-
Notifications
You must be signed in to change notification settings - Fork 9
/
evaluate_uap.py
283 lines (213 loc) · 8.32 KB
/
evaluate_uap.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
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
'''Code to evaluate the UAP for any target CNN.
Almost similar to the code to train generator i.e. train_generator.py
except the optimization step; this code does only the forward pass
'''
from misc.layers import *
from classify import *
from new_generator import *
import tensorflow as tf
import numpy as np
import argparse
import h5py
import math
import random
import sys
import copy
from skimage.io import imread
from skimage.transform import resize
import time
import os
from random import shuffle
from math import ceil
from scipy import misc
from itertools import combinations
#config = tf.ConfigProto(gpu_options=tf.GPUOptions(per_process_gpu_memory_fraction=0.5))
config = tf.ConfigProto(gpu_options=tf.GPUOptions(allow_growth=True))
size=227
batch_size = 64
test_path = 'ilsvrc_valid.hdf5'
test_file = h5py.File(test_path, "r")
test_num = test_file["valid_img"].shape[0]
print "{} {}".format("Should be 1000",test_num)
def resize_caffenet(inp):
for i in xrange(inp.shape[0]):
img = resize(np.uint8(inp[i]) , (227,227))*255.0
img = img.reshape((1,227,227,3))
if i==0:
out = img
else:
out = np.vstack([out,img])
return out
def add_perturbation(inp, perturbation):
inp = inp + perturbation
return inp
def log_loss(prob_vec, adv_prob_vec, top_prob):
size = prob_vec.get_shape().as_list()[0]
for i in xrange(size):
if i==0:
loss= adv_prob_vec[i][top_prob[i][0]]
else:
loss = loss + adv_prob_vec[i][top_prob[i][0]]
mean = (loss/size)
gen_loss = -tf.log(1-mean)
return gen_loss,mean
def cosine_loss(prob_vec, adv_prob_vec, top_prob):
size = prob_vec.get_shape().as_list()[0]
for i in xrange(size):
a1 = tf.square(prob_vec[i])
sum1 = tf.reduce_sum(a1)
sqrt = tf.sqrt(sum1)
unitv1 = tf.divide(a1,sqrt)
a1 = tf.square(adv_prob_vec[i])
sum1 = tf.reduce_sum(a1)
sqrt = tf.sqrt(sum1)
unitv2 = tf.divide(a1,sqrt)
mul = tf.multiply(unitv1,unitv2)
error = tf.reduce_sum(mul)
if i==0:
loss= error#v_adv[i][topk[i][0]]
else:
loss = loss + error#v_adv[i][topk[i][0]]
cosine_loss = (loss/size)
return cosine_loss
def validation_results(prob_adv,prob_real):
nfool=0
size = prob_real.shape[0]
for i in xrange(size):
if prob_real[i]!=prob_adv[i]:
nfool = nfool+1
return nfool, 100*float(nfool)/size
def tensor_read(sess,img_list):
filename_queue = tf.train.string_input_producer(img_list,num_epochs=2,shuffle=True)
image_reader = tf.WholeFileReader()
_, image_file = image_reader.read(filename_queue)
image = tf.image.decode_png(image_file)
image = tf.image.resize_images(image, [224, 224])
image.set_shape((224, 224, 3))
batch_size = 64
num_preprocess_threads = 1
min_queue_examples = 0
image = tf.train.shuffle_batch(
[image],
batch_size=batch_size,
num_threads=num_preprocess_threads,
capacity=min_queue_examples + 3 * batch_size,
min_after_dequeue=min_queue_examples)
# Required to get the filename matching to run.
sess.run(tf.local_variables_initializer())
sess.run(tf.global_variables_initializer())
# Coordinate the loading of image files.
coord = tf.train.Coordinator()
threads = tf.train.start_queue_runners(coord=coord)
#return sess.run(image)
image_tensor = sess.run(image)
coord.request_stop()
coord.join(threads)
return image_tensor
def read_images_as_array(img_list):
for i,name in enumerate(img_list):
img = imread(name)
img = resize(img,(size,size))*255.0
img = img.reshape((1,size,size,3))
# print i,name,np.mean(img)
if i==0:
cln_img = img
#adv_img = adv_im
else:
cln_img = np.vstack([cln_img,img])
#adv_img = np.vstack(adv_img,adv_im)
return cln_img#, adv_img
def tensor_preprocess(tensor, size=size):
mean = [103.939, 116.779, 123.68]
a = tensor[:,:,0]
a -= mean[2]
a = tf.reshape(a,shape=(size,size,1))
b = tensor[:,:,1]
b -= mean[1]
b = tf.reshape(b,shape=(size,size,1))
c = tensor[:,:,2]
c -= mean[0]
c = tf.reshape(c,shape=(size,size,1))
tensor = tf.concat([c,b,a],axis=2)
#tensor[:,:,[0,1,2]] = tensor[:,:,[2,1,0]]
return tensor
def whole_process(tensor):
l = tensor.get_shape().as_list()[0]
for i in xrange(l):
if i==0:
temp = tensor_preprocess(tensor[i])
temp = tf.reshape(temp, shape = [1,size,size,3])
else:
temp = tf.concat([temp, tf.reshape(tensor_preprocess(tensor[i]),shape=[1,size,size,3])],axis=0)
return temp
sess = tf.Session(config=config)
#sess2 = tf.Session(config=config)
img_dir = '/data1/nithish/WORK/Resources/ilsvrc_train/'
curr_dir = os.getcwd()
with sess.as_default():
G = Generator()
w = G.generate()
w = tf.pad(w , [[0,0],[1,2],[1,2],[0,0]] , mode='REFLECT')
saver = tf.train.Saver(tf.trainable_variables())
img_list = open('text_files/new_train.txt').readlines()
for i in xrange(len(img_list)):
img_list[i] = img_list[i].split()[0]
random.shuffle(img_list)
valid_list_og = open('text_files/new_valid.txt').readlines()
for i in xrange(len(valid_list_og)):
valid_list_og[i] = valid_list_og[i].split()[0]
valid_list = open('text_files/full_valid.txt').readlines()
for i in xrange(len(valid_list)):
valid_list[i] = valid_list[i].split()[0]
batch_data = tf.placeholder(dtype=tf.float32, shape=(batch_size,size,size,3), name='clean_train_data')
val_data = tf.placeholder(dtype=tf.float32, shape=(batch_size,size,size,3), name='clean_validation_data')
class_hist = np.zeros([1000], dtype = np.int32)
random_adv_batch = add_perturbation(batch_data,w[0])
adv_valid_data = add_perturbation(val_data,w[0])
v,topk = scores(whole_process(batch_data))
v_adv, topk_adv = scores(whole_process(random_adv_batch))
v = v['prob']
v_adv = v_adv['prob']
q1_loss,mean_q1 = log_loss(v,v_adv,topk)
cos_loss = cosine_loss(v,v_adv,topk)
q1_loss = q1_loss
cos_loss = cos_loss
lr = 1e-3
train_op = tf.train.AdamOptimizer(lr).minimize(q1_loss)
epoch=0
total_epochs = 60
iterations = len(img_list) / batch_size
remaining = len(img_list) % batch_size
val_iterations = 1000/batch_size
sess.run(tf.global_variables_initializer())
G_list = ['cosine10','res4f_10','res5c_10']
for i in G_list:
model = 'variations/resnet50_'+i+'.00280.ckpt'
saver.restore(sess,model)
print "{} {}".format(i,"Loaded")
for j in xrange(100):
z = np.random.uniform(low=-1. , high=1. , size = (G.batch_size,G.z_dim))
print "{} {}".format(j+1,"z sampled")
if True:
total_fool=0
print "{}".format("############### VALIDATION PHASE STARTED ################")
for y in xrange(val_iterations):
i_s = y * batch_size # index of the first image in this batch
i_e = min([(y + 1) * batch_size, test_num])
train_batch = test_file["test_img"][i_s:i_e, ...]
train_batch = resize_caffenet(train_batch)
cln_top, adv_top = sess.run([topk, topk_adv],feed_dict={
G.zn:z,#np.random.uniform(low=-1. , high=1. , size = (G.batch_size,G.z_dim)),
batch_data:train_batch} )
adv_top = np.reshape(adv_top , (adv_top.shape[0]))
for i in xrange(adv_top.shape[0]):
class_hist[adv_top[i]] = class_hist[adv_top[i]] + 1
nfool,foolr = validation_results(cln_top,adv_top)
total_fool = total_fool + nfool
print "{} {} {} {} {} {} {} {} {}".format(y+1,"out of",val_iterations,"batches done and number of fools for this batch is",nfool,"/",batch_size,"and fooling rate",100*float(total_fool)/((y+1)*batch_size))
foolr = 100*float(total_fool)/(val_iterations*batch_size)
print "{} {} {}".format("Fooling rate",foolr,total_fool)
np.save('cvpr_final/ablation_exp/'+i+'/'+str(j+1)+'.npy',class_hist)
print "{}".format("############### VALIDATION PHASE ENDED ################")
epoch = epoch + 1
print "Ablation experiment Done!"