 2024-01-11 19:21:40 发布

阅读量331

点赞数 1

分类专栏： keras

本文链接：https://blog.csdn.net/wzw12315/article/details/91819403

版权

keras 专栏收录该内容

20 篇文章 1 订阅

订阅专栏

1、大型卷积神经网络

import numpy as np
from keras.datasets import cifar10
from keras.models import Sequential
from keras.layers import Dense
from keras.layers import Dropout
from keras.layers import Flatten
from keras.layers.convolutional import Conv2D
from keras.layers.convolutional import MaxPooling2D
from keras.optimizers import SGD
from keras.constraints import maxnorm
from keras.utils import np_utils
from keras import backend
backend.set_image_data_format('channels_first')

# 设定随机种子
seed = 7
np.random.seed(seed=seed)

# 导入数据
(X_train, y_train), (X_validation, y_validation) = cifar10.load_data()

# 格式化数据到0-1之前
X_train = X_train.astype('float32')
X_validation = X_validation.astype('float32')
X_train = X_train / 255.0
X_validation = X_validation / 255.0

# one-hot编码
y_train = np_utils.to_categorical(y_train)
y_validation = np_utils.to_categorical(y_validation)
num_classes = y_train.shape[1]

def create_model(epochs=25):
    model = Sequential()
    model.add(Conv2D(32, (3, 3), input_shape=(3, 32, 32), padding='same', activation='relu', kernel_constraint=maxnorm(3)))
    model.add(Dropout(0.2))
    model.add(Conv2D(32, (3, 3), activation='relu', padding='same', kernel_constraint=maxnorm(3)))
    model.add(MaxPooling2D(pool_size=(2, 2)))
    model.add(Conv2D(64, (3, 3), activation='relu', padding='same', kernel_constraint=maxnorm(3)))
    model.add(Dropout(0.2))
    model.add(Conv2D(64, (3, 3), activation='relu', padding='same', kernel_constraint=maxnorm(3)))
    model.add(MaxPooling2D(pool_size=(2, 2)))
    model.add(Conv2D(128, (3, 3), activation='relu', padding='same', kernel_constraint=maxnorm(3)))
    model.add(Dropout(0.2))
    model.add(Conv2D(128, (3, 3), activation='relu', padding='same', kernel_constraint=maxnorm(3)))
    model.add(MaxPooling2D(pool_size=(2, 2)))
    model.add(Flatten())
    model.add(Dropout(0.2))
    model.add(Dense(1024, activation='relu', kernel_constraint=maxnorm(3)))
    model.add(Dropout(0.2))
    model.add(Dense(512, activation='relu', kernel_constraint=maxnorm(3)))
    model.add(Dropout(0.2))
    model.add(Dense(10, activation='softmax'))
    # 默认情况下，这里的学习率是线性衰减的
    # learningRate =  learningRate * 1/(1 + decay*epoch)
    # 当decay衰减为0时（默认值），对学习率没有影响；当使用非0学习率衰减时，学习率呈线性衰减
    # decay不为0的情况下，学习率随着epochs的增大而减少
    lrate = 0.01
    decay = lrate / epochs
    sgd = SGD(lr=lrate, momentum=0.9, decay=decay, nesterov=False)
    model.compile(loss='categorical_crossentropy', optimizer=sgd, metrics=['accuracy'])
    return model

epochs = 25
model = create_model(epochs)
model.fit(x=X_train, y=y_train, epochs=epochs, batch_size=32, verbose=2)
scores = model.evaluate(x=X_validation, y=y_validation, verbose=0)
print('Accuracy: %.2f%%' % (scores[1] * 100))

改进型模型

import keras
import numpy as np
from keras.datasets import cifar10
from keras.models import Sequential
from keras.layers import Dropout, Activation
from keras.layers import Conv2D, MaxPooling2D, GlobalAveragePooling2D
from keras.initializers import RandomNormal
from keras import optimizers
from keras.callbacks import LearningRateScheduler, TensorBoard

batch_size = 128
epochs = 200
iterations = 391
num_classes = 10
dropout = 0.5
log_filepath = './nin'


def normalize_preprocessing(x_train, x_validation):

    #首先把x_train和x_validation设置数据类型为float32
    x_train = x_train.astype('float32')
    x_validation = x_validation.astype('float32')
    mean = [125.307, 122.95, 113.865]
    std = [62.9932, 62.0887, 66.7048]
    #然后分别对每一个图像数据的每一个通道进行归一化操作
    for i in range(3):
        x_train[:, :, :, i] = (x_train[:, :, :, i] - mean[i]) / std[i]
        x_validation[:, :, :, i] = (x_validation[:, :, :, i] - mean[i]) / std[i]

    return x_train, x_validation


def scheduler(epoch):
    if epoch <= 60:
        return 0.05
    if epoch <= 120:
        return 0.01
    if epoch <= 160:
        return 0.002
    return 0.0004


def build_model():
    model = Sequential()

    model.add(Conv2D(192, (5, 5), padding='same', kernel_regularizer=keras.regularizers.l2(0.0001),
                     kernel_initializer=RandomNormal(stddev=0.01), input_shape=x_train.shape[1:],
                     activation='relu'))
    model.add(Conv2D(160, (1, 1), padding='same', kernel_regularizer=keras.regularizers.l2(0.0001),
                     kernel_initializer=RandomNormal(stddev=0.05), activation='relu'))
    model.add(Conv2D(96, (1, 1), padding='same', kernel_regularizer=keras.regularizers.l2(0.0001),
                     kernel_initializer=RandomNormal(stddev=0.05), activation='relu'))
    model.add(MaxPooling2D(pool_size=(3, 3), strides=(2, 2), padding='same'))

    model.add(Dropout(dropout))

    model.add(Conv2D(192, (5, 5), padding='same', kernel_regularizer=keras.regularizers.l2(0.0001),
                     kernel_initializer=RandomNormal(stddev=0.05), activation='relu'))
    model.add(Conv2D(192, (1, 1), padding='same', kernel_regularizer=keras.regularizers.l2(0.0001),
                     kernel_initializer=RandomNormal(stddev=0.05), activation='relu'))
    model.add(Conv2D(192, (1, 1), padding='same', kernel_regularizer=keras.regularizers.l2(0.0001),
                     kernel_initializer=RandomNormal(stddev=0.05), activation='relu'))
    model.add(MaxPooling2D(pool_size=(3, 3), strides=(2, 2), padding='same'))

    model.add(Dropout(dropout))

    model.add(Conv2D(192, (3, 3), padding='same', kernel_regularizer=keras.regularizers.l2(0.0001),
                     kernel_initializer=RandomNormal(stddev=0.05), activation='relu'))
    model.add(Conv2D(192, (1, 1), padding='same', kernel_regularizer=keras.regularizers.l2(0.0001),
                     kernel_initializer=RandomNormal(stddev=0.05), activation='relu'))
    #在这里并没有像传统一样把activation设置为softmax
    model.add(Conv2D(10, (1, 1), padding='same', kernel_regularizer=keras.regularizers.l2(0.0001),
                     kernel_initializer=RandomNormal(stddev=0.05), activation='relu'))

    #
    model.add(GlobalAveragePooling2D())
    # 单独增加一个激活层
    model.add(Activation('softmax'))

    sgd = optimizers.SGD(lr=0.1, momentum=0.9, nesterov=True)
    model.compile(loss='categorical_crossentropy', optimizer=sgd, metrics=['accuracy'])
    return model


if __name__ == '__main__':
    np.random.seed(seed=7)
    # load data
    (x_train, y_train), (x_validation, y_validation) = cifar10.load_data()
    y_train = keras.utils.to_categorical(y_train, num_classes)
    y_validation = keras.utils.to_categorical(y_validation, num_classes)

    x_train, x_validation = normalize_preprocessing(x_train, x_validation)

    # build network
    model = build_model()
    print(model.summary())

    # set callback
    # 使用了回调函数
    tb_cb = TensorBoard(log_dir=log_filepath, histogram_freq=0)
    change_lr = LearningRateScheduler(scheduler)
    cbks = [change_lr, tb_cb]


    # set data augmentation
    print('Using real-time data augmentation.')
    from keras.preprocessing.image import ImageDataGenerator
    datagen = ImageDataGenerator(horizontal_flip=True, width_shift_range=0.125, height_shift_range=0.125,
                                 fill_mode='constant', cval=0.)
    datagen.fit(x_train)

    # start training
    model.fit_generator(datagen.flow(x_train, y_train, batch_size=batch_size), steps_per_epoch=iterations,
                        epochs=epochs, callbacks=cbks, validation_data=(x_validation, y_validation), verbose=2)

    #model.fit(x_train, y_train, batch_size=batch_size, epochs=epochs, callbacks=cbks,
    #         validation_data=(x_validation, y_validation), verbose=2)
    #保存模型
    model.save('nin.h5')