基于python与resnet_model的迁移学习设计与实现

from keras.optimizers import SGD from keras.layers import Input, Dense, Conv2D, MaxPooling2D, AveragePooling2D, ZeroPadding2D, add, Flatten, Activation from keras.layers.normalization import BatchNormalization from keras.models import Model import matplotlib.pyplot as plt import numpy as np from keras.utils import plot_model from sklearn.metrics import log_loss from load_data import load_cifar10_data def identity_block(input_tensor, kernel_size, filters, stage, block): """ 直接相加，并不需要1*1卷积 参数： input_tensor: 输入 kernel_size: 卷积核大小 filters: 卷积核个数，需要按顺序指定3个，例如（64,64,256） stage和block:主要为了绘图观察，指定好名字 """ nb_filter1, nb_filter2, nb_filter3 = filters conv_name_base = 'res' + str(stage) + block + '_branch' bn_name_base = 'bn' + str(stage) + block + '_branch' x = Conv2D(nb_filter1, (1, 1), name=conv_name_base + '2a')(input_tensor) x = BatchNormalization(axis=bn_axis, name=bn_name_base + '2a')(x) x = Activation('relu')(x) x = Conv2D(nb_filter2, (kernel_size, kernel_size), padding='same', name=conv_name_base + '2b')(x) x = BatchNormalization(axis=bn_axis, name=bn_name_base + '2b')(x) x = Activation('relu')(x) x = Conv2D(nb_filter3, (1, 1), name=conv_name_base + '2c')(x) x = BatchNormalization(axis=bn_axis, name=bn_name_base + '2c')(x) x = add([x, input_tensor], name = conv_name_base +'_add') x = Activation('relu')(x) return x def conv_block(input_tensor, kernel_size, filters, stage, block, strides=(2, 2)): """ 需要1*1卷积 输入参数与identity_block一致 """ nb_filter1, nb_filter2, nb_filter3 = filters conv_name_base = 'res' + str(stage) + block + '_branch' bn_name_base = 'bn' + str(stage) + block + '_branch' x = Conv2D(nb_filter1, (1, 1), strides=strides, name=conv_name_base + '2a')(input_tensor) x = BatchNormalization(axis=bn_axis, name=bn_name_base + '2a')(x) x = Activation('relu')(x) x = Conv2D(nb_filter2, (kernel_size, kernel_size), padding='same', name=conv_name_base + '2b')(x) x = BatchNormalization(axis=bn_axis, name=bn_name_base + '2b')(x) x = Activation('relu')(x) x = Conv2D(nb_filter3, (1, 1), name=conv_name_base + '2c')(x) x = BatchNormalization(axis=bn_axis, name=bn_name_base + '2c')(x) shortcut = Conv2D(nb_filter3, (1, 1), strides=strides, name=conv_name_base + '1')(input_tensor) shortcut = BatchNormalization(axis=bn_axis, name=bn_name_base + '1')(shortcut) x = add([x, shortcut], name = conv_name_base +'_add') x = Activation('relu')(x) return x def resnet50_model(img_rows, img_cols, color_type=3, num_classes=None): """ 网络模型定义，需要下载好权重参数文件 """ #按照Tensorflow格式 global bn_axis bn_axis = 3 img_input = Input(shape=(img_rows, img_cols, color_type)) #网络模型定义 x = ZeroPadding2D((3, 3))(img_input) x = Conv2D(64, (7, 7), strides=(2, 2), name='conv1')(x) x = BatchNormalization(axis=bn_axis, name='bn_conv1')(x) x = Activation('relu')(x) x = MaxPooling2D((3, 3), strides=(2, 2))(x) x = conv_block(x, 3, [64, 64, 256], stage=2, block='a', strides=(1, 1)) x = identity_block(x, 3, [64, 64, 256], stage=2, block='b') x = identity_block(x, 3, [64, 64, 256], stage=2, block='c') x = conv_block(x, 3, [128, 128, 512], stage=3, block='a') x = identity_block(x, 3, [128, 128, 512], stage=3, block='b') x = identity_block(x, 3, [128, 128, 512], stage=3, block='c') x = identity_block(x, 3, [128, 128, 512], stage=3, block='d') x = conv_block(x, 3, [256, 256, 1024], stage=4, block='a') x = identity_block(x, 3, [256, 256, 1024], stage=4, block='b') x = identity_block(x, 3, [256, 256, 1024], stage=4, block='c') x = identity_block(x, 3, [256, 256, 1024], stage=4, block='d') x = identity_block(x, 3, [256, 256, 1024], stage=4, block='e') x = identity_block(x, 3, [256, 256, 1024], stage=4, block='f') x = conv_block(x, 3, [512, 512, 2048], stage=5, block='a') x = identity_block(x, 3, [512, 512, 2048], stage=5, block='b') x = identity_block(x, 3, [512, 512, 2048], stage=5, block='c') # 全连接层 x_fc = AveragePooling2D((7, 7), name='avg_pool')(x) x_fc = Flatten()(x_fc) x_fc = Dense(1000, activation='softmax', name='fc1000')(x_fc) # 创建模型 model = Model(img_input, x_fc) # 加载权重参数 """ for layer in model.layers: layer.trainable = False model.layers[:100] """ weights_path = 'resnet50_weights_tf_dim_ordering_tf_kernels.h5' model.load_weights(weights_path) # 迁移学习时全连接层用自己的开始学习 x_newfc = AveragePooling2D((7, 7), name='avg_pool')(x) x_newfc = Flatten()(x_newfc) x_newfc = Dense(num_classes, activation='softmax', name='fc10')(x_newfc) # 创建新的模型 model = Model(img_input, x_newfc) # 训练 sgd = SGD(lr=1e-3, decay=1e-6, momentum=0.9, nesterov=True) model.compile(optimizer=sgd, loss='categorical_crossentropy', metrics=['accuracy']) return model if __name__ == '__main__': # 定义好输入的数据 img_rows, img_cols = 224, 224 # h,w channel = 3 num_classes = 10 batch_size = 16 nb_epoch = 10 # 加载数据 X_train, Y_train, X_valid, Y_valid = load_cifar10_data(img_rows, img_cols) # 网络模型 model = resnet50_model(img_rows, img_cols, channel, num_classes) plot_model(model, to_file='model.png',show_shapes=True) # Fine-tuning H = model.fit(X_train, Y_train, batch_size=batch_size, epochs=nb_epoch, shuffle=True, verbose=1, validation_data=(X_valid, Y_valid), ) plt.style.use("ggplot") plt.figure() plt.ylim(0,3) N = nb_epoch plt.plot(np.arange(0, N), H.history["loss"], label="train_loss") plt.plot(np.arange(0, N), H.history["val_loss"], label="val_loss") plt.plot(np.arange(0, N), H.history["acc"], label="train_acc") plt.plot(np.arange(0, N), H.history["val_acc"], label="val_acc") plt.title("Training Loss and Accuracy") plt.xlabel("Epoch #") plt.ylabel("Loss/Accuracy") plt.legend(loc="upper left") plt.savefig('result.png')