Keras 使用自己的数据分类，并使用tensorboard记录的简单实例

Keras 使用自己的数据分类，并使用tensorboard记录的简单实例

1.使用的分类图片按照不同类别保存在不同文件夹子中，并且切分好训练集和测试集，如下图显示

注意：文件名建议使用标签名

from keras.models import Sequentialimport kerasfrom keras.preprocessing.image import ImageDataGeneratorfrom keras.layers import Dense, Dropout, Flattenfrom keras.layers import Conv2D, MaxPooling2Dfrom keras.callbacks import TensorBoardimport timeimport osimport tensorflow as tf#制定GPU，限制GPU内存os.environ["CUDA_VISIBLE_DEVICES"] = "0"from keras.backend.tensorflow_backend import set_sessionconfig = tf.ConfigProto()config.gpu_options.per_process_gpu_memory_fraction = 0.7set_session(tf.Session(config=config))BATCHSIZE = 100IMG_SIZE = (100,100)#训练集，测试集文件路径train_path = '../data/train'test_path = '../data/test's_time = time.strftime("%Y%m%d%H%M%S", time.localtime())  #时间戳#image_batch_generatortrain_datagen = ImageDataGenerator(         rescale=1./255)test_datagen = ImageDataGenerator( rescale=1./255 )#训练集batch生成器train_generator = train_datagen.flow_from_directory(        train_path,        target_size=IMG_SIZE,        batch_size=BATCHSIZE,        color_mode='grayscale',        classes=['original','tampered'],        class_mode='categorical')#测试集batch生成器validation_generator = test_datagen.flow_from_directory(        test_path,        target_size=IMG_SIZE,        color_mode='grayscale',        batch_size=BATCHSIZE,        classes=['original','tampered'],        class_mode='categorical')#网络结构model = Sequential()model.add(Conv2D(16, (3, 3), activation='relu', input_shape=(IMG_SIZE[0],IMG_SIZE[1],1)))model.add(Conv2D(16, (3, 3), activation='relu'))model.add(MaxPooling2D(pool_size=(2, 2)))model.add(Dropout(0.25))model.add(Conv2D(32, (3, 3), activation='relu'))#model.add(Conv2D(64, (3, 3), activation='relu'))model.add(MaxPooling2D(pool_size=(2, 2)))model.add(Dropout(0.25))model.add(Flatten())model.add(Dense(128, activation='relu'))model.add(Dropout(0.5))model.add(Dense(2, activation='softmax'))#优化器adam = keras.optimizers.Adam(lr=0.001, beta_1=0.9, beta_2=0.999, epsilon=1e-08)model.compile(loss='binary_crossentropy',              optimizer=adam,              metrics=['accuracy'])#logs文件路径logs_path = 'F:/zy/logs/log_%s'%(s_time)try:    os.makedirs(logs_path)except:    pass#将loss ，acc， val_loss ,val_acc记录tensorboardtensorboard = TensorBoard(log_dir=logs_path, histogram_freq=1,write_graph=True,write_batch_performance=True)#模型训练model.fit_generator(        train_generator,        steps_per_epoch=60,        epochs=50,        verbose=1,        validation_data=validation_generator,        validation_steps=60,        callbacks=[tensorboard])