就叫乾龙呀丶

用Tensorflow2搭建经典的卷积神经网络（LeNet,AlexNet,VGG16,ResNet50）

最近看了卷积神经网络的一些经典论文，用Tensorflow2进行了复现，其中包括LeNet,AlexNet,VGG16,ResNet50,并且在数据集上进行了测试，能够正常训练，结果表明搭建的没有问题。

下面进入正题：

import tensorflow as tf
from tensorflow import keras
from tensorflow.keras import layers
import numpy as np
import matplotlib.pyplot as plt
%matplotlib inline

加载mnist数据集用于测试LeNet

mnist_data = np.load("E:/deep_learning/jupyternotebook工作空间/0531/mnist/mnist.npz")
x_train = mnist_data['x_train'] / 255.0
y_train = mnist_data['y_train'] / 255.0
x_test = mnist_data['x_test']
y_test = mnist_data['y_test']

print(x_train.shape)
print(x_test.shape)
print(y_train.shape)
print(y_test.shape)

(60000, 28, 28)
(10000, 28, 28)
(60000,)
(10000,)

x_train0 = x_train.reshape((-1,28,28,1))
x_test0 = x_test.reshape((-1,28,28,1))

plt.imshow(x_test[0])
plt.title('The number is '+ str(y_test[0]))

Text(0.5, 1.0, 'The number is 7')

LeNet

LeNet模型网络结构示意图

def get_LeNet():
    model = keras.Sequential(name='LeNet')
    model.add(layers.Conv2D(filters=6,kernel_size=(5,5),strides=1,activation='sigmoid',
                            input_shape=(28,28,1),name='Conv1'))
    model.add(layers.MaxPool2D(pool_size=(2,2),strides=2,name='Pool1'))
    
    model.add(layers.Conv2D(16,(5,5),strides=1,activation='sigmoid',name='Conv2'))
    model.add(layers.MaxPool2D(pool_size=(2,2),strides=2,name='Pool2'))
    
    model.add(layers.Conv2D(120,(4,4),strides=1,activation='sigmoid',name='Conv3'))
    
    model.add(layers.Flatten(name='Flatten'))
    model.add(layers.Dense(64,activation='sigmoid',name='FC1'))
    model.add(layers.Dense(10,activation='softmax',name='FC2'))
    return model

LeNet = get_LeNet()
LeNet.summary()

Model: "LeNet"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
Conv1 (Conv2D)               (None, 24, 24, 6)         156       
_________________________________________________________________
Pool1 (MaxPooling2D)         (None, 12, 12, 6)         0         
_________________________________________________________________
Conv2 (Conv2D)               (None, 8, 8, 16)          2416      
_________________________________________________________________
Pool2 (MaxPooling2D)         (None, 4, 4, 16)          0         
_________________________________________________________________
Conv3 (Conv2D)               (None, 1, 1, 120)         30840     
_________________________________________________________________
Flatten (Flatten)            (None, 120)               0         
_________________________________________________________________
FC1 (Dense)                  (None, 64)                7744      
_________________________________________________________________
FC2 (Dense)                  (None, 10)                650       
=================================================================
Total params: 41,806
Trainable params: 41,806
Non-trainable params: 0
_________________________________________________________________

由上图可知，LeNet为5层网络，共41806个参数

#训练
LeNet.compile(optimizer=keras.optimizers.Adam(learning_rate=0.001),
                   loss='sparse_categorical_crossentropy',   #计算稀疏的分类交叉熵损失
                    metrics=['accuracy'])

LeNet.fit(x_train0,y_train,batch_size=64,epochs=1,validation_split=0.15)
LeNet.evaluate(x_test0,y_test,batch_size=64)

AlexNet

def alexNet_model(numclasses):
    alexNet = keras.Sequential(name='AlexNet')
    #第一层
    alexNet.add(layers.Conv2D(filters=96,kernel_size=(11,11),strides=(4,4),
                              activation='relu',input_shape=(227,227,3),name='Conv1'))
    alexNet.add(layers.BatchNormalization(name='BN1'))
    alexNet.add(layers.MaxPool2D(pool_size=(3,3),strides=(2,2),name='Pool1'))
    
    #第二层
    alexNet.add(layers.Conv2D(filters=256,kernel_size=(5,5),strides=(1,1),padding='same',
                              activation='relu',name='Conv2'))
    alexNet.add(layers.BatchNormalization(name='BN2'))
    alexNet.add(layers.MaxPool2D(pool_size=(3,3),strides=(2,2),name='Pool2'))
    
    #第三层
    alexNet.add(layers.Conv2D(filters=384,kernel_size=(3,3),strides=(1,1),padding='same',
                              activation='relu',name='Conv3'))
    
    #第四层
    alexNet.add(layers.Conv2D(filters=192,kernel_size=(3,3),strides=(1,1),padding='same',
                              activation='relu',name='Conv4'))
    
    #第五层
    alexNet.add(layers.Conv2D(filters=256,kernel_size=(3,3),strides=(1,1),padding='same',
                              activation='relu',name='Conv5'))
    alexNet.add(layers.BatchNormalization(name='BN3'))
    alexNet.add(layers.MaxPool2D(pool_size=(3,3),strides=(2,2),name='Pool3'))
    
    #全连接层
    alexNet.add(layers.Flatten())
    alexNet.add(layers.Dense(4096,activation='relu',name='fc1'))
    alexNet.add(layers.Dense(4096,activation='relu',name='fc2'))
    alexNet.add(layers.Dense(numclasses,activation='softmax',name='output'))
    
    return alexNet

AlexNet = alexNet_model(1000)
AlexNet.summary()

Model: "AlexNet"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
Conv1 (Conv2D)               (None, 55, 55, 96)        34944     
_________________________________________________________________
BN1 (BatchNormalization)     (None, 55, 55, 96)        384       
_________________________________________________________________
Pool1 (MaxPooling2D)         (None, 27, 27, 96)        0         
_________________________________________________________________
Conv2 (Conv2D)               (None, 27, 27, 256)       614656    
_________________________________________________________________
BN2 (BatchNormalization)     (None, 27, 27, 256)       1024      
_________________________________________________________________
Pool2 (MaxPooling2D)         (None, 13, 13, 256)       0         
_________________________________________________________________
Conv3 (Conv2D)               (None, 13, 13, 384)       885120    
_________________________________________________________________
Conv4 (Conv2D)               (None, 13, 13, 192)       663744    
_________________________________________________________________
Conv5 (Conv2D)               (None, 13, 13, 256)       442624    
_________________________________________________________________
BN3 (BatchNormalization)     (None, 13, 13, 256)       1024      
_________________________________________________________________
Pool3 (MaxPooling2D)         (None, 6, 6, 256)         0         
_________________________________________________________________
flatten (Flatten)            (None, 9216)              0         
_________________________________________________________________
fc1 (Dense)                  (None, 4096)              37752832  
_________________________________________________________________
fc2 (Dense)                  (None, 4096)              16781312  
_________________________________________________________________
output (Dense)               (None, 1000)              4097000   
=================================================================
Total params: 61,274,664
Trainable params: 61,273,448
Non-trainable params: 1,216
_________________________________________________________________

AlexNet共有61,273,448个训练参数

构造手势识别数据用于测试网络

import os
import random
from PIL import Image

#生成数据列表
data_path = 'E:\deep_learning\jupyternotebook工作空间\数据集\手势数字\Dataset'
charater_folders = os.listdir(data_path)

charater_folders

['0', '1', '2', '3', '4', '5', '6', '7', '8', '9']

img = Image.open(r'E:\deep_learning\jupyternotebook工作空间\数据集\手势数字\Dataset\0\IMG_1118.JPG')
plt.imshow(img)
plt.title("The lebel is 0")

Text(0.5, 1.0, 'The lebel is 0')

#生成数据列表
if(os.path.exists('./train_data.list')):
    os.remove('./train_data.list')
if(os.path.exists('./test_data.list')):
    os.remove('./test_data.list')
    
for charater_folder in charater_folders:
    with open('./train_data.txt','a') as f_train:
        with open('./test_data.txt','a') as f_test:
            imgname_list = os.listdir(os.path.join(data_path,charater_folder)) 
            count = 0
            for imgname in imgname_list:
                if imgname == '.DS_Store':
                    continue
                if count%10 == 0:#以10%的比例划分测试数据集
                    f_test.write(os.path.join(data_path,charater_folder,imgname) + '\t' + charater_folder + '\n')
                else:
                    f_train.write(os.path.join(data_path,charater_folder,imgname) + '\t' + charater_folder + '\n')
                    
                count += 1
print("列表已生成")

列表已生成

#加载数据集
def img_preprocess(img_path):
	'''数据预处理'''
    img = Image.open(img_path)
    img = img.resize((227,227),Image.ANTIALIAS)
    img = np.array(img).astype('float32')
    img = img / 255.0
    return img

def build_dataSet(datalist_path):
    img_list = []
    label_list = []
    with open(datalist_path,'r') as f:
        lines = f.readlines()
        random.shuffle(lines) #打乱数据
        for line in lines:
            imgpath,label = line.split('\t')
            label_list.append(int(label))
            img = img_preprocess(imgpath)
            img_list.append(img)
            
    return np.array(img_list),np.array(label_list)

img0 = img_preprocess(r'数据集\手势数字\Dataset\0\IMG_1118.JPG')  
img0.shape

(227, 227, 3)

plt.imshow(img0)

train_x,train_y = build_dataSet('train_data.txt')
test_x,test_y = build_dataSet('test_data.txt')

print(train_x.shape)
print(train_y.shape)
print(test_x.shape)
print(test_y.shape)

(1852, 227, 227, 3)
(1852,)
(210, 227, 227, 3)
(210,)

my_alexNet = alexNet_model(10)
my_alexNet.compile(optimizer=keras.optimizers.Adam(learning_rate=0.001),
                   loss='sparse_categorical_crossentropy',   #计算稀疏的分类交叉熵损失
                    metrics=['accuracy'])
my_alexNet.fit(train_x,train_y,batch_size=4,epochs=1)
my_alexNet.evaluate(test_x,test_y,batch_size=4)

#电脑太垃圾，跑一轮要半小时，不跑了。

VGGNet

def basic_net(model,filters,is_three):
    model.add(layers.Conv2D(filters=filters,kernel_size=3,strides=1,activation='relu',padding='same'))
    model.add(layers.Conv2D(filters=filters,kernel_size=3,strides=1,activation='relu',padding='same'))
    if is_three:
        model.add(layers.Conv2D(filters=filters,kernel_size=3,strides=1,activation='relu',padding='same'))
    model.add(layers.MaxPool2D(pool_size=(2,2),strides=(2,2)))
    return model

def get_vgg16(numclasses,input_shape):
    model = keras.Sequential(name='VGG16')
    model.add(layers.Conv2D(filters=64,kernel_size=3,strides=1,activation='relu',padding='same',input_shape=input_shape))
    model.add(layers.Conv2D(filters=64,kernel_size=3,strides=1,activation='relu',padding='same'))
    model.add(layers.MaxPool2D(pool_size=(2,2),strides=(2,2)))
    
    model = basic_net(model,128,is_three=False)
    
    model = basic_net(model,256,is_three=True)
    
    model = basic_net(model,512,is_three=True)
    
    model = basic_net(model,512,is_three=True)
    
    model.add(layers.Flatten())
    model.add(layers.Dense(4096,activation='relu'))
    model.add(layers.Dense(4096,activation='relu'))
    model.add(layers.Dense(numclasses,activation='softmax'))
    return model

VGG16 = get_vgg16(numclasses=1000,input_shape=(224,224,3))
keras.utils.plot_model(VGG16,to_file='VGG16.png')
VGG16.summary()

Failed to import pydot. You must install pydot and graphviz for `pydotprint` to work.
Model: "VGG16"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
conv2d_52 (Conv2D)           (None, 224, 224, 64)      1792      
_________________________________________________________________
conv2d_53 (Conv2D)           (None, 224, 224, 64)      36928     
_________________________________________________________________
max_pooling2d_20 (MaxPooling (None, 112, 112, 64)      0         
_________________________________________________________________
conv2d_54 (Conv2D)           (None, 112, 112, 128)     73856     
_________________________________________________________________
conv2d_55 (Conv2D)           (None, 112, 112, 128)     147584    
_________________________________________________________________
max_pooling2d_21 (MaxPooling (None, 56, 56, 128)       0         
_________________________________________________________________
conv2d_56 (Conv2D)           (None, 56, 56, 256)       295168    
_________________________________________________________________
conv2d_57 (Conv2D)           (None, 56, 56, 256)       590080    
_________________________________________________________________
conv2d_58 (Conv2D)           (None, 56, 56, 256)       590080    
_________________________________________________________________
max_pooling2d_22 (MaxPooling (None, 28, 28, 256)       0         
_________________________________________________________________
conv2d_59 (Conv2D)           (None, 28, 28, 512)       1180160   
_________________________________________________________________
conv2d_60 (Conv2D)           (None, 28, 28, 512)       2359808   
_________________________________________________________________
conv2d_61 (Conv2D)           (None, 28, 28, 512)       2359808   
_________________________________________________________________
max_pooling2d_23 (MaxPooling (None, 14, 14, 512)       0         
_________________________________________________________________
conv2d_62 (Conv2D)           (None, 14, 14, 512)       2359808   
_________________________________________________________________
conv2d_63 (Conv2D)           (None, 14, 14, 512)       2359808   
_________________________________________________________________
conv2d_64 (Conv2D)           (None, 14, 14, 512)       2359808   
_________________________________________________________________
max_pooling2d_24 (MaxPooling (None, 7, 7, 512)         0         
_________________________________________________________________
flatten_4 (Flatten)          (None, 25088)             0         
_________________________________________________________________
dense_12 (Dense)             (None, 4096)              102764544 
_________________________________________________________________
dense_13 (Dense)             (None, 4096)              16781312  
_________________________________________________________________
dense_14 (Dense)             (None, 1000)              4097000   
=================================================================
Total params: 138,357,544
Trainable params: 138,357,544
Non-trainable params: 0
_________________________________________________________________

VGG16共有138,357,544个训练参数

ResNet

搭建ResNet50

def res_block(x,filters,is_same,s=1):
    '''构造残差快'''
    x_shortcut = x
    f1,f2,f3 = filters
    x = layers.Conv2D(filters=f1,kernel_size=(1,1),strides=(s,s),padding='same')(x)
    x = layers.BatchNormalization()(x)
    x = layers.Activation('relu')(x)
    
    x = layers.Conv2D(filters=f2,kernel_size=(3,3),strides=(1,1),padding='same')(x)
    x = layers.BatchNormalization()(x)
    x = layers.Activation('relu')(x)
    
    x = layers.Conv2D(filters=f3,kernel_size=(1,1),strides=(1,1),padding='same')(x)
    x = layers.BatchNormalization()(x)
    if is_same:
        x_shortcut = layers.Conv2D(filters=f3,kernel_size=(1,1),strides=(1,1),padding='valid')(x_shortcut)
        x = layers.Add()([x,x_shortcut])
        x = layers.Activation('relu')(x)
    else:
        x_shortcut = layers.Conv2D(filters=f3,kernel_size=(1,1),strides=(2,2),padding='valid')(x_shortcut)
        x_shortcut = layers.BatchNormalization()(x_shortcut)
        x = layers.Add()([x,x_shortcut])
        x = layers.Activation('relu')(x)
    return x

def get_ResNet50(numclasses,input_shape):
    #浅层网络
    inputs = layers.Input(shape=input_shape)
    x = layers.ZeroPadding2D(padding=(3,3))(inputs)
    x = layers.Conv2D(filters=64,kernel_size=(7,7),strides=(2,2),activation='relu')(x)
    x = layers.ZeroPadding2D(padding=(1,1))(inputs)
    x = layers.MaxPool2D(pool_size=(3,3),strides=(2,2))(x)
    
    #残差块1 *3
    x = res_block(x,[64,64,256],is_same=True)
    x = res_block(x,[64,64,256],is_same=True)
    x = res_block(x,[64,64,256],is_same=True)
    
    #残差块2 *4
    x = res_block(x,[128,128,512],is_same=False,s=2)
    x = res_block(x,[128,128,512],is_same=True)
    x = res_block(x,[128,128,512],is_same=True)
    x = res_block(x,[128,128,512],is_same=True)
    
    #残差块3 *6
    x = res_block(x,[256,256,1024],is_same=False,s=2)
    x = res_block(x,[256,256,1024],is_same=True)
    x = res_block(x,[256,256,1024],is_same=True)
    x = res_block(x,[256,256,1024],is_same=True)
    x = res_block(x,[256,256,1024],is_same=True)
    x = res_block(x,[256,256,1024],is_same=True)
    
    #残差块4 *3
    x = res_block(x,[512,512,2048],is_same=False,s=2)
    x = res_block(x,[512,512,2048],is_same=True)
    x = res_block(x,[512,512,2048],is_same=True)
    
    #全局平均池化
    x = layers.GlobalAveragePooling2D()(x)
    x = layers.Flatten()(x)
    outputs = layers.Dense(numclasses,activation='softmax')(x)
    
    model = keras.Model(inputs=inputs,outputs=outputs,name='ResNet50')
    return model

ResNet50 = get_ResNet50(1000,(224,224,3))
ResNet50.summary()

Model: "ResNet50"
__________________________________________________________________________________________________
Layer (type)                    Output Shape         Param #     Connected to                     
==================================================================================================
input_5 (InputLayer)            [(None, 224, 224, 3) 0                                            
__________________________________________________________________________________________________
zero_padding2d_9 (ZeroPadding2D (None, 226, 226, 3)  0           input_5[0][0]                    
__________________________________________________________________________________________________
max_pooling2d_29 (MaxPooling2D) (None, 112, 112, 3)  0           zero_padding2d_9[0][0]           
__________________________________________________________________________________________________
conv2d_182 (Conv2D)             (None, 112, 112, 64) 256         max_pooling2d_29[0][0]           
__________________________________________________________________________________________________
batch_normalization_90 (BatchNo (None, 112, 112, 64) 256         conv2d_182[0][0]                 
__________________________________________________________________________________________________
activation_81 (Activation)      (None, 112, 112, 64) 0           batch_normalization_90[0][0]     
__________________________________________________________________________________________________
conv2d_183 (Conv2D)             (None, 112, 112, 64) 36928       activation_81[0][0]              
__________________________________________________________________________________________________
batch_normalization_91 (BatchNo (None, 112, 112, 64) 256         conv2d_183[0][0]                 
__________________________________________________________________________________________________
activation_82 (Activation)      (None, 112, 112, 64) 0           batch_normalization_91[0][0]     
__________________________________________________________________________________________________
conv2d_184 (Conv2D)             (None, 112, 112, 256 16640       activation_82[0][0]              
__________________________________________________________________________________________________
batch_normalization_92 (BatchNo (None, 112, 112, 256 1024        conv2d_184[0][0]                 
__________________________________________________________________________________________________
conv2d_185 (Conv2D)             (None, 112, 112, 256 1024        max_pooling2d_29[0][0]           
__________________________________________________________________________________________________
add_25 (Add)                    (None, 112, 112, 256 0           batch_normalization_92[0][0]     
                                                                 conv2d_185[0][0]                 
__________________________________________________________________________________________________
activation_83 (Activation)      (None, 112, 112, 256 0           add_25[0][0]                     
__________________________________________________________________________________________________
conv2d_186 (Conv2D)             (None, 112, 112, 64) 16448       activation_83[0][0]              
__________________________________________________________________________________________________
batch_normalization_93 (BatchNo (None, 112, 112, 64) 256         conv2d_186[0][0]                 
__________________________________________________________________________________________________
activation_84 (Activation)      (None, 112, 112, 64) 0           batch_normalization_93[0][0]     
__________________________________________________________________________________________________
conv2d_187 (Conv2D)             (None, 112, 112, 64) 36928       activation_84[0][0]              
__________________________________________________________________________________________________
batch_normalization_94 (BatchNo (None, 112, 112, 64) 256         conv2d_187[0][0]                 
__________________________________________________________________________________________________
activation_85 (Activation)      (None, 112, 112, 64) 0           batch_normalization_94[0][0]     
__________________________________________________________________________________________________
conv2d_188 (Conv2D)             (None, 112, 112, 256 16640       activation_85[0][0]              
__________________________________________________________________________________________________
batch_normalization_95 (BatchNo (None, 112, 112, 256 1024        conv2d_188[0][0]                 
__________________________________________________________________________________________________
conv2d_189 (Conv2D)             (None, 112, 112, 256 65792       activation_83[0][0]              
__________________________________________________________________________________________________
add_26 (Add)                    (None, 112, 112, 256 0           batch_normalization_95[0][0]     
                                                                 conv2d_189[0][0]                 
__________________________________________________________________________________________________
activation_86 (Activation)      (None, 112, 112, 256 0           add_26[0][0]                     
__________________________________________________________________________________________________
conv2d_190 (Conv2D)             (None, 112, 112, 64) 16448       activation_86[0][0]              
__________________________________________________________________________________________________
batch_normalization_96 (BatchNo (None, 112, 112, 64) 256         conv2d_190[0][0]                 
__________________________________________________________________________________________________
activation_87 (Activation)      (None, 112, 112, 64) 0           batch_normalization_96[0][0]     
__________________________________________________________________________________________________
conv2d_191 (Conv2D)             (None, 112, 112, 64) 36928       activation_87[0][0]              
__________________________________________________________________________________________________
batch_normalization_97 (BatchNo (None, 112, 112, 64) 256         conv2d_191[0][0]                 
__________________________________________________________________________________________________
activation_88 (Activation)      (None, 112, 112, 64) 0           batch_normalization_97[0][0]     
__________________________________________________________________________________________________
conv2d_192 (Conv2D)             (None, 112, 112, 256 16640       activation_88[0][0]              
__________________________________________________________________________________________________
batch_normalization_98 (BatchNo (None, 112, 112, 256 1024        conv2d_192[0][0]                 
__________________________________________________________________________________________________
conv2d_193 (Conv2D)             (None, 112, 112, 256 65792       activation_86[0][0]              
__________________________________________________________________________________________________
add_27 (Add)                    (None, 112, 112, 256 0           batch_normalization_98[0][0]     
                                                                 conv2d_193[0][0]                 
__________________________________________________________________________________________________
activation_89 (Activation)      (None, 112, 112, 256 0           add_27[0][0]                     
__________________________________________________________________________________________________
conv2d_194 (Conv2D)             (None, 56, 56, 128)  32896       activation_89[0][0]              
__________________________________________________________________________________________________
batch_normalization_99 (BatchNo (None, 56, 56, 128)  512         conv2d_194[0][0]                 
__________________________________________________________________________________________________
activation_90 (Activation)      (None, 56, 56, 128)  0           batch_normalization_99[0][0]     
__________________________________________________________________________________________________
conv2d_195 (Conv2D)             (None, 56, 56, 128)  147584      activation_90[0][0]              
__________________________________________________________________________________________________
batch_normalization_100 (BatchN (None, 56, 56, 128)  512         conv2d_195[0][0]                 
__________________________________________________________________________________________________
activation_91 (Activation)      (None, 56, 56, 128)  0           batch_normalization_100[0][0]    
__________________________________________________________________________________________________
conv2d_196 (Conv2D)             (None, 56, 56, 512)  66048       activation_91[0][0]              
__________________________________________________________________________________________________
conv2d_197 (Conv2D)             (None, 56, 56, 512)  131584      activation_89[0][0]              
__________________________________________________________________________________________________
batch_normalization_101 (BatchN (None, 56, 56, 512)  2048        conv2d_196[0][0]                 
__________________________________________________________________________________________________
batch_normalization_102 (BatchN (None, 56, 56, 512)  2048        conv2d_197[0][0]                 
__________________________________________________________________________________________________
add_28 (Add)                    (None, 56, 56, 512)  0           batch_normalization_101[0][0]    
                                                                 batch_normalization_102[0][0]    
__________________________________________________________________________________________________
activation_92 (Activation)      (None, 56, 56, 512)  0           add_28[0][0]                     
__________________________________________________________________________________________________
conv2d_198 (Conv2D)             (None, 56, 56, 128)  65664       activation_92[0][0]              
__________________________________________________________________________________________________
batch_normalization_103 (BatchN (None, 56, 56, 128)  512         conv2d_198[0][0]                 
__________________________________________________________________________________________________
activation_93 (Activation)      (None, 56, 56, 128)  0           batch_normalization_103[0][0]    
__________________________________________________________________________________________________
conv2d_199 (Conv2D)             (None, 56, 56, 128)  147584      activation_93[0][0]              
__________________________________________________________________________________________________
batch_normalization_104 (BatchN (None, 56, 56, 128)  512         conv2d_199[0][0]                 
__________________________________________________________________________________________________
activation_94 (Activation)      (None, 56, 56, 128)  0           batch_normalization_104[0][0]    
__________________________________________________________________________________________________
conv2d_200 (Conv2D)             (None, 56, 56, 512)  66048       activation_94[0][0]              
__________________________________________________________________________________________________
batch_normalization_105 (BatchN (None, 56, 56, 512)  2048        conv2d_200[0][0]                 
__________________________________________________________________________________________________
conv2d_201 (Conv2D)             (None, 56, 56, 512)  262656      activation_92[0][0]              
__________________________________________________________________________________________________
add_29 (Add)                    (None, 56, 56, 512)  0           batch_normalization_105[0][0]    
                                                                 conv2d_201[0][0]                 
__________________________________________________________________________________________________
activation_95 (Activation)      (None, 56, 56, 512)  0           add_29[0][0]                     
__________________________________________________________________________________________________
conv2d_202 (Conv2D)             (None, 56, 56, 128)  65664       activation_95[0][0]              
__________________________________________________________________________________________________
batch_normalization_106 (BatchN (None, 56, 56, 128)  512         conv2d_202[0][0]                 
__________________________________________________________________________________________________
activation_96 (Activation)      (None, 56, 56, 128)  0           batch_normalization_106[0][0]    
__________________________________________________________________________________________________
conv2d_203 (Conv2D)             (None, 56, 56, 128)  147584      activation_96[0][0]              
__________________________________________________________________________________________________
batch_normalization_107 (BatchN (None, 56, 56, 128)  512         conv2d_203[0][0]                 
__________________________________________________________________________________________________
activation_97 (Activation)      (None, 56, 56, 128)  0           batch_normalization_107[0][0]    
__________________________________________________________________________________________________
conv2d_204 (Conv2D)             (None, 56, 56, 512)  66048       activation_97[0][0]              
__________________________________________________________________________________________________
batch_normalization_108 (BatchN (None, 56, 56, 512)  2048        conv2d_204[0][0]                 
__________________________________________________________________________________________________
conv2d_205 (Conv2D)             (None, 56, 56, 512)  262656      activation_95[0][0]              
__________________________________________________________________________________________________
add_30 (Add)                    (None, 56, 56, 512)  0           batch_normalization_108[0][0]    
                                                                 conv2d_205[0][0]                 
__________________________________________________________________________________________________
activation_98 (Activation)      (None, 56, 56, 512)  0           add_30[0][0]                     
__________________________________________________________________________________________________
conv2d_206 (Conv2D)             (None, 56, 56, 128)  65664       activation_98[0][0]              
__________________________________________________________________________________________________
batch_normalization_109 (BatchN (None, 56, 56, 128)  512         conv2d_206[0][0]                 
__________________________________________________________________________________________________
activation_99 (Activation)      (None, 56, 56, 128)  0           batch_normalization_109[0][0]    
__________________________________________________________________________________________________
conv2d_207 (Conv2D)             (None, 56, 56, 128)  147584      activation_99[0][0]              
__________________________________________________________________________________________________
batch_normalization_110 (BatchN (None, 56, 56, 128)  512         conv2d_207[0][0]                 
__________________________________________________________________________________________________
activation_100 (Activation)     (None, 56, 56, 128)  0           batch_normalization_110[0][0]    
__________________________________________________________________________________________________
conv2d_208 (Conv2D)             (None, 56, 56, 512)  66048       activation_100[0][0]             
__________________________________________________________________________________________________
batch_normalization_111 (BatchN (None, 56, 56, 512)  2048        conv2d_208[0][0]                 
__________________________________________________________________________________________________
conv2d_209 (Conv2D)             (None, 56, 56, 512)  262656      activation_98[0][0]              
__________________________________________________________________________________________________
add_31 (Add)                    (None, 56, 56, 512)  0           batch_normalization_111[0][0]    
                                                                 conv2d_209[0][0]                 
__________________________________________________________________________________________________
activation_101 (Activation)     (None, 56, 56, 512)  0           add_31[0][0]                     
__________________________________________________________________________________________________
conv2d_210 (Conv2D)             (None, 28, 28, 256)  131328      activation_101[0][0]             
__________________________________________________________________________________________________
batch_normalization_112 (BatchN (None, 28, 28, 256)  1024        conv2d_210[0][0]                 
__________________________________________________________________________________________________
activation_102 (Activation)     (None, 28, 28, 256)  0           batch_normalization_112[0][0]    
__________________________________________________________________________________________________
conv2d_211 (Conv2D)             (None, 28, 28, 256)  590080      activation_102[0][0]             
__________________________________________________________________________________________________
batch_normalization_113 (BatchN (None, 28, 28, 256)  1024        conv2d_211[0][0]                 
__________________________________________________________________________________________________
activation_103 (Activation)     (None, 28, 28, 256)  0           batch_normalization_113[0][0]    
__________________________________________________________________________________________________
conv2d_212 (Conv2D)             (None, 28, 28, 1024) 263168      activation_103[0][0]             
__________________________________________________________________________________________________
conv2d_213 (Conv2D)             (None, 28, 28, 1024) 525312      activation_101[0][0]             
__________________________________________________________________________________________________
batch_normalization_114 (BatchN (None, 28, 28, 1024) 4096        conv2d_212[0][0]                 
__________________________________________________________________________________________________
batch_normalization_115 (BatchN (None, 28, 28, 1024) 4096        conv2d_213[0][0]                 
__________________________________________________________________________________________________
add_32 (Add)                    (None, 28, 28, 1024) 0           batch_normalization_114[0][0]    
                                                                 batch_normalization_115[0][0]    
__________________________________________________________________________________________________
activation_104 (Activation)     (None, 28, 28, 1024) 0           add_32[0][0]                     
__________________________________________________________________________________________________
conv2d_214 (Conv2D)             (None, 28, 28, 256)  262400      activation_104[0][0]             
__________________________________________________________________________________________________
batch_normalization_116 (BatchN (None, 28, 28, 256)  1024        conv2d_214[0][0]                 
__________________________________________________________________________________________________
activation_105 (Activation)     (None, 28, 28, 256)  0           batch_normalization_116[0][0]    
__________________________________________________________________________________________________
conv2d_215 (Conv2D)             (None, 28, 28, 256)  590080      activation_105[0][0]             
__________________________________________________________________________________________________
batch_normalization_117 (BatchN (None, 28, 28, 256)  1024        conv2d_215[0][0]                 
__________________________________________________________________________________________________
activation_106 (Activation)     (None, 28, 28, 256)  0           batch_normalization_117[0][0]    
__________________________________________________________________________________________________
conv2d_216 (Conv2D)             (None, 28, 28, 1024) 263168      activation_106[0][0]             
__________________________________________________________________________________________________
batch_normalization_118 (BatchN (None, 28, 28, 1024) 4096        conv2d_216[0][0]                 
__________________________________________________________________________________________________
conv2d_217 (Conv2D)             (None, 28, 28, 1024) 1049600     activation_104[0][0]             
__________________________________________________________________________________________________
add_33 (Add)                    (None, 28, 28, 1024) 0           batch_normalization_118[0][0]    
                                                                 conv2d_217[0][0]                 
__________________________________________________________________________________________________
activation_107 (Activation)     (None, 28, 28, 1024) 0           add_33[0][0]                     
__________________________________________________________________________________________________
conv2d_218 (Conv2D)             (None, 28, 28, 256)  262400      activation_107[0][0]             
__________________________________________________________________________________________________
batch_normalization_119 (BatchN (None, 28, 28, 256)  1024        conv2d_218[0][0]                 
__________________________________________________________________________________________________
activation_108 (Activation)     (None, 28, 28, 256)  0           batch_normalization_119[0][0]    
__________________________________________________________________________________________________
conv2d_219 (Conv2D)             (None, 28, 28, 256)  590080      activation_108[0][0]             
__________________________________________________________________________________________________
batch_normalization_120 (BatchN (None, 28, 28, 256)  1024        conv2d_219[0][0]                 
__________________________________________________________________________________________________
activation_109 (Activation)     (None, 28, 28, 256)  0           batch_normalization_120[0][0]    
__________________________________________________________________________________________________
conv2d_220 (Conv2D)             (None, 28, 28, 1024) 263168      activation_109[0][0]             
__________________________________________________________________________________________________
batch_normalization_121 (BatchN (None, 28, 28, 1024) 4096        conv2d_220[0][0]                 
__________________________________________________________________________________________________
conv2d_221 (Conv2D)             (None, 28, 28, 1024) 1049600     activation_107[0][0]             
__________________________________________________________________________________________________
add_34 (Add)                    (None, 28, 28, 1024) 0           batch_normalization_121[0][0]    
                                                                 conv2d_221[0][0]                 
__________________________________________________________________________________________________
activation_110 (Activation)     (None, 28, 28, 1024) 0           add_34[0][0]                     
__________________________________________________________________________________________________
conv2d_222 (Conv2D)             (None, 28, 28, 256)  262400      activation_110[0][0]             
__________________________________________________________________________________________________
batch_normalization_122 (BatchN (None, 28, 28, 256)  1024        conv2d_222[0][0]                 
__________________________________________________________________________________________________
activation_111 (Activation)     (None, 28, 28, 256)  0           batch_normalization_122[0][0]    
__________________________________________________________________________________________________
conv2d_223 (Conv2D)             (None, 28, 28, 256)  590080      activation_111[0][0]             
__________________________________________________________________________________________________
batch_normalization_123 (BatchN (None, 28, 28, 256)  1024        conv2d_223[0][0]                 
__________________________________________________________________________________________________
activation_112 (Activation)     (None, 28, 28, 256)  0           batch_normalization_123[0][0]    
__________________________________________________________________________________________________
conv2d_224 (Conv2D)             (None, 28, 28, 1024) 263168      activation_112[0][0]             
__________________________________________________________________________________________________
batch_normalization_124 (BatchN (None, 28, 28, 1024) 4096        conv2d_224[0][0]                 
__________________________________________________________________________________________________
conv2d_225 (Conv2D)             (None, 28, 28, 1024) 1049600     activation_110[0][0]             
__________________________________________________________________________________________________
add_35 (Add)                    (None, 28, 28, 1024) 0           batch_normalization_124[0][0]    
                                                                 conv2d_225[0][0]                 
__________________________________________________________________________________________________
activation_113 (Activation)     (None, 28, 28, 1024) 0           add_35[0][0]                     
__________________________________________________________________________________________________
conv2d_226 (Conv2D)             (None, 28, 28, 256)  262400      activation_113[0][0]             
__________________________________________________________________________________________________
batch_normalization_125 (BatchN (None, 28, 28, 256)  1024        conv2d_226[0][0]                 
__________________________________________________________________________________________________
activation_114 (Activation)     (None, 28, 28, 256)  0           batch_normalization_125[0][0]    
__________________________________________________________________________________________________
conv2d_227 (Conv2D)             (None, 28, 28, 256)  590080      activation_114[0][0]             
__________________________________________________________________________________________________
batch_normalization_126 (BatchN (None, 28, 28, 256)  1024        conv2d_227[0][0]                 
__________________________________________________________________________________________________
activation_115 (Activation)     (None, 28, 28, 256)  0           batch_normalization_126[0][0]    
__________________________________________________________________________________________________
conv2d_228 (Conv2D)             (None, 28, 28, 1024) 263168      activation_115[0][0]             
__________________________________________________________________________________________________
batch_normalization_127 (BatchN (None, 28, 28, 1024) 4096        conv2d_228[0][0]                 
__________________________________________________________________________________________________
conv2d_229 (Conv2D)             (None, 28, 28, 1024) 1049600     activation_113[0][0]             
__________________________________________________________________________________________________
add_36 (Add)                    (None, 28, 28, 1024) 0           batch_normalization_127[0][0]    
                                                                 conv2d_229[0][0]                 
__________________________________________________________________________________________________
activation_116 (Activation)     (None, 28, 28, 1024) 0           add_36[0][0]                     
__________________________________________________________________________________________________
conv2d_230 (Conv2D)             (None, 28, 28, 256)  262400      activation_116[0][0]             
__________________________________________________________________________________________________
batch_normalization_128 (BatchN (None, 28, 28, 256)  1024        conv2d_230[0][0]                 
__________________________________________________________________________________________________
activation_117 (Activation)     (None, 28, 28, 256)  0           batch_normalization_128[0][0]    
__________________________________________________________________________________________________
conv2d_231 (Conv2D)             (None, 28, 28, 256)  590080      activation_117[0][0]             
__________________________________________________________________________________________________
batch_normalization_129 (BatchN (None, 28, 28, 256)  1024        conv2d_231[0][0]                 
__________________________________________________________________________________________________
activation_118 (Activation)     (None, 28, 28, 256)  0           batch_normalization_129[0][0]    
__________________________________________________________________________________________________
conv2d_232 (Conv2D)             (None, 28, 28, 1024) 263168      activation_118[0][0]             
__________________________________________________________________________________________________
batch_normalization_130 (BatchN (None, 28, 28, 1024) 4096        conv2d_232[0][0]                 
__________________________________________________________________________________________________
conv2d_233 (Conv2D)             (None, 28, 28, 1024) 1049600     activation_116[0][0]             
__________________________________________________________________________________________________
add_37 (Add)                    (None, 28, 28, 1024) 0           batch_normalization_130[0][0]    
                                                                 conv2d_233[0][0]                 
__________________________________________________________________________________________________
activation_119 (Activation)     (None, 28, 28, 1024) 0           add_37[0][0]                     
__________________________________________________________________________________________________
conv2d_234 (Conv2D)             (None, 14, 14, 512)  524800      activation_119[0][0]             
__________________________________________________________________________________________________
batch_normalization_131 (BatchN (None, 14, 14, 512)  2048        conv2d_234[0][0]                 
__________________________________________________________________________________________________
activation_120 (Activation)     (None, 14, 14, 512)  0           batch_normalization_131[0][0]    
__________________________________________________________________________________________________
conv2d_235 (Conv2D)             (None, 14, 14, 512)  2359808     activation_120[0][0]             
__________________________________________________________________________________________________
batch_normalization_132 (BatchN (None, 14, 14, 512)  2048        conv2d_235[0][0]                 
__________________________________________________________________________________________________
activation_121 (Activation)     (None, 14, 14, 512)  0           batch_normalization_132[0][0]    
__________________________________________________________________________________________________
conv2d_236 (Conv2D)             (None, 14, 14, 2048) 1050624     activation_121[0][0]             
__________________________________________________________________________________________________
conv2d_237 (Conv2D)             (None, 14, 14, 2048) 2099200     activation_119[0][0]             
__________________________________________________________________________________________________
batch_normalization_133 (BatchN (None, 14, 14, 2048) 8192        conv2d_236[0][0]                 
__________________________________________________________________________________________________
batch_normalization_134 (BatchN (None, 14, 14, 2048) 8192        conv2d_237[0][0]                 
__________________________________________________________________________________________________
add_38 (Add)                    (None, 14, 14, 2048) 0           batch_normalization_133[0][0]    
                                                                 batch_normalization_134[0][0]    
__________________________________________________________________________________________________
activation_122 (Activation)     (None, 14, 14, 2048) 0           add_38[0][0]                     
__________________________________________________________________________________________________
conv2d_238 (Conv2D)             (None, 14, 14, 512)  1049088     activation_122[0][0]             
__________________________________________________________________________________________________
batch_normalization_135 (BatchN (None, 14, 14, 512)  2048        conv2d_238[0][0]                 
__________________________________________________________________________________________________
activation_123 (Activation)     (None, 14, 14, 512)  0           batch_normalization_135[0][0]    
__________________________________________________________________________________________________
conv2d_239 (Conv2D)             (None, 14, 14, 512)  2359808     activation_123[0][0]             
__________________________________________________________________________________________________
batch_normalization_136 (BatchN (None, 14, 14, 512)  2048        conv2d_239[0][0]                 
__________________________________________________________________________________________________
activation_124 (Activation)     (None, 14, 14, 512)  0           batch_normalization_136[0][0]    
__________________________________________________________________________________________________
conv2d_240 (Conv2D)             (None, 14, 14, 2048) 1050624     activation_124[0][0]             
__________________________________________________________________________________________________
batch_normalization_137 (BatchN (None, 14, 14, 2048) 8192        conv2d_240[0][0]                 
__________________________________________________________________________________________________
conv2d_241 (Conv2D)             (None, 14, 14, 2048) 4196352     activation_122[0][0]             
__________________________________________________________________________________________________
add_39 (Add)                    (None, 14, 14, 2048) 0           batch_normalization_137[0][0]    
                                                                 conv2d_241[0][0]                 
__________________________________________________________________________________________________
activation_125 (Activation)     (None, 14, 14, 2048) 0           add_39[0][0]                     
__________________________________________________________________________________________________
conv2d_242 (Conv2D)             (None, 14, 14, 512)  1049088     activation_125[0][0]             
__________________________________________________________________________________________________
batch_normalization_138 (BatchN (None, 14, 14, 512)  2048        conv2d_242[0][0]                 
__________________________________________________________________________________________________
activation_126 (Activation)     (None, 14, 14, 512)  0           batch_normalization_138[0][0]    
__________________________________________________________________________________________________
conv2d_243 (Conv2D)             (None, 14, 14, 512)  2359808     activation_126[0][0]             
__________________________________________________________________________________________________
batch_normalization_139 (BatchN (None, 14, 14, 512)  2048        conv2d_243[0][0]                 
__________________________________________________________________________________________________
activation_127 (Activation)     (None, 14, 14, 512)  0           batch_normalization_139[0][0]    
__________________________________________________________________________________________________
conv2d_244 (Conv2D)             (None, 14, 14, 2048) 1050624     activation_127[0][0]             
__________________________________________________________________________________________________
batch_normalization_140 (BatchN (None, 14, 14, 2048) 8192        conv2d_244[0][0]                 
__________________________________________________________________________________________________
conv2d_245 (Conv2D)             (None, 14, 14, 2048) 4196352     activation_125[0][0]             
__________________________________________________________________________________________________
add_40 (Add)                    (None, 14, 14, 2048) 0           batch_normalization_140[0][0]    
                                                                 conv2d_245[0][0]                 
__________________________________________________________________________________________________
activation_128 (Activation)     (None, 14, 14, 2048) 0           add_40[0][0]                     
__________________________________________________________________________________________________
global_average_pooling2d_1 (Glo (None, 2048)         0           activation_128[0][0]             
__________________________________________________________________________________________________
flatten_6 (Flatten)             (None, 2048)         0           global_average_pooling2d_1[0][0] 
__________________________________________________________________________________________________
dense_16 (Dense)                (None, 1000)         2049000     flatten_6[0][0]                  
==================================================================================================
Total params: 40,166,696
Trainable params: 40,114,216
Non-trainable params: 52,480
__________________________________________________________________________________________________

ResNet共有 40,114,216个训练参数

#训练
my_resnet50 = get_ResNet50(10,(227,227,3))
my_resnet50.compile(optimizer=keras.optimizers.Adam(learning_rate=0.001),
                   loss='sparse_categorical_crossentropy',   #计算稀疏的分类交叉熵损失
                    metrics=['accuracy'])

my_resnet50.fit(test_x,test_y,batch_size=1,epochs=1)

通过对比AlexNet,VGG16,ResNet50,可以发现，ResNet50层数最多，而且训练参数最少。

你可能感兴趣的:(经典网络代码复现,神经网络,tensorflow,图像识别,深度学习)

机器学习与深度学习间关系与区别 ℒℴѵℯ心·动ꦿ໊ོ꫞ 人工智能学习深度学习 python
一、机器学习概述定义机器学习（MachineLearning,ML）是一种通过数据驱动的方法，利用统计学和计算算法来训练模型，使计算机能够从数据中学习并自动进行预测或决策。机器学习通过分析大量数据样本，识别其中的模式和规律，从而对新的数据进行判断。其核心在于通过训练过程，让模型不断优化和提升其预测准确性。主要类型1.监督学习（SupervisedLearning）监督学习是指在训练数据集中包含输入
【iOS】MVC设计模式 Magnetic_h ios mvc 设计模式 objective-c 学习 ui
MVC前言如何设计一个程序的结构，这是一门专门的学问，叫做"架构模式"（architecturalpattern），属于编程的方法论。MVC模式就是架构模式的一种。它是Apple官方推荐的App开发架构，也是一般开发者最先遇到、最经典的架构。MVC各层controller层Controller/ViewController/VC（控制器）负责协调Model和View，处理大部分逻辑它将数据从Mod
element实现动态路由+面包屑软件技术NINI vue案例 vue.js 前端
el-breadcrumb是ElementUI组件库中的一个面包屑导航组件，它用于显示当前页面的路径，帮助用户快速理解和导航到应用的各个部分。在Vue.js项目中，如果你已经安装了ElementUI，就可以很方便地使用el-breadcrumb组件。以下是一个基本的使用示例：安装ElementUI（如果你还没有安装的话）:你可以通过npm或yarn来安装ElementUI。bash复制代码npmi
C语言如何定义宏函数？小九格物 c语言
在C语言中，宏函数是通过预处理器定义的，它在编译之前替换代码中的宏调用。宏函数可以模拟函数的行为，但它们不是真正的函数，因为它们在编译时不会进行类型检查，也不会分配存储空间。宏函数的定义通常使用#define指令，后面跟着宏的名称和参数列表，以及宏展开后的代码。宏函数的定义方式：1.基本宏函数：这是最简单的宏函数形式，它直接定义一个表达式。#defineSQUARE(x)((x)*(x))2.带参
html 中如何使用 uniapp 的部分方法某公司摸鱼前端 html uni-app 前端
示例代码：Documentconsole.log(window);效果展示：好了，现在就可以uni.使用相关的方法了
回溯 Leetcode 332 重新安排行程 mmaerd Leetcode刷题学习记录 leetcode 算法职场和发展
重新安排行程Leetcode332学习记录自代码随想录给你一份航线列表tickets，其中tickets[i]=[fromi,toi]表示飞机出发和降落的机场地点。请你对该行程进行重新规划排序。所有这些机票都属于一个从JFK（肯尼迪国际机场）出发的先生，所以该行程必须从JFK开始。如果存在多种有效的行程，请你按字典排序返回最小的行程组合。例如，行程[“JFK”,“LGA”]与[“JFK”,“LGB
Pyecharts数据可视化大屏：打造沉浸式数据分析体验我的运维人生信息可视化数据分析数据挖掘运维开发技术共享
Pyecharts数据可视化大屏：打造沉浸式数据分析体验在当今这个数据驱动的时代，如何将海量数据以直观、生动的方式展现出来，成为了数据分析师和企业决策者关注的焦点。Pyecharts，作为一款基于Python的开源数据可视化库，凭借其丰富的图表类型、灵活的配置选项以及高度的定制化能力，成为了构建数据可视化大屏的理想选择。本文将深入探讨如何利用Pyecharts打造数据可视化大屏，并通过实际代码案例
将cmd中命令输出保存为txt文本文件落难Coder Windows cmd window
最近深度学习本地的训练中我们常常要在命令行中运行自己的代码，无可厚非，我们有必要保存我们的炼丹结果，但是复制命令行输出到txt是非常麻烦的，其实Windows下的命令行为我们提供了相应的操作。其基本的调用格式就是：运行指令>输出到的文件名称或者具体保存路径测试下，我打开cmd并且ping一下百度：pingwww.baidu.com>./data.txt看下相同目录下data.txt的输出：如果你再
基于社交网络算法优化的二维最大熵图像分割智能算法研学社（Jack旭）智能优化算法应用图像分割算法 php 开发语言
智能优化算法应用：基于社交网络优化的二维最大熵图像阈值分割-附代码文章目录智能优化算法应用：基于社交网络优化的二维最大熵图像阈值分割-附代码1.前言2.二维最大熵阈值分割原理3.基于社交网络优化的多阈值分割4.算法结果：5.参考文献：6.Matlab代码摘要：本文介绍基于最大熵的图像分割，并且应用社交网络算法进行阈值寻优。1.前言阅读此文章前，请阅读《图像分割：直方图区域划分及信息统计介绍》htt
关于提高复杂业务逻辑代码可读性的思考编程经验分享开发经验 java 数据库开发语言
目录前言需求场景常规写法拆分方法领域对象总结前言实际工作中大部分时间都是在写业务逻辑，一般都是三层架构，表示层（Controller）接收客户端请求，并对入参做检验，业务逻辑层（Service）负责处理业务逻辑，一般开发都是在这一层中写具体的业务逻辑。数据访问层（Dao）是直接和数据库交互的，用于查数据给业务逻辑层，或者是将业务逻辑层处理后的数据写入数据库。简单的增删改查接口不用多说，基本上写好一
DIV+CSS+JavaScript技术制作网页（旅游主题网页设计与制作）云南大理 STU学生网页设计网页设计期末网页作业 html静态网页 html5期末大作业网页设计 web大作业
️精彩专栏推荐作者主页:【进入主页—获取更多源码】web前端期末大作业：【HTML5网页期末作业(1000套)】程序员有趣的告白方式：【HTML七夕情人节表白网页制作(110套)】文章目录二、网站介绍三、网站效果▶️1.视频演示2.图片演示四、网站代码HTML结构代码CSS样式代码五、更多源码二、网站介绍网站布局方面：计划采用目前主流的、能兼容各大主流浏览器、显示效果稳定的浮动网页布局结构。网站程
2023-04-17|篮球女孩长一木
1小学抑或初中阶段，在课外书了解到她的故事。“篮球女孩”。当时佩服她的顽强，也对生命多了一丝敬畏。今天刚好在公众号看到，长大后的“篮球女孩”。佩服之余又满是心疼。网络侵删祝那素未蒙面的女孩，未来一切顺遂。
使用Faiss进行高效相似度搜索 llzwxh888 faiss python
在现代AI应用中，快速和高效的相似度搜索是至关重要的。Faiss（FacebookAISimilaritySearch）是一个专门用于快速相似度搜索和聚类的库，特别适用于高维向量。本文将介绍如何使用Faiss来进行相似度搜索，并结合Python代码演示其基本用法。什么是Faiss？Faiss是一个由FacebookAIResearch团队开发的开源库，主要用于高维向量的相似性搜索和聚类。Faiss
数组去重好奇的猫猫猫
整理自js中基础数据结构数组去重问题思考？如何去除数组中重复的项例如数组：[1,3,4,3,5]我们在做去重的时候，一开始想到的肯定是，逐个比较，外面一层循环，内层后一个与前一个一比较，如果是久不将当前这一项放进新的数组，挨个比较完之后返回一个新的去过重复的数组不好的实践方式上述方法效率极低，代码量还多，思考？有没有更好的方法这时候不禁一想当然有了！！！hashtable啊，通过对象的hash办法
在一台Ubuntu计算机上构建Hyperledger Fabric网络落叶无声9 区块链超级账本 Hyperledger fabric 区块链 ubuntu 构建 hyperledger fabric
在一台Ubuntu计算机上构建HyperledgerFabric网络Hyperledgerfabric是一个开源的区块链应用程序平台，为开发基于区块链的应用程序提供了一个起点。当我们提到HyperledgerFabric网络时，我们指的是使用HyperledgerFabric的正在运行的系统。即使只使用最少数量的组件，部署Fabric网络也不是一件容易的事。Fabric社区创建了一个名为Cello
GitHub上克隆项目 bigbig猩猩 github
从GitHub上克隆项目是一个简单且直接的过程，它允许你将远程仓库中的项目复制到你的本地计算机上，以便进行进一步的开发、测试或学习。以下是一个详细的步骤指南，帮助你从GitHub上克隆项目。一、准备工作1.安装Git在克隆GitHub项目之前，你需要在你的计算机上安装Git工具。Git是一个开源的分布式版本控制系统，用于跟踪和管理代码变更。你可以从Git的官方网站（https://git-scm.
Day1笔记-Python简介&标识符和关键字&输入输出 ~在杰难逃~ Python python 开发语言大数据数据分析数据挖掘
大家好，从今天开始呢，杰哥开展一个新的专栏，当然，数据分析部分也会不定时更新的，这个新的专栏主要是讲解一些Python的基础语法和知识，帮助0基础的小伙伴入门和学习Python，感兴趣的小伙伴可以开始认真学习啦！一、Python简介【了解】1.计算机工作原理编程语言就是用来定义计算机程序的形式语言。我们通过编程语言来编写程序代码，再通过语言处理程序执行向计算机发送指令，让计算机完成对应的工作，编程
【JS】执行时长(100分) |思路参考+代码解析（C++） l939035548 JS 算法数据结构 c++
题目为了充分发挥GPU算力，需要尽可能多的将任务交给GPU执行，现在有一个任务数组，数组元素表示在这1秒内新增的任务个数且每秒都有新增任务。假设GPU最多一次执行n个任务，一次执行耗时1秒，在保证GPU不空闲情况下，最少需要多长时间执行完成。题目输入第一个参数为GPU一次最多执行的任务个数，取值范围[1,10000]第二个参数为任务数组长度，取值范围[1,10000]第三个参数为任务数组，数字范围
ARM中断处理过程落汤老狗嵌入式linux
一、前言本文主要以ARM体系结构下的中断处理为例，讲述整个中断处理过程中的硬件行为和软件动作。具体整个处理过程分成三个步骤来描述：1、第二章描述了中断处理的准备过程2、第三章描述了当发生中的时候，ARM硬件的行为3、第四章描述了ARM的中断进入过程4、第五章描述了ARM的中断退出过程本文涉及的代码来自3.14内核。另外，本文注意描述ARM指令集的内容，有些sourcecode为了简短一些，删除了T
Python 实现图片裁剪（附代码） | Python工具剑客阿良_ALiang
前言本文提供将图片按照自定义尺寸进行裁剪的工具方法，一如既往的实用主义。环境依赖ffmpeg环境安装，可以参考我的另一篇文章：windowsffmpeg安装部署_阿良的博客-CSDN博客本文主要使用到的不是ffmpeg，而是ffprobe也在上面这篇文章中的zip包中。ffmpy安装：pipinstallffmpy-ihttps://pypi.douban.com/simple代码不废话了，上代码
【华为OD技术面试真题 - 技术面】- python八股文真题题库（4) 算法大师华为od 面试 python
华为OD面试真题精选专栏：华为OD面试真题精选目录:2024华为OD面试手撕代码真题目录以及八股文真题目录文章目录华为OD面试真题精选**1.Python中的`with`**用途和功能自动资源管理示例：文件操作上下文管理协议示例代码工作流程解析优点2.\_\_new\_\_和**\_\_init\_\_**区别__new____init__区别总结3.**切片（Slicing）操作**基本切片语法
【华为OD技术面试真题 - 技术面】-测试八股文真题题库（1）算法大师华为od 面试 python 算法前端
华为OD面试真题精选专栏：华为OD面试真题精选目录:2024华为OD面试手撕代码真题目录以及八股文真题目录文章目录华为OD面试真题精选1.黑盒测试和白盒测试的区别2.假设我们公司现在开发一个类似于微信的软件1.0版本，现在要你测试这个功能：打开聊天窗口，输入文本，限制字数在200字以内。问你怎么提取测试点。功能测试性能测试安全性测试可用性测试跨平台兼容性测试网络环境测试3.接口测试的工具你了解哪些
python os 环境变量 CV矿工 python 开发语言 numpy
环境变量：环境变量是程序和操作系统之间的通信方式。有些字符不宜明文写进代码里，比如数据库密码，个人账户密码，如果写进自己本机的环境变量里，程序用的时候通过os.environ.get（）取出来就行了。os.environ是一个环境变量的字典。环境变量的相关操作importos"""设置/修改环境变量：os.environ[‘环境变量名称’]=‘环境变量值’#其中key和value均为string类
Python爬虫解析工具之xpath使用详解 eqa11 python 爬虫开发语言
文章目录Python爬虫解析工具之xpath使用详解一、引言二、环境准备1、插件安装2、依赖库安装三、xpath语法详解1、路径表达式2、通配符3、谓语4、常用函数四、xpath在Python代码中的使用1、文档树的创建2、使用xpath表达式3、获取元素内容和属性五、总结Python爬虫解析工具之xpath使用详解一、引言在Python爬虫开发中，数据提取是一个至关重要的环节。xpath作为一门
01-Git初识 Meereen Git git
01-Git初识概念：一个免费开源，分布式的代码版本控制系统，帮助开发团队维护代码作用：记录代码内容。切换代码版本，多人开发时高效合并代码内容如何学：个人本机使用：Git基础命令和概念多人共享使用：团队开发同一个项目的代码版本管理Git配置用户信息配置：用户名和邮箱，应用在每次提交代码版本时表明自己的身份命令：查看git版本号git-v配置用户名gitconfig--globaluser.name
【PG】常见数据库、表属性设置江无羡数据库
PG的常见属性配置方法数据库复制、备份相关表的复制标识单表操作批量表操作链接数据库复制、备份相关表的复制标识单表操作通过ALTER语句单独更改一张表的复制标识。ALTERTABLE[tablename]REPLICAIDENTITYFULL;批量表操作通过代码块的方式，对某个schema中的所有表一起更新其复制标识。SELECTtablename,CASErelreplidentWHEN'd'TH
《在战“疫”中成长致敬生活》观后感梅子刘的刀
（作者：周晨）今天上午，我看了“我是接班人”网络大课堂《在战役中成长致敬生活》。有很多人拿出自己攒下的钱，默默地捐给了武汉，有几千块钱的、有几万块钱的，也有十几万块钱的。连小朋友也把自己的压岁钱捐给了武汉。有名环卫工人把自己五年的积蓄全部捐给了武汉。有名外卖小哥为医护人员买鞋子送吃的。还有已经治愈出院的新型肺炎病人捐了400毫升的血浆。还有位叫大树的叔叔，虽然他没有钱，但是他地里有蔬菜，捐了几大卡
【Git】常见命令(仅笔记) 好想有猫猫 Git Linux学习笔记 git 笔记 elasticsearch linux c++
文章目录创建/初始化本地仓库添加本地仓库配置项提交文件查看仓库状态回退仓库查看日志分支删除文件暂存工作区代码远程仓库使用`.gitigore`文件让git不追踪一些文件标签创建/初始化本地仓库gitinit添加本地仓库配置项gitconfig-l#以列表形式显示配置项gitconfiguser.name"ljh"#配置user.namegitconfiguser.email"[email protected]
【华为OD技术面试真题 - 技术面】- python八股文真题题库（1）算法大师华为od 面试 python
华为OD面试真题精选专栏：华为OD面试真题精选目录:2024华为OD面试手撕代码真题目录以及八股文真题目录文章目录华为OD面试真题精选1.数据预处理流程数据预处理的主要步骤工具和库2.介绍线性回归、逻辑回归模型线性回归（LinearRegression）模型形式：关键点：逻辑回归（LogisticRegression）模型形式：关键点：参数估计与评估：3.python浅拷贝及深拷贝浅拷贝（Shal
2019-01-19 王小康KK
姓名:王康公司:扬州市方圆建筑工程有限公司2018年3月16日～3月18日上海361期《六项精进》感谢二组学员【日精进打卡第307天】【知～学习】《六项精进》大纲3遍共862遍《大学》通篇3遍共860遍《六项精进》全书40页【经典名句】思想决定行为，行为决定习惯，习惯决定性格，性格决定命运。【行～实践】一、修身：（对自己个人）1、践行六项精进的理念。二、齐家：（对家庭和家人）1、和女朋友视频聊天。
jdk tomcat 环境变量配置 Array_06 java jdk tomcat
Win7 下如何配置java环境变量 1。准备jdk包，win7系统，tomcat安装包（均上网下载即可） 2。进行对jdk的安装，尽量为默认路径（但要记住啊！！以防以后配置用。。。） 3。分别配置高级环境变量。电脑-->右击属性-->高级环境变量-->环境变量。分别配置 : path &nbs
Spring调SDK包报java.lang.NoSuchFieldError错误 bijian1013 java spring
在工作中调另一个系统的SDK包，出现如下java.lang.NoSuchFieldError错误。 org.springframework.web.util.NestedServletException: Handler processing failed; nested exception is java.l
LeetCode[位运算] - #136 数组中的单一数 Cwind java 题解位运算 LeetCode Algorithm
原题链接：#136 Single Number 要求：给定一个整型数组，其中除了一个元素之外，每个元素都出现两次。找出这个元素注意：算法的时间复杂度应为O(n)，最好不使用额外的内存空间难度：中等分析：题目限定了线性的时间复杂度，同时不使用额外的空间，即要求只遍历数组一遍得出结果。由于异或运算 n XOR n = 0, n XOR 0 = n，故将数组中的每个元素进
qq登陆界面开发 15700786134 qq
今天我们来开发一个qq登陆界面，首先写一个界面程序，一个界面首先是一个Frame对象，即是一个窗体。然后在这个窗体上放置其他组件。代码如下： public class First { public void initul(){ jf=ne
Linux的程序包管理器RPM 被触发 linux
在早期我们使用源代码的方式来安装软件时，都需要先把源程序代码编译成可执行的二进制安装程序，然后进行安装。这就意味着每次安装软件都需要经过预处理-->编译-->汇编-->链接-->生成安装文件--> 安装，这个复杂而艰辛的过程。为简化安装步骤，便于广大用户的安装部署程序，程序提供商就在特定的系统上面编译好相关程序的安装文件并进行打包，提供给大家下载，我们只需要根据自己的
socket通信遇到EOFException 肆无忌惮_ EOFException
java.io.EOFException at java.io.ObjectInputStream$PeekInputStream.readFully(ObjectInputStream.java:2281) at java.io.ObjectInputStream$BlockDataInputStream.readShort(ObjectInputStream.java:
基于spring的web项目定时操作知了ing java Web
废话不多说，直接上代码，很简单配置一下项目启动就行 1，web.xml <?xml version="1.0" encoding="UTF-8"?> <web-app xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns="h
树形结构的数据库表Schema设计矮蛋蛋 schema
原文地址： http://blog.csdn.net/MONKEY_D_MENG/article/details/6647488 程序设计过程中，我们常常用树形结构来表征某些数据的关联关系，如企业上下级部门、栏目结构、商品分类等等，通常而言，这些树状结构需要借助于数据库完成持久化。然而目前的各种基于关系的数据库，都是以二维表的形式记录存储数据信息，
maven将jar包和源码一起打包到本地仓库 alleni123 maven
http://stackoverflow.com/questions/4031987/how-to-upload-sources-to-local-maven-repository <project> ... <build> <plugins> <plugin> <groupI
java IO操作与 File 获取文件或文件夹的大小，可读，等属性！！！百合不是茶
类 File File是指文件和目录路径名的抽象表示形式。 1，何为文件：标准文件（txt doc mp3...）目录文件（文件夹）虚拟内存文件 2，File类中有可以创建文件的 createNewFile（）方法,在创建新文件的时候需要try{} catch(）{}因为可能会抛出异常；也有可以判断文件是否是一个标准文件的方法isFile();这些防抖都
Spring注入有继承关系的类（2） bijian1013 java spring
被注入类的父类有相应的属性，Spring可以直接注入相应的属性，如下所例：1.AClass类 package com.bijian.spring.test4; public class AClass { private String a; private String b; public String getA() { retu
30岁转型期你能否成为成功人士 bijian1013 成长励志
很多人由于年轻时走了弯路，到了30岁一事无成，这样的例子大有人在。但同样也有一些人，整个职业生涯都发展得很优秀，到了30岁已经成为职场的精英阶层。由于做猎头的原因，我们接触很多30岁左右的经理人，发现他们在职业发展道路上往往有很多致命的问题。在30岁之前，他们的职业生涯表现很优秀，但从30岁到40岁这一段，很多人
【Velocity四】Velocity与Java互操作 bit1129 velocity
Velocity出现的目的用于简化基于MVC的web应用开发，用于替代JSP标签技术，那么Velocity如何访问Java代码.本篇继续以Velocity三http://bit1129.iteye.com/blog/2106142中的例子为基础， POJO package com.tom.servlets; public
【Hive十一】Hive数据倾斜优化 bit1129 hive
什么是Hive数据倾斜问题操作：join,group by,count distinct 现象：任务进度长时间维持在99%（或100%），查看任务监控页面，发现只有少量（1个或几个）reduce子任务未完成；查看未完成的子任务，可以看到本地读写数据量积累非常大，通常超过10GB可以认定为发生数据倾斜。原因：key分布不均匀倾斜度衡量：平均记录数超过50w且
在nginx中集成lua脚本：添加自定义Http头，封IP等 ronin47 nginx lua csrf
Lua是一个可以嵌入到Nginx配置文件中的动态脚本语言，从而可以在Nginx请求处理的任何阶段执行各种Lua代码。刚开始我们只是用Lua 把请求路由到后端服务器，但是它对我们架构的作用超出了我们的预期。下面就讲讲我们所做的工作。强制搜索引擎只索引mixlr.com Google把子域名当作完全独立的网站，我们不希望爬虫抓取子域名的页面，降低我们的Page rank。 location /{
java-3.求子数组的最大和 bylijinnan java
package beautyOfCoding; public class MaxSubArraySum { /** * 3.求子数组的最大和题目描述：输入一个整形数组，数组里有正数也有负数。数组中连续的一个或多个整数组成一个子数组，每个子数组都有一个和。求所有子数组的和的最大值。要求时间复杂度为O(n)。例如输入的数组为1, -2, 3, 10, -4,
Netty源码学习-FileRegion bylijinnan java netty
今天看org.jboss.netty.example.http.file.HttpStaticFileServerHandler.java 可以直接往channel里面写入一个FileRegion对象，而不需要相应的encoder： //pipeline（没有诸如“FileRegionEncoder”的handler）： public ChannelPipeline ge
使用ZeroClipboard解决跨浏览器复制到剪贴板的问题 cngolon 跨浏览器复制到粘贴板 Zero Clipboard
Zero Clipboard的实现原理 Zero Clipboard 利用透明的Flash让其漂浮在复制按钮之上，这样其实点击的不是按钮而是 Flash ，这样将需要的内容传入Flash，再通过Flash的复制功能把传入的内容复制到剪贴板。 Zero Clipboard的安装方法首先需要下载 Zero Clipboard的压缩包，解压后把文件夹中两个文件：ZeroClipboard.js
单例模式 cuishikuan 单例模式
第一种（懒汉，线程不安全）： public class Singleton { 2 private static Singleton instance; 3 pri
spring+websocket的使用 dalan_123
一、spring配置文件 <?xml version="1.0" encoding="UTF-8"?><beans xmlns="http://www.springframework.org/schema/beans" xmlns:xsi="http://www.w3.or
细节问题：ZEROFILL的用法范围。 dcj3sjt126com mysql
1、zerofill把月份中的一位数字比如1，2，3等加前导0 mysql> CREATE TABLE t1 (year YEAR(4), month INT(2) UNSIGNED ZEROFILL, -> day
Android开发10——Activity的跳转与传值 dcj3sjt126com Android开发
Activity跳转与传值，主要是通过Intent类，Intent的作用是激活组件和附带数据。一、Activity跳转方法一Intent intent = new Intent(A.this, B.class); startActivity(intent) 方法二Intent intent = new Intent();intent.setCla
jdbc 得到表结构、主键 eksliang jdbc 得到表结构、主键
转自博客：http://blog.csdn.net/ocean1010/article/details/7266042 假设有个con DatabaseMetaData dbmd = con.getMetaData(); rs = dbmd.getColumns(con.getCatalog(), schema, tableName, null); rs.getSt
Android 应用程序开关GPS gqdy365 android
要在应用程序中操作GPS开关需要权限： <uses-permission android:name="android.permission.WRITE_SECURE_SETTINGS" /> 但在配置文件中添加此权限之后会报错，无法再eclipse里面正常编译，怎么办？ 1、方法一：将项目放到Android源码中编译； 2、方法二：网上有人说cl
Windows上调试MapReduce zhiquanliu mapreduce
1.下载hadoop2x-eclipse-plugin https://github.com/winghc/hadoop2x-eclipse-plugin.git 把 hadoop2.6.0-eclipse-plugin.jar 放到eclipse plugin 目录中。 2.下载 hadoop2.6_x64_.zip http://dl.iteye.com/topics/download/d2b
如何看待一些知名博客推广软文的行为？ justjavac 博客
本文来自我在知乎上的一个回答：http://www.zhihu.com/question/23431810/answer/24588621 互联网上的两种典型心态：当初求种像条狗，如今撸完嫌人丑当初搜贴像条犬，如今读完嫌人软你为啥感觉不舒服呢？难道非得要作者把自己的劳动成果免费给你用，你才舒服？就如同 Google 关闭了 Gooled Reader，那是
sql优化总结 macroli sql
为了是自己对sql优化有更好的原则性，在这里做一下总结，个人原则如有不对请多多指教。谢谢！要知道一个简单的sql语句执行效率，就要有查看方式，一遍更好的进行优化。一、简单的统计语句执行时间 declare @d datetime ---定义一个datetime的变量set @d=getdate() ---获取查询语句开始前的时间select user_id
Linux Oracle中常遇到的一些问题及命令总结超声波 oracle linux
1.linux更改主机名 (1)#hostname oracledb　　　　临时修改主机名 (2) vi /etc/sysconfig/network 　　修改hostname (3) vi /etc/hosts　　　　　　　　修改IP对应的主机名 2.linux重启oracle实例及监听的各种方法（注意操作的顺序应该是先监听，后数据库实例） &nbs
hive函数大全及使用示例 superlxw1234 hadoop hive函数
具体说明及示例参见附件文档。文档目录：目录一、关系运算： 4 1. 等值比较: = 4 2. 不等值比较: <> 4 3. 小于比较: < 4 4. 小于等于比较: <= 4 5. 大于比较: > 5 6. 大于等于比较: >= 5 7. 空值判断: IS NULL 5
Spring 4.2新特性-使用@Order调整配置类加载顺序 wiselyman spring 4
4.1 @Order Spring 4.2 利用@Order控制配置类的加载顺序 4.2 演示两个演示bean package com.wisely.spring4_2.order; public class Demo1Service { } package com.wisely.spring4_2.order; public class