def evaluate(net_file, model_file):
""" main
"""
#1, build model
net_path = os.path.dirname(net_file)
if net_path not in sys.path:
sys.path.insert(0, net_path)
from lenet import LeNet as MyNet
#1, define network topology
images = fluid.layers.data(name='image',
shape=[1, 28, 28],
dtype='float32')
label = fluid.layers.data(name='label', shape=[1], dtype='int64')
net = MyNet({'data': images})
prediction = net.layers['prob']
acc = fluid.layers.accuracy(input=prediction, label=label)
place = fluid.CPUPlace()
exe = fluid.Executor(place)
exe.run(fluid.default_startup_program())
#2, load weights
if model_file.find('.npy') > 0:
net.load(data_path=model_file, exe=exe, place=place)
else:
net.load(data_path=model_file, exe=exe)
#3, test this model
test_program = fluid.default_main_program().clone()
test_reader = paddle.batch(paddle.dataset.mnist.test(), batch_size=128)
feeder = fluid.DataFeeder(feed_list=[images, label], place=place)
fetch_list = [acc, prediction]
print('go to test model using test set')
acc_val = test_model(exe, test_program, \
fetch_list, test_reader, feeder)
print('test accuracy is [%.4f], expected value[0.919]' % (acc_val))
import os
import cv2
import numpy as np
from lenet import LeNet
data_dir = "mnist/test"
net = LeNet()
net.load("lenet.npy")
files = os.listdir(data_dir)
images = []
labels = []
for f in files:
img = cv2.imread(os.path.join(data_dir, f), cv2.IMREAD_GRAYSCALE)
img = cv2.resize(img, (32, 32))
img = img.astype(np.float32).reshape(32, 32, 1) / 255.0
images.append(img)
labels.append(int(f[0]))
x = np.array(images)
y = np.array(labels)
predict = net.predict(x)
tp = np.sum(predict == y)
accuracy = float(tp) / len(files)
print("accuracy=%f" % accuracy)
def main(use_cuda):
"""
Advbox demo which demonstrate how to use advbox.
"""
TOTAL_NUM = 500
IMG_NAME = 'image'
LABEL_NAME = 'label'
weight_file = "fluid/lenet/lenet.npy"
#1, define network topology
images = fluid.layers.data(name=IMG_NAME,
shape=[1, 28, 28],
dtype='float32')
# gradient should flow
images.stop_gradient = False
label = fluid.layers.data(name=LABEL_NAME, shape=[1], dtype='int64')
net = LeNet({'data': images})
prediction = net.layers['prob']
cost = fluid.layers.cross_entropy(input=prediction, label=label)
avg_cost = fluid.layers.mean(x=cost)
#根据配置选择使用CPU资源还是GPU资源
place = fluid.CUDAPlace(0) if use_cuda else fluid.CPUPlace()
exe = fluid.Executor(place)
#这句很关键 没有的话会报错
# AttributeError: 'NoneType' object has no attribute 'get_tensor'
exe.run(fluid.default_startup_program())
#加载参数
net.load(data_path=weight_file, exe=exe, place=place)
BATCH_SIZE = 1
test_reader = paddle.batch(paddle.reader.shuffle(
paddle.dataset.mnist.test(), buf_size=128 * 10),
batch_size=BATCH_SIZE)
# advbox demo
m = PaddleModel(fluid.default_main_program(),
IMG_NAME,
LABEL_NAME,
prediction.name,
avg_cost.name, (-1, 1),
channel_axis=1)
attack = FGSM(m)
# attack = FGSMT(m)
attack_config = {"epsilons": 0.3}
# use test data to generate adversarial examples
total_count = 0
fooling_count = 0
for data in test_reader():
total_count += 1
adversary = Adversary(data[0][0], data[0][1])
# FGSM non-targeted attack
adversary = attack(adversary, **attack_config)
if adversary.is_successful():
fooling_count += 1
print(
'attack success, original_label=%d, adversarial_label=%d, count=%d'
% (data[0][1], adversary.adversarial_label, total_count))
else:
print('attack failed, original_label=%d, count=%d' %
(data[0][1], total_count))
if total_count >= TOTAL_NUM:
print(
"[TEST_DATASET]: fooling_count=%d, total_count=%d, fooling_rate=%f"
% (fooling_count, total_count,
float(fooling_count) / total_count))
break
print("fgsm attack done")
