-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathvis.py
216 lines (191 loc) · 8.49 KB
/
vis.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
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
import cnn_vis
import deepdream
import keras.backend as K
from keras.applications import *
from keras.preprocessing import image
from keras.models import load_model
import numpy as np
class ConvNet_Vis():
#constructor
def __init__(self, image_path, model=None, name='custom_convnet'):
if model == None:
self.select_model()
self.model_dict = {
1: 'Xception',
2: 'VGG16',
3: 'VGG19',
4: 'ResNet50',
5: 'InceptionV3',
6: 'InceptionResNetV2',
7: 'MobileNet',
8: 'MobileNetV2',
9: 'DenseNet121',
10: 'DenseNet169',
11: 'DenseNet201',
12: 'NASNetMobile',
13: 'NASNetLarge'
}
self.set_model()
self.name = self.model_dict[self.model_no]
else:
if type(model) == str:
if os.path.exists(model):
self.model = load_model(model)
else:
raise ValueError("Model path doesn't exist.")
else:
self.model = model
self.name = name
# choose visulization method
self.selcet_vis()
self.image_path = image_path
self.load_image()
self.generate_vis()
def select_model(self):
try:
print('Choose the number for ConvNet architecture from following list: ')
print('1. Xception')
print('2. VGG16')
print('3. VGG19')
print('4. ResNet50')
print('5. InceptionV3')
print('6. InceptionResNetV2')
print('7. MobileNet')
print('8. MobileNetV2')
print('9. DenseNet121')
print('10. DenseNet169')
print('11. DenseNet201')
print('12. NASNetMobile')
print('13. NASNetLarge')
self.model_no = int(input())
except:
print('Choose right number :)')
self.select_model()
def set_model(self):
if self.model_dict[self.model_no] == 'Xception':
K.clear_session()
self.model = xception.Xception(include_top=True, weights='imagenet', input_tensor=None, input_shape=None, pooling=None, classes=1000)
from keras.applications.xception import preprocess_input, decode_predictions
self.preprocess_fun = preprocess_input
self.decode_fun = decode_predictions
self.target_size = (299, 299)
if self.model_dict[self.model_no] == 'VGG16':
K.clear_session()
self.model = vgg16.VGG16(include_top=True, weights='imagenet', input_tensor=None, input_shape=None, pooling=None, classes=1000)
from keras.applications.vgg16 import preprocess_input, decode_predictions
self.preprocess_fun = preprocess_input
self.decode_fun = decode_predictions
self.target_size = (224, 224)
if self.model_dict[self.model_no] == 'VGG19':
K.clear_session()
self.model = vgg19.VGG19(include_top=True, weights='imagenet', input_tensor=None, input_shape=None, pooling=None, classes=1000)
from keras.applications.vgg19 import preprocess_input, decode_predictions
self.preprocess_fun = preprocess_input
self.decode_fun = decode_predictions
self.target_size = (224, 224)
if self.model_dict[self.model_no] == 'ResNet50':
K.clear_session()
self.model = resnet50.ResNet50(include_top=True, weights='imagenet', input_tensor=None, input_shape=None, pooling=None, classes=1000)
from keras.applications.resnet50 import preprocess_input, decode_predictions
self.preprocess_fun = preprocess_input
self.decode_fun = decode_predictions
self.target_size = (224, 224)
if self.model_dict[self.model_no] == 'InceptionV3':
K.clear_session()
self.model = inception_v3.InceptionV3(include_top=True, weights='imagenet', input_tensor=None, input_shape=None, pooling=None, classes=1000)
from keras.applications.inception_v3 import preprocess_input, decode_predictions
self.preprocess_fun = preprocess_input
self.decode_fun = decode_predictions
self.target_size = (299, 299)
if self.model_dict[self.model_no] == 'InceptionResNetV2':
K.clear_session()
self.model = inception_resnet_v2.InceptionResNetV2(include_top=True, weights='imagenet', input_tensor=None, input_shape=None, pooling=None, classes=1000)
from keras.applications.inception_resnet_v2 import preprocess_input, decode_predictions
self.preprocess_fun = preprocess_input
self.decode_fun = decode_predictions
self.target_size = (299, 299)
if self.model_dict[self.model_no] == 'MobileNet':
K.clear_session()
self.model = mobilenet.MobileNet(input_shape=None, alpha=1.0, depth_multiplier=1, dropout=1e-3, include_top=True, weights='imagenet', input_tensor=None, pooling=None, classes=1000)
from keras.applications.mobilenet import preprocess_input, decode_predictions
self.preprocess_fun = preprocess_input
self.decode_fun = decode_predictions
self.target_size = (224, 224)
if self.model_dict[self.model_no] == 'MobileNetV2':
K.clear_session()
self.model = mobilenet_v2.MobileNetV2(input_shape=None, alpha=1.0, depth_multiplier=1, include_top=True, weights='imagenet', input_tensor=None, pooling=None, classes=1000)
from keras.applications.mobilenet_v2 import preprocess_input, decode_predictions
self.preprocess_fun = preprocess_input
self.decode_fun = decode_predictions
self.target_size = (224, 224)
if self.model_dict[self.model_no] == 'DenseNet121':
K.clear_session()
self.model = densenet.DenseNet121(include_top=True, weights='imagenet', input_tensor=None, input_shape=None, pooling=None, classes=1000)
from keras.applications.densenet import preprocess_input, decode_predictions
self.preprocess_fun = preprocess_input
self.decode_fun = decode_predictions
self.target_size = (224, 224)
if self.model_dict[self.model_no] == 'DenseNet169':
K.clear_session()
self.model = densenet.DenseNet169(include_top=True, weights='imagenet', input_tensor=None, input_shape=None, pooling=None, classes=1000)
from keras.applications.densenet import preprocess_input, decode_predictions
self.preprocess_fun = preprocess_input
self.decode_fun = decode_predictions
self.target_size = (224, 224)
if self.model_dict[self.model_no] == 'DenseNet201':
K.clear_session()
self.model = densenet.DenseNet201(include_top=True, weights='imagenet', input_tensor=None, input_shape=None, pooling=None, classes=1000)
from keras.applications.densenet import preprocess_input, decode_predictions
self.preprocess_fun = preprocess_input
self.decode_fun = decode_predictions
self.target_size = (224, 224)
if self.model_dict[self.model_no] == 'NASNetMobile':
K.clear_session()
self.model = nasnet.NASNetMobile(input_shape=None, include_top=True, weights='imagenet', input_tensor=None, pooling=None, classes=1000)
from keras.applications.nasnet import preprocess_input, decode_predictions
self.preprocess_fun = preprocess_input
self.decode_fun = decode_predictions
self.target_size = (224, 224)
if self.model_dict[self.model_no] == 'NASNetLarge':
K.clear_session()
self.model = nasnet.NASNetLarge(input_shape=None, include_top=True, weights='imagenet', input_tensor=None, pooling=None, classes=1000)
from keras.applications.nasnet import preprocess_input, decode_predictions
self.preprocess_fun = preprocess_input
self.decode_fun = decode_predictions
self.target_size = (331, 331)
def selcet_vis(self):
try:
print('Choose the number for Visulization method from following list: ')
print('1. Activation Visualization')
print('2. Deconvolution Visualization')
print('3. Deep Dream Visualization')
self.vis_method_no = int(input())
except:
print('Choose right number :)')
self.selcet_vis()
def load_image(self):
img = image.load_img(self.image_path, target_size=self.target_size)
x = image.img_to_array(img)
x = self.preprocess_fun(x)
x = np.expand_dims(x, axis=0)
self.loaded_image = x
def generate_vis(self):
sess = K.get_session()
layers = ['r', 'p', 'c']
with sess.as_default():
if self.vis_method_no == 1:
is_success = cnn_vis.activation_visualization(sess_graph_path = None,
value_feed_dict = {self.model.get_layer('input_1').input : self.loaded_image},
layers=layers, path_logdir='./vis/activation/' + self.model_dict[self.model_no] + '/log/',
path_outdir='./vis/activation/' + self.model_dict[self.model_no] + '/output/')
if self.vis_method_no == 2:
is_success = cnn_vis.deconv_visualization(sess_graph_path = None,
value_feed_dict = {self.model.get_layer('input_1').input : self.loaded_image},
layers=layers, path_logdir='./vis/deconv/' + self.model_dict[self.model_no] + '/log/',
path_outdir='./vis/deconv/' + self.model_dict[self.model_no] + '/output/')
if self.vis_method_no == 3:
deepdream.deep_dream(self.model, self.name, self.image_path, self.target_size, self.preprocess_fun, result_prefix = './vis/deepdream/' + self.name + '/output/')
def main():
ConvNet_Vis('./images/cat.jpg')
if __name__ == "__main__":
main()