RCNN的python实现

前言

1.关于RCNN的详细介绍可见我的另一篇博客：
https://blog.csdn.net/u014796085/article/details/83352821
2.由于项目比较大，并且需要相关数据集，所以这篇博客的代码不适用于需要整个项目的人，这里仅把项目主体部分拆分解读，来为对项目某部分有疑问的人提供参考，并且对本人的代码工作作一个记录，方便以后复习。
3.该程序是在github相关项目基础上修改而来，参考的github项目：
https://github.com/yangxue0827/RCNN

代码实现与解读

大训练集预训练

# Building 'AlexNet'
def create_alexnet(num_classes):
    network = input_data(shape=[None, config.IMAGE_SIZE, config.IMAGE_SIZE, 3])
    # 4维输入张量，卷积核个数，卷积核尺寸，步长
    network = conv_2d(network, 96, 11, strides=4, activation='relu')
    network = max_pool_2d(network, 3, strides=2)
    # 数据归一化
    network = local_response_normalization(network)
    network = conv_2d(network, 256, 5, activation='relu')
    network = max_pool_2d(network, 3, strides=2)
    network = local_response_normalization(network)
    network = conv_2d(network, 384, 3, activation='relu')
    network = conv_2d(network, 384, 3, activation='relu')
    network = conv_2d(network, 256, 3, activation='relu')
    network = max_pool_2d(network, 3, strides=2)
    network = local_response_normalization(network)
    network = fully_connected(network, 4096, activation='tanh')
    network = dropout(network, 0.5)
    network = fully_connected(network, 4096, activation='tanh')
    network = dropout(network, 0.5)
    network = fully_connected(network, num_classes, activation='softmax')
    momentum = tflearn.Momentum(learning_rate=0.001, lr_decay=0.95, decay_step=200)
    network = regression(network, optimizer=momentum,
                         loss='categorical_crossentropy')
    return network

定义alexnet网络，这里num_classes是大训练集对应的分类数量。

def load_data(datafile, num_class, save=False, save_path='dataset.pkl'):
    fr = codecs.open(datafile, 'r', 'utf-8')
    train_list = fr.readlines()
    labels = []
    images = []
    # 对每一个训练样本
    for line in train_list:
        tmp = line.strip().split(' ')
        fpath = tmp[0]
        img = cv2.imread(fpath)
        # 样本resize到227x227，转为矩阵保存
        img = prep.resize_image(img, config.IMAGE_SIZE, config.IMAGE_SIZE)
        np_img = np.asarray(img, dtype="float32")
        images.append(np_img)

        index = int(tmp[1])
        label = np.zeros(num_class)
        label[index] = 1
        labels.append(label)
    if save:
        # 序列化保存
        pickle.dump((images, labels), open(save_path, 'wb'))
    fr.close()
    return images, labels

样本预处理，所有样本resize后转矩阵保存。

def train(network, X, Y, save_model_path):
    # Training
    model = tflearn.DNN(network, checkpoint_path='model_alexnet',
                        max_checkpoints=1, tensorboard_verbose=2, tensorboard_dir='output')
    if os.path.isfile(save_model_path + '.index'):
        model.load(save_model_path)
        print('load model...')

    model.fit(X, Y, n_epoch=200, validation_set=0.1, shuffle=True,
              show_metric=True, batch_size=64, snapshot_step=200,
              snapshot_epoch=False, run_id='alexnet_oxflowers17') # epoch = 1000
    # Save the model
    model.save(save_model_path)
    print('save model...')

训练模型，X、Y为样本，迭代次数200次，训练集中取出10%作为验证集（用来计算模型预测正确率），每次迭代训练64个样本。

小数据集微调

# Use a already trained alexnet with the last layer redesigned
def create_alexnet(num_classes, restore=False):
    # Building 'AlexNet'
    network = input_data(shape=[None, config.IMAGE_SIZE, config.IMAGE_SIZE, 3])
    network = conv_2d(network, 96, 11, strides=4, activation='relu')
    network = max_pool_2d(network, 3, strides=2)
    network = local_response_normalization(network)
    network = conv_2d(network, 256, 5, activation='relu')
    network = max_pool_2d(netw