TensorFlow-2.x-10-AlexNet的实现

        AlexNet可以说是使得深度学习大火的一个网络模型，它构建了多层CNN，增大感受野，加入LRN等方法，使得网络拥有极高的特征提取能力，但受限于设备的元素，当时用了2块GPU来提升训练效率。相关结构及原理可参考一下这篇文章：

深入理解AlexNet网络

        本章使用tf-2.x来实现AlexNet，由于设备的原因，修改了全连接层神经元个数、输出层个数。又因为LRN对模型的提升作用不大，且容易造成模型训练难以收敛等因素，所以并没有使用LRN。

1、读取数据集并Resize

        下面的代码随机从数据集中选取batch_size个训练数据和测试数据，并将图片大小resize成224*224大小。

class Data_load():
    def __init__(self):
        fashion_mnist = tf.keras.datasets.fashion_mnist
        (self.train_images, self.train_labels), (self.test_images, self.test_labels)\
            = fashion_mnist.load_data()
        # 数据维度扩充成[n,h,w,c]的模式
        self.train_images = np.expand_dims(self.train_images.astype(np.float32) / 255.0, axis=-1)
        self.test_images = np.expand_dims(self.test_images.astype(np.float32)/255.0,axis=-1)
        # 标签
        self.train_labels = self.train_labels.astype(np.int32)
        self.test_labels = self.test_labels.astype(np.int32)
        # 训练和测试的数据个数
        self.num_train, self.num_test = self.train_images.shape[0], self.test_images.shape[0]
    def get_train_batch(self,batch_size):
        # 随机取batch_size个索引
        index = np.random.randint(0, np.shape(self.train_images)[0], batch_size)
        # resize
        resized_images = tf.image.resize_with_pad(self.train_images[index], 224, 224 )
        return resized_images.numpy(), self.train_labels[index]
    def get_test_batch(self,batch_size):
        index = np.random.randint(0, np.shape(self.test_images)[0], batch_size)
        # resize
        resized_images = tf.image.resize_with_pad(self.test_images[index], 224, 224 )
        return resized_images.numpy(), self.test_labels[index]

2、定义模型

根据AlexNet的网络结构，定义相关模型结构：

def AlexNet():
    net=tf.keras.Sequential()
    net.add(tf.keras.layers.Conv2D(96,11,(4,4),"same",activation="relu"))
    net.add(tf.keras.layers.MaxPool2D(pool_size=3, strides=2))
    net.add(tf.keras.layers.Conv2D(filters=256, kernel_size=5, padding='same', activation='relu'))
    net.add(tf.keras.layers.MaxPool2D(pool_size=3, strides=2))
    net.add(tf.keras.layers.Conv2D(filters=384, kernel_size=3, padding='same', activation='relu'))
    net.add(tf.keras.layers.Conv2D(filters=384, kernel_size=3, padding='same', activation='relu'))
    net.add(tf.keras.layers.Conv2D(filters=256, kernel_size=3, padding='same', activation='relu'))
    net.add(tf.keras.layers.MaxPool2D(pool_size=3, strides=2))
    net.add(tf.keras.layers.Flatten())
    net.add(tf.keras.layers.Dense(512, activation='relu'))# 为了方便训练 神经元个数改小，原来是1024
    net.add(tf.keras.layers.Dropout(0.5))
    net.add(tf.keras.layers.Dense(256, activation='relu'))# 为了方便训练 神经元个数改小，原来是1024
    net.add(tf.keras.layers.Dropout(0.5))
    net.add(tf.keras.layers.Dense(10, activation='sigmoid'))

    return net

3、训练

训练比较简单，利用keras方法即可：

def train(num_epoches,net):
    optimizer = tf.keras.optimizers.SGD(learning_rate=0.01, momentum=0.0, nesterov=False)
    net.compile(optimizer=optimizer,
                  loss='sparse_categorical_crossentropy',
                  metrics=['accuracy'])

    num_iter = data_load.num_train // batch_size
    for e in range(num_epoches):
        for n in range(num_iter):
            x_batch, y_batch = data_load.get_train_batch(batch_size)
            test_x_batch, test_y_batch = data_load.get_test_batch(batch_size)
            net.fit(x_batch, y_batch,validation_data=(test_x_batch, test_y_batch))
train(5,net)

训练结果：

附上所有源码：

import tensorflow as tf
from tensorflow.keras.datasets import fashion_mnist
from tensorflow import data as tfdata
import numpy as np

# 将 GPU 的显存使用策略设置为 “仅在需要时申请显存空间”。
for gpu in tf.config.experimental.list_physical_devices('GPU'):
    tf.config.experimental.set_memory_growth(gpu, True)

# 1、读取数据
'''由于Imagenet数据集是一个比较庞大的数据集，且网络的输入为224*224，所以，我们定义一个方法，
来读取数据集并将数据resize到224*224的大小'''

class Data_load():
    def __init__(self):
        fashion_mnist = tf.keras.datasets.fashion_mnist
        (self.train_images, self.train_labels), (self.test_images, self.test_labels)\
            = fashion_mnist.load_data()
        # 数据维度扩充成[n,h,w,c]的模式
        self.train_images = np.expand_dims(self.train_images.astype(np.float32) / 255.0, axis=-1)
        self.test_images = np.expand_dims(self.test_images.astype(np.float32)/255.0,axis=-1)
        # 标签
        self.train_labels = self.train_labels.astype(np.int32)
        self.test_labels = self.test_labels.astype(np.int32)
        # 训练和测试的数据个数
        self.num_train, self.num_test = self.train_images.shape[0], self.test_images.shape[0]
    def get_train_batch(self,batch_size):
        # 随机取batch_size个索引
        index = np.random.randint(0, np.shape(self.train_images)[0], batch_size)
        # resize
        resized_images = tf.image.resize_with_pad(self.train_images[index], 224, 224 )
        return resized_images.numpy(), self.train_labels[index]
    def get_test_batch(self,batch_size):
        index = np.random.randint(0, np.shape(self.test_images)[0], batch_size)
        # resize
        resized_images = tf.image.resize_with_pad(self.test_images[index], 224, 224 )
        return resized_images.numpy(), self.test_labels[index]


batch_size=64
data_load=Data_load()
x_train_batch,y_train_batch=data_load.get_train_batch(batch_size)
print("x_batch shape:",x_train_batch.shape,"y_batch shape:", y_train_batch.shape)

# 2、定义模型
def AlexNet():
    net=tf.keras.Sequential()
    net.add(tf.keras.layers.Conv2D(96,11,(4,4),"same",activation="relu"))
    net.add(tf.keras.layers.MaxPool2D(pool_size=3, strides=2))
    net.add(tf.keras.layers.Conv2D(filters=256, kernel_size=5, padding='same', activation='relu'))
    net.add(tf.keras.layers.MaxPool2D(pool_size=3, strides=2))
    net.add(tf.keras.layers.Conv2D(filters=384, kernel_size=3, padding='same', activation='relu'))
    net.add(tf.keras.layers.Conv2D(filters=384, kernel_size=3, padding='same', activation='relu'))
    net.add(tf.keras.layers.Conv2D(filters=256, kernel_size=3, padding='same', activation='relu'))
    net.add(tf.keras.layers.MaxPool2D(pool_size=3, strides=2))
    net.add(tf.keras.layers.Flatten())
    net.add(tf.keras.layers.Dense(512, activation='relu'))# 为了方便训练 神经元个数改小，原来是1024
    net.add(tf.keras.layers.Dropout(0.5))
    net.add(tf.keras.layers.Dense(256, activation='relu'))# 为了方便训练 神经元个数改小，原来是1024
    net.add(tf.keras.layers.Dropout(0.5))
    net.add(tf.keras.layers.Dense(10, activation='sigmoid'))

    return net
net=AlexNet()
# 查看网络结构
X = tf.random.uniform((1,224,224,1))
for layer in net.layers:
    X = layer(X)
    print(layer.name, 'output shape\t', X.shape)


def train(num_epoches,net):
    optimizer = tf.keras.optimizers.SGD(learning_rate=0.01, momentum=0.0, nesterov=False)
    net.compile(optimizer=optimizer,
                  loss='sparse_categorical_crossentropy',
                  metrics=['accuracy'])

    num_iter = data_load.num_train // batch_size
    for e in range(num_epoches):
        for n in range(num_iter):
            x_batch, y_batch = data_load.get_train_batch(batch_size)
            test_x_batch, test_y_batch = data_load.get_test_batch(batch_size)
            net.fit(x_batch, y_batch,validation_data=(test_x_batch, test_y_batch))
train(5,net)