# coding: utf-8

# In[5]:

from keras.applications.resnet50 import ResNet50

from keras.applications.resnet50 import preprocess_input

from keras.applications.imagenet_utils import decode_predictions

import os

import numpy as np

from keras.layers import merge, Input, Lambda

from keras.layers import Dense, Activation, Flatten

from keras.layers import Convolution2D, MaxPooling2D, ZeroPadding2D, AveragePooling2D

from keras.layers import BatchNormalization

from keras.models import Model

import pickle

import matplotlib.pyplot as plt

from keras.datasets import mnist

from keras.layers import Conv2D, Dense, GlobalAveragePooling2D

from keras.models import Model, Input

from keras.utils import to_categorical

import os

import numpy as np

import pandas as pd

import keras

from pandas import DataFrame

from keras import models

from keras.layers import Conv1D, Concatenate, merge

from keras.models import load_model, Sequential

from keras.callbacks import EarlyStopping, ModelCheckpoint

from sklearn.model_selection import train_test_split

import keras,pickle

from keras.layers import Activation, Dense, InputLayer, MaxPool1D, Flatten, Dropout, AvgPool1D, Input, concatenate, Concatenate, Add, Reshape

from keras.utils import np_utils

from keras.utils import np_utils

from keras.layers.merge import Concatenate

from keras.models import Model

np.random.seed(10)

from tensorflow import set_random_seed

set_random_seed(15)

from pprint import pprint

from keras.layers.embeddings import Embedding

from numpy import newaxis

from keras import backend as K

# from keras.utils import plot_model

# In[22]:

def my_simple_cnn(inputs) :

return x

input1 = Input(shape=(224,224, 3))

input2 = Input(shape=(224,224, 3))

input3 = Input(shape=(224,224, 3))

input4 = Input(shape=(224,224, 3))

input5 = Input(shape=(224,224, 3))

input6 = Input(shape=(224,224, 3))

input7 = Input(shape=(224,224, 3))

input8 = Input(shape=(224,224, 3))

input9 = Input(shape=(224,224, 3))

input10 = Input(shape=(224,224, 3))

input11 = Input(shape=(224,224, 3))

input12 = Input(shape=(224,224, 3))

input13 = Input(shape=(224,224, 3))

input14 = Input(shape=(224,224, 3))

input15 = Input(shape=(224,224, 3))

input16 = Input(shape=(224,224, 3))

input17 = Input(shape=(224,224, 3))

input18 = Input(shape=(224,224, 3))

input19 = Input(shape=(224,224, 3))

input20 = Input(shape=(224,224, 3))

input21 = Input(shape=(224,224, 3))

input22 = Input(shape=(224,224, 3))

input23 = Input(shape=(224,224, 3))

input24 = Input(shape=(224,224, 3))

input25 = Input(shape=(224,224, 3))

input26 = Input(shape=(224,224, 3))

input27 = Input(shape=(224,224, 3))

input28 = Input(shape=(224,224, 3))

input29 = Input(shape=(224,224, 3))

input30 = Input(shape=(224,224, 3))

output1 = my_simple_cnn(input1)

output2 = my_simple_cnn(input2)

output3 = my_simple_cnn(input3)

output4 = my_simple_cnn(input4)

output5 = my_simple_cnn(input5)

output6 = my_simple_cnn(input6)

output7 = my_simple_cnn(input7)

output8 = my_simple_cnn(input8)

output9 = my_simple_cnn(input9)

output10 = my_simple_cnn(input10)

output11 = my_simple_cnn(input11)

output12 = my_simple_cnn(input12)

output13 = my_simple_cnn(input13)

output14 = my_simple_cnn(input14)

output15 = my_simple_cnn(input15)

output16 = my_simple_cnn(input16)

output17 = my_simple_cnn(input17)

output18 = my_simple_cnn(input18)

output19 = my_simple_cnn(input19)

output20= my_simple_cnn(input20)

output21 = my_simple_cnn(input21)

output22 = my_simple_cnn(input22)

output23 = my_simple_cnn(input23)

output24 = my_simple_cnn(input24)

output25 = my_simple_cnn(input25)

output26 = my_simple_cnn(input26)

output27 = my_simple_cnn(input27)

output28 = my_simple_cnn(input28)

output29 = my_simple_cnn(input29)

output30 = my_simple_cnn(input30)

my_inputs = [input1,input2,input3,input4,input5,input6,

input7,input8,input9,input10,input11,input12,input13,input14,input15,input16,input17,input18,input19,input20,

input21,input22,input23,input24,input25,input26,input27,input28,input29,input30]

my_outputs = [output1,output2,output3,output4,output5,output6,

output7,output8,output9,output10,output11,output12,output13,output14,

output15,output16,output17,output18,output19,output20,output21,output22,

output23,output24,output25,output26,output27,output28,output29,output30]

concat_output = Concatenate(axis = -1)(my_outputs)

predictions = Flatten()(concat_output)

predictions = Dense(300, activation='relu')(predictions)

predictions = Dense(200, activation='relu')(predictions)

predictions = Dense(100, activation='relu')(predictions)

predictions = Dense(2, activation='softmax')(predictions)

from keras.optimizers import SGD

opt = SGD(lr=0.001)

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

model.compile(optimizer=opt,

loss='categorical_crossentropy',

metrics=['accuracy'])

print(model.summary())

# In[23]:

from glob import glob

from random import shuffle

pat_list0 = glob('C:/Users/korea/Desktop/sungmo_RESNET/data/30image/0/*')

pat_list1 = glob('C:/Users/korea/Desktop/sungmo_RESNET/data/30image/1/*')

from pprint import pprint

######1

file_list0_1 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_1.append(i + "/" + str(tmp_list[0]))

pprint(file_list0_1)

file_list1_1 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_1.append(i + "/" + str(tmp_list[0]))

pprint(file_list1_1)

######2

file_list0_2 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_2.append(i + "/" + str(tmp_list[1]))

pprint(file_list0_2)

file_list1_2 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_2.append(i + "/" + str(tmp_list[1]))

pprint(file_list1_2)

######3

file_list0_3 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_3.append(i + "/" + str(tmp_list[2]))

pprint(file_list0_3)

file_list1_3 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_3.append(i + "/" + str(tmp_list[2]))

pprint(file_list1_3)

######4

file_list0_4 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_4.append(i + "/" + str(tmp_list[3]))

pprint(file_list0_4)

file_list1_4 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_4.append(i + "/" + str(tmp_list[3]))

pprint(file_list1_4)

######5

file_list0_5 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_5.append(i + "/" + str(tmp_list[4]))

pprint(file_list0_5)

file_list1_5 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_5.append(i + "/" + str(tmp_list[4]))

pprint(file_list1_5)

######6

file_list0_6 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_6.append(i + "/" + str(tmp_list[5]))

pprint(file_list0_6)

file_list1_6 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_6.append(i + "/" + str(tmp_list[5]))

pprint(file_list1_6)

######7

file_list0_7 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_7.append(i + "/" + str(tmp_list[6]))

pprint(file_list0_7)

file_list1_7 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_7.append(i + "/" + str(tmp_list[6]))

pprint(file_list1_7)

######8

file_list0_8 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_8.append(i + "/" + str(tmp_list[7]))

pprint(file_list0_8)

file_list1_8 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_8.append(i + "/" + str(tmp_list[7]))

pprint(file_list1_8)

######9

file_list0_9 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_9.append(i + "/" + str(tmp_list[8]))

pprint(file_list0_9)

file_list1_9 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_9.append(i + "/" + str(tmp_list[8]))

pprint(file_list1_9)

######10

file_list0_10 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_10.append(i + "/" + str(tmp_list[9]))

pprint(file_list0_10)

file_list1_10 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_10.append(i + "/" + str(tmp_list[9]))

pprint(file_list1_10)

######11

file_list0_11 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_11.append(i + "/" + str(tmp_list[10]))

pprint(file_list0_11)

file_list1_11 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_11.append(i + "/" + str(tmp_list[10]))

pprint(file_list1_11)

######12

file_list0_12 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_12.append(i + "/" + str(tmp_list[11]))

pprint(file_list0_12)

file_list1_12 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_12.append(i + "/" + str(tmp_list[11]))

pprint(file_list1_12)

######13

file_list0_13 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_13.append(i + "/" + str(tmp_list[12]))

pprint(file_list0_13)

file_list1_13 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_13.append(i + "/" + str(tmp_list[12]))

pprint(file_list1_13)

######14

file_list0_14 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_14.append(i + "/" + str(tmp_list[13]))

pprint(file_list0_14)

file_list1_14 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_14.append(i + "/" + str(tmp_list[13]))

pprint(file_list1_14)

######15

file_list0_15 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_15.append(i + "/" + str(tmp_list[14]))

pprint(file_list0_15)

file_list1_15 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_15.append(i + "/" + str(tmp_list[14]))

pprint(file_list1_15)

######16

file_list0_16 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_16.append(i + "/" + str(tmp_list[15]))

pprint(file_list0_16)

file_list1_16 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_16.append(i + "/" + str(tmp_list[15]))

pprint(file_list1_16)

######17

file_list0_17 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_17.append(i + "/" + str(tmp_list[16]))

pprint(file_list0_17)

file_list1_17 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_17.append(i + "/" + str(tmp_list[16]))

pprint(file_list1_17)

######18

file_list0_18 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_18.append(i + "/" + str(tmp_list[17]))

pprint(file_list0_18)

file_list1_18 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_18.append(i + "/" + str(tmp_list[17]))

pprint(file_list1_18)

######19

file_list0_19 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_19.append(i + "/" + str(tmp_list[18]))

pprint(file_list0_19)

file_list1_19 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_19.append(i + "/" + str(tmp_list[18]))

pprint(file_list1_19)

######20

file_list0_20 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_20.append(i + "/" + str(tmp_list[19]))

pprint(file_list0_20)

file_list1_20 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_20.append(i + "/" + str(tmp_list[19]))

pprint(file_list1_20)

######21

file_list0_21 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_21.append(i + "/" + str(tmp_list[20]))

pprint(file_list0_21)

file_list1_21 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_21.append(i + "/" + str(tmp_list[20]))

pprint(file_list1_21)

######22

file_list0_22 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_22.append(i + "/" + str(tmp_list[21]))

pprint(file_list0_22)

file_list1_22 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_22.append(i + "/" + str(tmp_list[21]))

pprint(file_list1_22)

######23

file_list0_23 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_23.append(i + "/" + str(tmp_list[22]))

pprint(file_list0_23)

file_list1_23 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_23.append(i + "/" + str(tmp_list[22]))

pprint(file_list1_23)

######24

file_list0_24 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_24.append(i + "/" + str(tmp_list[23]))

pprint(file_list0_24)

file_list1_24 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_24.append(i + "/" + str(tmp_list[23]))

pprint(file_list1_24)

######25

file_list0_25 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_25.append(i + "/" + str(tmp_list[24]))

pprint(file_list0_25)

file_list1_25 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_25.append(i + "/" + str(tmp_list[24]))

pprint(file_list1_25)

######26

file_list0_26 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_26.append(i + "/" + str(tmp_list[25]))

pprint(file_list0_26)

file_list1_26 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_26.append(i + "/" + str(tmp_list[25]))

pprint(file_list1_26)

######27

file_list0_27 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_27.append(i + "/" + str(tmp_list[26]))

pprint(file_list0_27)

file_list1_27 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_27.append(i + "/" + str(tmp_list[26]))

pprint(file_list1_27)

######28

file_list0_28 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_28.append(i + "/" + str(tmp_list[27]))

pprint(file_list0_28)

file_list1_28 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_28.append(i + "/" + str(tmp_list[27]))

pprint(file_list1_28)

######29

file_list0_29 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_29.append(i + "/" + str(tmp_list[28]))

pprint(file_list0_29)

file_list1_29 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_29.append(i + "/" + str(tmp_list[28]))

pprint(file_list1_29)

######30

file_list0_30 = []

for i in pat_list0:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list0_30.append(i + "/" + str(tmp_list[29]))

pprint(file_list0_30)

file_list1_30 = []

for i in pat_list1:

tmp_list = os.listdir(i)

tmp_list = sorted(tmp_list, key = int)

file_list1_30.append(i + "/" + str(tmp_list[29]))

pprint(file_list1_30)

def my_new_axis(x):

x= x[newaxis,:,:,:]

return x

# 첫번째 CNN에 들어갈 Input임

input_image_for_cnn_1 = pickle.load(open(file_list0_1[0],'rb'))

input_image_for_cnn_1 = my_new_axis(input_image_for_cnn_1)

for i in range(len(file_list0_1)):

img = pickle.load(open(file_list0_1[i],'rb'))

img = img/255.0

img = my_new_axis(img)

input_image_for_cnn_1 = np.append(input_image_for_cnn_1, img, axis = 0)

input_image_for_cnn_1 = input_image_for_cnn_1[1:,:,:,:]

for j in range(len(file_list1_1)):

img = pickle.load(open(file_list1_1[j],'rb'))

img = img/255.0

img = my_new_axis(img)

input_image_for_cnn_1 = np.append(input_image_for_cnn_1, img, axis = 0)

print(input_image_for_cnn_1.shape)

# 두번째 CNN에 들어갈 Input임

input_image_for_cnn_2 = pickle.load(open(file_list0_2[0],'rb'))

input_image_for_cnn_2 = my_new_axis(input_image_for_cnn_2)

for i in range(len(file_list0_2)):

img = pickle.load(open(file_list0_2[i],'rb'))

img = img/255.0

img = my_new_axis(img)

input_image_for_cnn_2 = np.append(input_image_for_cnn_2, img, axis = 0)

input_image_for_cnn_2 = input_image_for_cnn_2[1:,:,:,:]

for j in range(len(file_list1_2)):

img = pickle.load(open(file_list1_2[j],'rb'))

img = img/255.0

img = my_new_axis(img)

input_image_for_cnn_2 = np.append(input_image_for_cnn_2, img, axis = 0)

print(input_image_for_cnn_2.shape)

# 세번째 CNN에 들어갈 Input임

input_image_for_cnn_3 = pickle.load(open(file_list0_3[0],'rb'))

input_image_for_cnn_3 = my_new_axis(input_image_for_cnn_3)

for i in range(len(file_list0_3)):

img = pickle.load(open(file_list0_3[i],'rb'))

img = img/255.0

img = my_new_axis(img)

input_image_for_cnn_3 = np.append(input_image_for_cnn_3, img, axis = 0)

input_image_for_cnn_3 = input_image_for_cnn_3[1:,:,:,:]

for j in range(len(file_list1_3)):

img = pickle.load(open(file_list1_3[j],'rb'))

img = img/255.0

img = my_new_axis(img)

input_image_for_cnn_3 = np.append(input_image_for_cnn_3, img, axis = 0)

print(input_image_for_cnn_3.shape)

# 4번째 CNN에 들어갈 Input임

input_image_for_cnn_4 = pickle.load(open(file_list0_4[0],'rb'))

input_image_for_cnn_4 = my_new_axis(input_image_for_cnn_4)

for i in range(len(file_list0_4)):

img = pickle.load(open(file_list0_4[i],'rb'))

img = img/255.0

img = my_new_axis(img)

input_image_for_cnn_4 = np.append(input_image_for_cnn_4, img, axis = 0)

input_image_for_cnn_4 = input_image_for_cnn_4[1:,:,:,:]

for j in range(len(file_list1_4)):

img = pickle.load(open(file_list1_4[j],'rb'))

img = img/255.0

img = my_new_axis(img)

input_image_for_cnn_4 = np.append(input_image_for_cnn_4, img, axis = 0)

print(input_image_for_cnn_4.shape)

# 5번째 CNN에 들어갈 Input임

input_image_for_cnn_5 = pickle.load(open(file_list0_5[0],'rb'))

input_image_for_cnn_5 = my_new_axis(input_image_for_cnn_5)

for i in range(len(file_list0_5)):

img = pickle.load(open(file_list0_5[i],'rb'))

img = img/255.0

img = my_new_axis(img)

input_image_for_cnn_5 = np.append(input_image_for_cnn_5, img, axis = 0)

input_image_for_cnn_5 = input_image_for_cnn_5[1:,:,:,:]

for j in range(len(file_list1_5)):

img = pickle.load(open(file_list1_5[j],'rb'))

img = img/255.0

img = my_new_axis(img)

input_image_for_cnn_5 = np.append(input_image_for_cnn_5, img, axis = 0)

print(input_image_for_cnn_5.shape)

# 6번째 CNN에 들어갈 Input임

input_image_for_cnn_6 = pickle.load(open(file_list0_6[0],'rb'))

input_image_for_cnn_6 = my_new_axis(input_image_for_cnn_6)

for i in range(len(file_list0_6)):

img = pickle.load(open(file_list0_6[i],'rb'))

img = img/255.0

img = my_new_axis(img)

input_image_for_cnn_6 = np.append(input_image_for_cnn_6, img, axis = 0)

input_image_for_cnn_6 = input_image_for_cnn_6[1:,:,:,:]

for j in range(len(file_list1_6)):

img = pickle.load(open(file_list1_6[j],'rb'))

img = img/255.0

img = my_new_axis(img)

input_image_for_cnn_6 = np.append(input_image_for_cnn_6, img, axis = 0)

print(input_image_for_cnn_6.shape)

# 7번째 CNN에 들어갈 Input임

input_image_for_cnn_7 = pickle.load(open(file_list0_7[0],'rb'))

input_image_for_cnn_7 = my_new_axis(input_image_for_cnn_7)

for i in range(len(file_list0_7)):

img = pickle.load(open(file_list0_7[i],'rb'))

img = img/255.0

img = my_new_axis(img)

input_image_for_cnn_7 = np.append(input_image_for_cnn_7, img, axis = 0)

input_image_for_cnn_7 = input_image_for_cnn_7[1:,:,:,:]

for j in range(len(file_list1_7)):

img = pickle.load(open(file_list1_7[j],'rb'))

img = img/255.0

img = my_new_axis(img)

input_image_for_cnn_7 = np.append(input_image_for_cnn_7, img, axis = 0)

print(input_image_for_cnn_7.shape)

# 8번째 CNN에 들어갈 Input임

input_image_for_cnn_8 = pickle.load(open(file_list0_8[0],'rb'))

input_image_for_cnn_8 = my_new_axis(input_image_for_cnn_8)

for i in range(len(file_list0_8)):

img = pickle.load(open(file_list0_8[i],'rb'))

img = img/255.0

img = my_new_axis(img)

input_image_for_cnn_8 = np.append(input_image_for_cnn_8, img, axis = 0)

input_image_for_cnn_8 = input_image_for_cnn_8[1:,:,:,:]

for j in range(len(file_list1_8)):

img = pickle.load(open(file_list1_8[j],'rb'))

img = img/255.0

img = my_new_axis(img)

input_image_for_cnn_8 = np.append(input_image_for_cnn_8, img, axis = 0)

print(input_image_for_cnn_8.shape)

# 9번째 CNN에 들어갈 Input임

input_image_for_cnn_9 = pickle.load(open(file_list0_9[0],'rb'))

input_image_for_cnn_9 = my_new_axis(input_image_for_cnn_9)

for i in range(len(file_list0_9)):

img = pickle.load(open(file_list0_9[i],'rb'))

img = img/255.0

img = my_new_axis(img)

input_image_for_cnn_9 = np.append(input_image_for_cnn_9, img, axis = 0)

input_image_for_cnn_9 = input_image_for_cnn_9[1:,:,:,:]

for j in range(len(file_list1_9)):

img = pickle.load(open(file_list1_9[j],'rb'))

img = img/255.0

img = my_new_axis(img)

input_image_for_cnn_9 = np.append(input_image_for_cnn_9, img, axis = 0)

print(input_image_for_cnn_9.shape)

# 10번째 CNN에 들어갈 Input임

input_image_for_cnn_10 = pickle.load(open(file_list0_10[0],'rb'))

input_image_for_cnn_10 = my_new_axis(input_image_for_cnn_10)

for i in range(len(file_list0_10)):

img = pickle.load(open(file_list0_10[i],'rb'))

img = img/255.0

img = my_new_axis(img)

input_image_for_cnn_10 = np.append(input_image_for_cnn_10, img, axis = 0)

input_image_for_cnn_10 = input_image_for_cnn_10[1:,:,:,:]

for j in range(len(file_list1_10)):

img = pickle.load(open(file_list1_10[j],'rb'))

img = img/255.0

img = my_new_axis(img)

input_image_for_cnn_10 = np.append(input_image_for_cnn_10, img, axis = 0)

print(input_image_for_cnn_10.shape)

# 11번째 CNN에 들어갈 Input임

input_image_for_cnn_11 = pickle.load(open(file_list0_11[0],'rb'))

input_image_for_cnn_11 = my_new_axis(input_image_for_cnn_11)

for i in range(len(file_list0_11)):

img = pickle.load(open(file_list0_11[i],'rb'))

img = img/255.0

img = my_new_axis(img)

input_image_for_cnn_11 = np.append(input_image_for_cnn_11, img, axis = 0)

input_image_for_cnn_11 = input_image_for_cnn_11[1:,:,:,:]

for j in range(len(file_list1_11)):

img = pickle.load(open(file_list1_11[j],'rb'))

img = img/255.0

img = my_new_axis(img)

input_image_for_cnn_11 = np.append(input_image_for_cnn_11, img, axis = 0)

print(input_image_for_cnn_11.shape)

# 12번째 CNN에 들어갈 Input임

input_image_for_cnn_12 = pickle.load(open(file_list0_12[0],'rb'))

input_image_for_cnn_12 = my_new_axis(input_image_for_cnn_12)

for i in range(len(file_list0_12)):

img = pickle.load(open(file_list0_12[i],'rb'))

img = img/255.0

img = my_new_axis(img)

input_image_for_cnn_12 = np.append(input_image_for_cnn_12, img, axis = 0)

input_image_for_cnn_12 = input_image_for_cnn_12[1:,:,:,:]

for j in range(len(file_list1_12)):

img = pickle.load(open(file_list1_12[j],'rb'))

img = img/255.0

img = my_new_axis(img)

input_image_for_cnn_12 = np.append(input_image_for_cnn_12, img, axis = 0)

print(input_image_for_cnn_12.shape)

# 13번째 CNN에 들어갈 Input임

input_image_for_cnn_13 = pickle.load(open(file_list0_13[0],'rb'))

input_image_for_cnn_13 = my_new_axis(input_image_for_cnn_13)

for i in range(len(file_list0_13)):

img = pickle.load(open(file_list0_13[i],'rb'))

img = img/255.0

img = my_new_axis(img)

input_image_for_cnn_13 = np.append(input_image_for_cnn_13, img, axis = 0)

input_image_for_cnn_13 = input_image_for_cnn_13[1:,:,:,:]

for j in range(len(file_list1_13)):

img = pickle.load(open(file_list1_13[j],'rb'))

img = img/255.0

img = my_new_axis(img)

input_image_for_cnn_13 = np.append(input_image_for_cnn_13, img, axis = 0)

print(input_image_for_cnn_13.shape)

# 14번째 CNN에 들어갈 Input임

input_image_for_cnn_14 = pickle.load(open(file_list0_14[0],'rb'))

input_image_for_cnn_14 = my_new_axis(input_image_for_cnn_14)