学习笔记(01):轻松入门Tensorflow-LeNet模型构建

最新推荐文章于 2024-08-07 21:05:59 发布

齐玮

最新推荐文章于 2024-08-07 21:05:59 发布

阅读量374

点赞数 7

分类专栏：研发管理文章标签：编程语言移动开发研发管理人工智能深度学习

本文链接：https://blog.csdn.net/xinqiwei/article/details/104270616

版权

研发管理专栏收录该内容

1 篇文章 0 订阅

订阅专栏

立即学习:https://edu.csdn.net/course/play/26266/334243?utm_source=blogtoedu

给你们源代码

#MNIST数据集预处理
import tensorflow as tf

mnist = tf.keras.datasets.mnist
(x_train, y_train), (x_test, y_test) = mnist.load_data()
print(x_train.shape, y_train.shape)
print(x_test.shape, y_test.shape)

import matplotlib.pyplot as plt
image_index = 1234
plt.imshow(x_train[image_index],cmap='Greys')
plt.show()
print(y_train[image_index])

#格式化
import numpy as np
x_train = np.pad(x_train,((0,0),(2,2),(2,2)),'constant',constant_values=0)
print(x_train.shape)
x_train = x_train.astype('float32')
x_train /=255
x_train=x_train.reshape(x_train.shape[0],32,32,1)
print(x_train.shape)

# LeNet模型构建
# 定义模型
# 模型的构建:  tf.keras.Model 和tf.keras.layers
# 模型的损失函数：tf.keras.losses
# 模型的优化器：   tf.keras.optimizer
# 模型的评估：   tf.keras.metrics
class LetNet(tf.keras.Model):
    def __init__(self):
        super().__init__()
        #此处添加初始化代码（包含call方法中会用到的层）
        self.conv_layer_1 = tf.keras.layers.Conv2D(
            filters=6,
            kernel_size=(5,5),
            padding='valid',
            activation=tf.nn.relu)
        self.pool_layer_1=tf.keras.layers.MaxPooling2D(pool_size=(2,2),padding='same')

        self.conv_layer_2=tf.keras.layers.Conv2D(
            filters=16,
            kernel_size=(5,5),
            padding='valid',
            activation=tf.nn.relu)
        self.pool_layer_2=tf.keras.layers.MaxPooling2D(padding='same')
        self.flatten=tf.keras.layers.Flatten()
        self.fc_layer_1=tf.keras.layers.Dense(
            units=120,
            activation=tf.nn.relu)
        self.fc_layer_2=tf.keras.layers.Dense(
            units=84,
            activation=tf.nn.relu)
        self.output_layer=tf.keras.layers.Dense(
            units=10,
            activation=tf.nn.softmax)
    def call(self,inputs):    #[batch_size,28,28,1]
        x=self.conv_layer_1(inputs)
        x=self.pool_layer_1(x)
        x=self.conv_layer_2(x)
        x=self.pool_layer_2(x)
        x=self.flatten(x)
        x=self.fc_layer_1(x)
        x=self.fc_layer_2(x)
        output=self.output_layer(x)

        return output
    #还可以添加自定义的方法

    #模型实例化方法
    model=tf.keras.models.Sequential([
        tf.keras.layers.Conv2D(filters=6,kernel_size=(5,5),padding='valid',activation=tf.nn.relu,input_shape=(32,32,1)),
        tf.keras.layers.AveragePooling2D(pool_size=(2,2),strides=(2,2),padding='same'),
        tf.keras.layers.Conv2D(filters=16,kernel_size=(5,5),padding='valid',activation=tf.nn.relu),
        tf.keras.layers.AveragePooling2D(pool_size=(2,2),strides=(2,2),padding='same'),
        tf.keras.layers.Flatten(),
        tf.keras.layers.Dense(units=120,activation=tf.nn.relu),
        # tf.keras.layers.Conv2D(filters=120,kernel_size=(5,5),strides=(1,1),activation='tanh',padding='valid'),
        # tf.keras.layers.Flatten(),
        tf.keras.layers.Dense(units=84,activation=tf.nn.relu),
        tf.keras.layers.Dense(units=10,activation=tf.nn.softmax)
    ])
    model.summary()