阅读982 返回首页    go 阿里云 go 技术社区[云栖]


Kaggle首席技术官发布——(Kaggle)NIPS 2017对抗学习挑战赛起步指南

首发地址:https://yq.aliyun.com/articles/126898


https://yq.aliyun.com/cloud

NIPS,是关于机器学习领域的顶级会议,也是令众多学者振奋的学术盛会。12 年底举办的NIPS将新增一个议程,NIPS 2017Competition Track,从23个候选提案中选择了五个数据驱动的比赛项目。近日谷歌大脑研究员Ian Goodfellow在社媒平台中强烈推荐了由他组织的Adversarial Attacks and Defences(对抗攻击防御)比赛。为什么组织这样一个比赛呢,这是因为当前图像分类器非常容易被精心设计的对抗图像所欺骗,这些图像给原始图像及正确分类图像添加了微小变化,这些图像几乎不容易被人眼察觉,但会导致图像分类器错误地对错误的分类充满自信。

这项比赛是在kaggle

Kaggle

|

Kaggle1

下面具体来说下这三个相关比赛的侧重点:

  • 1Non Targeted Adversarial Attack,竞赛者所提交的系统需要对任何类别的图像进行处理,使得某个通用机器学习分类器无法识别。 
  • 2Targeted Adversarial Attack,竞赛者所提交的系统需要对给定的图像进行处理,使得某个通用机器学习分类器无法识别。 
  • 3Defense Against Adversarial Attack,竞赛者需要构建一个机器学习分类器,拥有足够的鲁棒性使之能正确鉴别对抗性的图像。 

根据前两个挑战的对抗性攻击,防御挑战的得分取决于分类器的好坏,另外前两个挑战的得分是基于在第三个挑战中对抗性攻击的伎俩有多好。

21a34e36727ac2ad9ff700c60b8ff65587a6a03d

 下面,我们将通过一些代码示例来生成非目标和目标的对抗图像,然后看看Inception V3googleNet Inception V3

注:下面许多的代码是基于Alextensorflow

/

import os
from cleverhans.attacks import FastGradientMethod
from io import BytesIO
import IPython.display
import numpy as np
import pandas as pd
from PIL import Image
from scipy.misc import imread
from scipy.misc import imsave
import tensorflow as tf
from tensorflow.contrib.slim.nets import inception

slim = tf.contrib.slim
tensorflow_master = ""
checkpoint_path   = "../input/inception-v3/inception_v3.ckpt"
input_dir         = "../input/nips-2017-adversarial-learning-development-set/images/"
max_epsilon       = 16.0
image_width       = 299
image_height      = 299
batch_size        = 16

eps = 2.0 * max_epsilon / 255.0
batch_shape = [batch_size, image_height, image_width, 3]
num_classes = 1001

def load_images(input_dir, batch_shape):
    images = np.zeros(batch_shape)
    filenames = []
    idx = 0
    batch_size = batch_shape[0]
    for filepath in sorted(tf.gfile.Glob(os.path.join(input_dir, '*.png'))):
        with tf.gfile.Open(filepath, "rb") as f:
            images[idx, :, :, :] = imread(f, mode='RGB').astype(np.float)*2.0/255.0 - 1.0
        filenames.append(os.path.basename(filepath))
        idx += 1
        if idx == batch_size:
            yield filenames, images
            filenames = []
            images = np.zeros(batch_shape)
            idx = 0
    if idx > 0:
        yield filenames, images

def show_image(a, fmt='png'):
    a = np.uint8((a+1.0)/2.0*255.0)
    f = BytesIO()
    Image.fromarray(a).save(f, fmt)
    IPython.display.display(IPython.display.Image(data=f.getvalue()))

class InceptionModel(object):
    def __init__(self, num_classes):
        self.num_classes = num_classes
        self.built = False

    def __call__(self, x_input):
        """Constructs model and return probabilities for given input."""
        reuse = True if self.built else None
        with slim.arg_scope(inception.inception_v3_arg_scope()):
            _, end_points = inception.inception_v3(
                            x_input, num_classes=self.num_classes, is_training=False,
                            reuse=reuse)
        self.built = True
        output = end_points['Predictions']
        probs = output.op.inputs[0]
        return probs

接下来,我们将在元数据中加载一批图像。

categories = pd.read_csv("../input/nips-2017-adversarial-learning-development-set/categories.csv")
image_classes = pd.read_csv("../input/nips-2017-adversarial-learning-development-set/images.csv")
image_iterator = load_images(input_dir, batch_shape)

# get first batch of images
filenames, images = next(image_iterator)

image_metadata = pd.DataFrame({"ImageId": [f[:-4] for f in filenames]}).merge(image_classes,
                                                                              on="ImageId")
true_classes = image_metadata["TrueLabel"].tolist()
target_classes = true_labels = image_metadata["TargetClass"].tolist()
true_classes_names = (pd.DataFrame({"CategoryId": true_classes})
                        .merge(categories, on="CategoryId")["CategoryName"].tolist())
target_classes_names = (pd.DataFrame({"CategoryId": target_classes})
                          .merge(categories, on="CategoryId")["CategoryName"].tolist())

print("Here's an example of one of the images in the development set")
show_image(images[0])

下面是开发集中的一个图像示例,是不是很可爱?

6e24caf7c3f6ccf673acdf87815f21b4dc218536

tensorflow的类别。

tf.logging.set_verbosity(tf.logging.INFO)

with tf.Graph().as_default():
    x_input = tf.placeholder(tf.float32, shape=batch_shape)
    model = InceptionModel(num_classes)

    fgsm  = FastGradientMethod(model)
    x_adv = fgsm.generate(x_input, eps=eps, clip_min=-1., clip_max=1.)

    saver = tf.train.Saver(slim.get_model_variables())
    session_creator = tf.train.ChiefSessionCreator(
                      scaffold=tf.train.Scaffold(saver=saver),
                      checkpoint_filename_with_path=checkpoint_path,
                      master=tensorflow_master)

    with tf.train.MonitoredSession(session_creator=session_creator) as sess:
        nontargeted_images = sess.run(x_adv, feed_dict={x_input: images})

print("The original image is on the left, and the nontargeted adversarial image is on the right. They look very similar, don't they? It's very clear both are gondolas")
show_image(np.concatenate([images[1], nontargeted_images[1]], axis=1))

INFO:tensorflow:Restoring parameters from ../input/inception-v3/inception_v3.ckpt

左边是原始图像,右边是非目标对抗图像。它们看起来很相似,很明显都是小船。

d983a119799fccd4d47358a61d26aeb5957a1ecb

tensorflow

注意:目前不工作,只是产生对抗图像而没有正确的目标。

all_images_target_class = {image_metadata["ImageId"][i]+".png": image_metadata["TargetClass"][i]
                           for i in image_metadata.index}

with tf.Graph().as_default():
    x_input = tf.placeholder(tf.float32, shape=batch_shape)

    with slim.arg_scope(inception.inception_v3_arg_scope()):
        logits, end_points = inception.inception_v3(
            x_input, num_classes=num_classes, is_training=False)

    target_class_input = tf.placeholder(tf.int32, shape=[batch_size])
    one_hot_target_class = tf.one_hot(target_class_input, num_classes)
    cross_entropy = tf.losses.softmax_cross_entropy(one_hot_target_class,
                                                    logits,
                                                    label_smoothing=0.1,
                                                    weights=1.0)
    cross_entropy += tf.losses.softmax_cross_entropy(one_hot_target_class,
                                                     end_points['AuxLogits'],
                                                     label_smoothing=0.1,
                                                     weights=0.4)
    x_adv = x_input - eps * tf.sign(tf.gradients(cross_entropy, x_input)[0])
    x_adv = tf.clip_by_value(x_adv, -1.0, 1.0)

    saver = tf.train.Saver(slim.get_model_variables())
    session_creator = tf.train.ChiefSessionCreator(
        scaffold=tf.train.Scaffold(saver=saver),
        checkpoint_filename_with_path=checkpoint_path,
        master=tensorflow_master)

    with tf.train.MonitoredSession(session_creator=session_creator) as sess:
        target_class_for_batch = ([all_images_target_class[n] for n in filenames]
                                  + [0] * (batch_size - len(filenames)))
        targeted_images = sess.run(x_adv,
                                   feed_dict={x_input: images,
                                              target_class_input: target_class_for_batch})
        
print("The original image is on the left, and the targeted adversarial image is on the right. Again, they look very similar, don't they? It's very clear both are butterflies")
show_image(np.concatenate([images[2], targeted_images[2]], axis=1))

INFO:tensorflow:Restoring parameters from ../input/inception-v3/inception_v3.ckpt

左边是原始图像,右边是目标对抗图像。同样可以发现它们看起来很相似,很明显都是蝴蝶。

ae81b393060b2a5ae4a02ea6021ba2f9ca1202f2

接下来,我们将看到,当把这些生成的对抗图像送入原始分类器运行时会发生什么

with tf.Graph().as_default():
    x_input = tf.placeholder(tf.float32, shape=batch_shape)

    with slim.arg_scope(inception.inception_v3_arg_scope()):
        _, end_points = inception.inception_v3(x_input, num_classes=num_classes, is_training=False)
    
    predicted_labels = tf.argmax(end_points['Predictions'], 1)

    saver = tf.train.Saver(slim.get_model_variables())
    session_creator = tf.train.ChiefSessionCreator(
                      scaffold=tf.train.Scaffold(saver=saver),
                      checkpoint_filename_with_path=checkpoint_path,
                      master=tensorflow_master)

    with tf.train.MonitoredSession(session_creator=session_creator) as sess:
        predicted_classes = sess.run(predicted_labels, feed_dict={x_input: images})
        predicted_nontargeted_classes = sess.run(predicted_labels, feed_dict={x_input: nontargeted_images})
        predicted_targeted_classes = sess.run(predicted_labels, feed_dict={x_input: targeted_images})

predicted_classes_names = (pd.DataFrame({"CategoryId": predicted_classes})
                           .merge(categories, on="CategoryId")["CategoryName"].tolist())

predicted_nontargeted_classes_names = (pd.DataFrame({"CategoryId": predicted_nontargeted_classes})
                          .merge(categories, on="CategoryId")["CategoryName"].tolist())

predicted_targeted_classes_names = (pd.DataFrame({"CategoryId": predicted_targeted_classes})
                          .merge(categories, on="CategoryId")["CategoryName"].tolist())

INFO:tensorflow:Restoring parameters from ../input/inception-v3/inception_v3.ckpt

下面我们将展示这个批次中的所有图像以及它们的分类的类别。每个集合中的左图是原始图像中间是非目标对抗形象,右图目标对抗图像

for i in range(len(images)):
    print("UNMODIFIED IMAGE (left)",
          "\n\tPredicted class:", predicted_classes_names[i],
          "\n\tTrue class:     ", true_classes_names[i])
    print("NONTARGETED ADVERSARIAL IMAGE (center)",
          "\n\tPredicted class:", predicted_nontargeted_classes_names[i])
    print("TARGETED ADVERSARIAL IMAGE (right)",
          "\n\tPredicted class:", predicted_targeted_classes_names[i],
          "\n\tTarget class:   ", target_classes_names[i])
    show_image(np.concatenate([images[i], nontargeted_images[i], targeted_images[i]], axis=1))

原图(左图)

预测类别:大熊猫

真实类别:大熊猫

非目标对抗图片(中间图)

预测类别:萨摩耶犬类

目标对抗图片(右图)

预测类别:土狗

真实类别:肉饼

eb92800602fbd72a7c747a2c569444094ca99f60

原图(左图)

预测类别:

真实类别:

非目标对抗图片(中间图)

   预测类别:堤坝

目标对抗图片(右图)

预测类别:堤坝

真实类别:

5e571645baa26b5deca4bed6500cad8630a366b5

原图(左图)

预测类别:灰蝶

真实类别:灰蝶

非目标对抗图片(中间图)

   预测类别:小环蝴蝶

目标对抗图片(右图)

预测类别:小环蝴蝶

真实类别:西班牙可卡犬

b79d9aa23bb377733ff8998d5c333bc7857f52c6

原图(左图)

预测类别:灰蝶

真实类别:灰蝶

非目标对抗图片(中间图)

预测类别:

目标对抗图片(右图)

预测类别:

真实类别:

add1a3bbd9d559f58915acbf43f56c24442d0978

原图(左图)

预测类别:美洲黑鸭

真实类别:美洲黑鸭

非目标对抗图片(中间图)

测类别:短尾鹦鹉

目标对抗图片(右图)

预测类别:短尾鹦鹉

真实类别:泉水

6bacc1251fca1bfb113c5f7101823bdab44ad176

原图(左图)

预测类别:短尾鹦鹉

真实类别:短尾鹦鹉

非目标对抗图片(中间图)

测类别:篮球

目标对抗图片(右图)

预测类别:篮球

真实类别:单峰骆驼

4f2a60da42e587d07d38b095ea7c9d5e00ecfc57

原图(左图)

预测类别:球员

真实类别:球员

非目标对抗图片(中间图)

测类别:鸵鸟

目标对抗图片(右图)

预测类别:鸵鸟

真实类别:金库

a29b1b39b10756a621ba2789f792721602c05fce

原图(左图)

预测类别:鸵鸟

真实类别:鸵鸟

非目标对抗图片(中间图)

测类别:日晷

目标对抗图片(右图)

预测类别:日晷

真实类别:海上钻井平台

e8b79ea0ca3198133822051dfd075d87b609180f

原图(左图)

预测类别:加农炮

真实类别:加农炮

非目标对抗图片(中间图)

测类别:虎甲虫

目标对抗图片(右图)

预测类别:虎甲虫

真实类别:特浓咖啡机

729a1a61ac901d8dc85277574645c78b0ea2caff

原图(左图)

预测类别:长角天牛

真实类别:长角天牛

非目标对抗图片(中间图)

测类别:特浓咖啡

目标对抗图片(右图)

预测类别:特浓咖啡

真实类别:髓内钉

eac6a118765788b3386fe7e6b67a4d333cfb3bfa

原图(左图)

预测类别:特浓咖啡

真实类别:特浓咖啡

非目标对抗图片(中间图)

测类别:淋浴帽

目标对抗图片(右图)

预测类别:淋浴帽

真实类别:

e733599f936c88b5a4fd0f20f24dae05cdba20c4

原图(左图)

预测类别:雪橇

真实类别:雪橇

非目标对抗图片(中间图)

测类别:玫瑰果

目标对抗图片(右图)

预测类别:蜜蜂

真实类别:

a211e66285f2cfb282119cba53a3a3111efbe656

原图(左图)

预测类别:大钢琴

真实类别:大钢琴

非目标对抗图片(中间图)

测类别:书桌

目标对抗图片(右图)

预测类别:餐桌

真实类别:

918452d7ad531e90c26a1633e764229be314b561

原图(左图)

预测类别:间歇喷泉

真实类别:间歇喷泉

非目标对抗图片(中间图)

测类别:沉船

目标对抗图片(右图)

预测类别:海狸

真实类别:

43f7b4cccf1b1b4fdde3e337d8a33b34d1addb23

原图(左图)

预测类别:图书馆

真实类别:图书馆

非目标对抗图片(中间图)

测类别:书店

目标对抗图片(右图)

预测类别:书店

真实类别:安全别针

baf38a228b7c8335232dd22af76aa54106c751ab

原图(左图)

预测类别:松鸭

真实类别:松鸭

非目标对抗图片(中间图)

测类别:松鸭

目标对抗图片(右图)

预测类别:黄雀

真实类别:针鼹

901ac96bde922ad713c80a3fdb6b57ffca7a80d1

Ben HamnerKaggle

6fc3a6a48252369b55117e13167a93a47ff90fe4

Linkedinhttps://www.linkedin.com/in/ben-hamner-98759712/

Github: https://github.com/benhamner

本文由北邮@老师推荐,阿里云云栖社区组织翻译。

Getting Started with the NIPS 2017 Adversarial Learning ChallengesBen Hamner

文章为简译,更为详细的内容,请查看

最后更新:2017-07-27 09:03:40

  上一篇:go  简单入门循环神经网络RNN:时间序列数据的首选神经网络
  下一篇:go  5个步骤 & 7个提示 | 一份开启Kaggle竞赛征途的初学者指南