-
Notifications
You must be signed in to change notification settings - Fork 3
/
Classification.py
148 lines (108 loc) · 4.09 KB
/
Classification.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
# coding: utf-8
# In[ ]:
from keras.layers import Input, Dense, Flatten, Dropout
from keras.layers.core import Reshape
from keras.models import Model
from keras.callbacks import ModelCheckpoint
from keras.layers.convolutional import MaxPooling2D,UpSampling2D,Conv2DTranspose
from keras.layers.convolutional import Convolution2D as Conv2D
from keras.layers.normalization import BatchNormalization
from keras.optimizers import Adadelta, RMSprop, adam, SGD
from keras import backend as K
from keras.models import model_from_json
from keras.utils import np_utils
import os
import os.path
import numpy as np
from PIL import Image
from numpy import *
from sklearn.utils import shuffle
from sklearn.cross_validation import train_test_split
import scipy.misc
import cv2
import matplotlib.pyplot as plt
import matplotlib
#from sklearn.metrics import classification_report,confusion_matrix
# In[ ]:
input_img = Input(shape=(128,128,1))
num_classes = 123
img_rows, img_cols = 128, 128
# In[ ]:
enco = Conv2D(16, (3, 3), activation='relu', padding='same')(input_img)
enco = BatchNormalization()(enco)
enco = Conv2D(16, (3, 3), activation='relu', padding='same')(enco)
enco = BatchNormalization()(enco)
enco = MaxPooling2D(pool_size=(2, 2))(enco)
enco = Conv2D(32, (3, 3), activation='relu', padding='same')(enco)
enco = BatchNormalization()(enco)
enco = Conv2D(32, (3, 3), activation='relu', padding='same')(enco)
enco = BatchNormalization()(enco)
enco = MaxPooling2D(pool_size=(2, 2))(enco)
enco = Conv2D(64, (3, 3), activation='relu', padding='same')(enco)
enco = BatchNormalization()(enco)
enco = Conv2D(64, (3, 3), activation='relu', padding='same')(enco)
enco = BatchNormalization()(enco)
enco = MaxPooling2D(pool_size=(2, 2))(enco)
enco = Conv2D(128, (3, 3), activation='relu', padding='same')(enco)
enco = BatchNormalization()(enco)
enco = Conv2D(128, (3, 3), activation='relu', padding='same')(enco)
enco = BatchNormalization()(enco)
#enco.load_weights("Only_Encoder.h5")
encoder = Model(input_img, enco)
encoder.load_weights("Only_Encoder.h5")
classify = Flatten()(enco)
classify = Dense(64, activation='relu')(classify)
classify = Dense(32, activation='relu')(classify)
classify = Dense(num_classes, activation='softmax')(classify)
#network = Model(enco, classify)
network = Model(input_img, classify)
rms=RMSprop(lr=0.001, rho=0.9, epsilon=1e-08, decay=0.001)
#network.compile(loss='mean_squared_error', optimizer=rms)
network.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])
#network.summary()
for layers in encoder.layers:
layers.trainable=False
#network.summary()
#exit()
# In[ ]:
path="Data"
basic_mat=[]
index=[]
epochs=100
batch_size1=64
# In[ ]:
for i in range(1,124):
path_major=path+'/'+str(i)
for j in range(1,101):
img=array(Image.open(path_major+"/"+str(j)+"_.jpg"))
#print shape(img)
img = cv2.cvtColor( img, cv2.COLOR_RGB2GRAY )
img=img.reshape(128,128,1)
basic_mat.append(img)
index.append(i-1)
network.summary()
# In[ ]:
data,Label = shuffle(basic_mat,index, random_state=2)
# In[ ]:
X_train, X_test, y_train, y_test = train_test_split(data, Label, test_size=0.2, random_state=2)
X_train = array(X_train)
y_train = array(y_train)
X_test = array(X_test)
y_test = array(y_test)
print(X_train.shape)
print(y_train.shape)
print(X_test.shape)
print(y_test.shape)
y_train = np_utils.to_categorical(y_train)
y_test = np_utils.to_categorical(y_test)
# In[ ]:
x_train = X_train.astype('float32') / 255.
x_test = X_test.astype('float32') / 255.
network.fit(x_train, y_train, validation_data=(x_test, y_test),batch_size=batch_size1, nb_epoch=epochs, verbose=2)
scores = network.evaluate(x_test, y_test, verbose=0)
print ("%s: %.2f%%" % (network.metrics_names[1], scores[1]*100))
network.save_weights("only_classify.h5")
# final_network.fit(x_train, y_train, validation_data=(x_test, y_test),batch_size=batch_size1, nb_epoch=epochs, verbose=2)
# scores_final = final_network.evaluate(x_test, y_test, verbose=0)
# print ("%s: %.2f%%" % (final_network.metrics_names[1], scores_final[1]*100))
# final_network.save_weights("complete_model.h5")