CNN:畳み込みニューラルネットワーク

##LeNet
###概要

1998年、Yann LeCun氏が発案

論文: http://yann.lecun.com/exdb/publis/pdf/lecun-98.pdf

###構造

##AlexNet
###概要
最近のディープラーニングブームの火付け役

2014年

①VGG16

import keras
from keras.layers import Conv2D, MaxPooling2D, Input, Dense, Flatten
from keras.models import Model

inputs = Input(shape=(224, 224, 3))
x = Conv2D(64, (3, 3), activation='relu', padding='same', name='block1_conv1')(inputs)
x = Conv2D(64, (3, 3), activation='relu', padding='same', name='block1_conv2')(x)
x = MaxPooling2D((2, 2), strides=(2, 2), padding='valid', name='block1_pool')(x)
x = Conv2D(128, (3, 3), activation='relu', padding='same', name='block2_conv1')(x)
x = Conv2D(128, (3, 3), activation='relu', padding='same', name='block2_conv2')(x)
x = MaxPooling2D((2, 2), strides=(2, 2), padding='valid', name='block2_pool')(x)
x = Conv2D(256, (3, 3), activation='relu', padding='same', name='block3_conv1')(x)
x = Conv2D(256, (3, 3), activation='relu', padding='same', name='block3_conv2')(x)
x = Conv2D(256, (3, 3), activation='relu', padding='same', name='block3_conv3')(x)
x = MaxPooling2D((2, 2), strides=(2, 2), padding='valid', name='block3_pool')(x)
x = Conv2D(512, (3, 3), activation='relu', padding='same', name='block4_conv1')(x)
x = Conv2D(512, (3, 3), activation='relu', padding='same', name='block4_conv2')(x)
x = Conv2D(512, (3, 3), activation='relu', padding='same', name='block4_conv3')(x)
x = MaxPooling2D((2, 2), strides=(2, 2), padding='valid', name='block4_pool')(x)
x = Conv2D(512, (3, 3), activation='relu', padding='same', name='block5_conv1')(x)
x = Conv2D(512, (3, 3), activation='relu', padding='same', name='block5_conv2')(x)
x = Conv2D(512, (3, 3), activation='relu', padding='same', name='block5_conv3')(x)
x = MaxPooling2D((2, 2), strides=(2, 2), padding='valid', name='block5_pool')(x)
flattened = Flatten(name='flatten')(x)
x = Dense(4096, activation='relu', name='fc1')(flattened)
x = Dense(4096, activation='relu', name='fc2')(x)
predictions = Dense(1000, activation='softmax', name='predictions')(x)

model = Model(inputs=inputs, outputs=predictions)

model.summary()

②AlexNet

import keras
from keras.layers import Conv2D, MaxPooling2D, Input, Dense, Flatten, BatchNormalization
from keras.models import Model

inputs = Input(shape=(227, 227, 3))

# Block 1
x = Conv2D(96, (11, 11), padding='valid', strides=(4, 4), activation='relu', name='conv1')(inputs) #out:55*55*96
x = MaxPooling2D((3, 3), padding='valid', strides=(2, 2), name='pool1')(x) #out:27*27*96
x = BatchNormalization(name='norm1')(x)

# Block 2
x = Conv2D(256, (5, 5), padding='same',  strides=(1, 1), activation='relu', name='conv2')(x) #out:27*27*256 padding:2
x = MaxPooling2D((3, 3), padding='valid',  strides=(2, 2), name='pool2')(x) #out:13*13*256
x = BatchNormalization(name='norm2')(x)

# Block 3
x = Conv2D(384, (3, 3), padding='same', strides=(1, 1), activation='relu', name='conv3')(x) #out:13*13*384 padding:1
x = Conv2D(384, (3, 3), padding='same', strides=(1, 1), activation='relu', name='conv4')(x) #out:13*13*384 padding:1
x = Conv2D(256, (3, 3), padding='same',  strides=(1, 1), activation='relu', name='conv5')(x) #out:13*13*256 padding:1
x = MaxPooling2D((3, 3), padding='valid',  strides=(2, 2), name='pool3')(x) #out:6*6*256

# Classification block
flattened = Flatten(name='flatten')(x) # out:9216
x = Dense(4096, activation='relu', name='fc1')(flattened)  # out:4096
x = Dense(4096, activation='relu', name='fc2')(x) # out:4096

predictions = Dense(1000, activation='softmax', name='predictions')(x)

model = Model(inputs=inputs, outputs=predictions)

model.summary()

③GoogLeNet

④ResNet