Adversarial attacks aim to confound machine learning systems, while remaining virtually imperceptible to humans. Attacks on image classification systems are typically gauged in terms of $p$-norm distortions in the pixel feature space. We perform a behavioral study, demonstrating that the pixel $p$-norm for any $0\le p \le \infty$, and several alternative measures including earth mover's distance, structural similarity index, and deep net embedding, do not fit human perception. Our result has the potential to improve the understanding of adversarial attack and defense strategies.
Deep neural networks are vulnerable against adversarial examples. In this paper, we propose to train and test the networks with randomly subsampled images with high drop rates. We show that this approach significantly improves robustness against adversarial examples in all cases of bounded L0, L2 and L_inf perturbations, while reducing the standard accuracy by a small value. We argue that subsampling pixels can be thought to provide a set of robust features for the input image and, thus, improves robustness without performing adversarial training.
The output of Deep Neural Networks (DNN) can be altered by a small perturbation of the input in a black box setting by making multiple calls to the DNN. However, the high computation and time required makes the existing approaches unusable. This work seeks to improve the One-pixel (few-pixel) black-box adversarial attacks to reduce the number of calls to the network under attack. The One-pixel attack uses a non-gradient optimization algorithm to find pixel-level perturbations under the constraint of a fixed number of pixels, which causes the network to predict the wrong label for a given image…
Despite extensive research into adversarial attacks, we do not know how adversarial attacks affect image pixels. Knowing how image pixels are affected by adversarial attacks has the potential to lead us to better adversarial defenses. Motivated by instances that we find where strong attacks do not transfer, we delve into adversarial examples at pixel level to scrutinize how adversarial attacks affect image pixel values. We consider several ImageNet architectures, InceptionV3, VGG19 and ResNet50, as well as several strong attacks. We find that attacks can have different effects at pixel level d…
Bennett University（贝内特大学）；Indian Statistical Institute（印度统计学院）；Duke University（杜克大学）；Indian Institute of Engineering Science & Technology, Shibpur（印度工程技术学院）；Kyungpook National University（庆北国立大学）
Constructing adversarial perturbations for deep neural networks is an important direction of research. Crafting image-dependent adversarial perturbations using white-box feedback has hitherto been the norm for such adversarial attacks. However, black-box attacks are much more practical for real-world applications. Universal perturbations applicable across multiple images are gaining popularity due to their innate generalizability. There have also been efforts to restrict the perturbations to a few pixels in the image. This helps to retain visual similarity with the original images making such …
Google（谷歌）；University of Wisconsin-Madison（威斯康星大学麦迪逊分校）
Wide adoption of artificial neural networks in various domains has led to an increasing interest in defending adversarial attacks against them. Preprocessing defense methods such as pixel discretization are particularly attractive in practice due to their simplicity, low computational overhead, and applicability to various systems. It is observed that such methods work well on simple datasets like MNIST, but break on more complicated ones like ImageNet under recently proposed strong white-box attacks. To understand the conditions for success and potentials for improvement, we study the pixel d…
Deep neural networks (DNNs) have been proven to be vulnerable to adversarial examples. A special branch of adversarial examples, namely sparse adversarial examples, can fool the target DNNs by perturbing only a few pixels. However, many existing sparse adversarial attacks use heuristic methods to select the pixels to be perturbed, and regard the pixel selection and the adversarial attack as two separate steps. From the perspective of neural network pruning, we propose a novel end-to-end sparse adversarial attack method, namely AutoAdversary, which can find the most important pixels automatical…
Recently, techniques have been developed to provably guarantee the robustness of a classifier to adversarial perturbations of bounded L_1 and L_2 magnitudes by using randomized smoothing: the robust classification is a consensus of base classifications on randomly noised samples where the noise is additive. In this paper, we extend this technique to the L_0 threat model. We propose an efficient and certifiably robust defense against sparse adversarial attacks by randomly ablating input features, rather than using additive noise. Experimentally, on MNIST, we can certify the classifications of o…
Recent research has found that neural networks are vulnerable to several types of adversarial attacks, where the input samples are modified in such a way that the model produces a wrong prediction that misclassifies the adversarial sample. In this paper we focus on black-box adversarial attacks, that can be performed without knowing the inner structure of the attacked model, nor the training procedure, and we propose a novel attack that is capable of correctly attacking a high percentage of samples by rearranging a small number of pixels within the attacked image. We demonstrate that our attac…
Adversarial machine learning has been both a major concern and a hot topic recently, especially with the ubiquitous use of deep neural networks in the current landscape. Adversarial attacks and defenses are usually likened to a cat-and-mouse game in which defenders and attackers evolve over the time. On one hand, the goal is to develop strong and robust deep networks that are resistant to malicious actors. On the other hand, in order to achieve that, we need to devise even stronger adversarial attacks to challenge these defense models. Most of existing attacks employs a single $\ell_p$ distanc…
Monash University（莫纳什大学）；Defence Science and Technology Group, Department of Defence（国防部国防科学技术集团）
Deep neural network image classifiers are reported to be susceptible to adversarial evasion attacks, which use carefully crafted images created to mislead a classifier. Many adversarial attacks belong to the category of dense attacks, which generate adversarial examples by perturbing all the pixels of a natural image. To generate sparse perturbations, sparse attacks have been recently developed, which are usually independent attacks derived by modifying a dense attack's algorithm with sparsity regularisations, resulting in reduced attack efficiency. In this paper, we aim to tackle this task fr…
We propose a novel technique that can generate natural-looking adversarial examples by bounding the variations induced for internal activation values in some deep layer(s), through a distribution quantile bound and a polynomial barrier loss function. By bounding model internals instead of individual pixels, our attack admits perturbations closely coupled with the existing features of the original input, allowing the generated examples to be natural-looking while having diverse and often substantial pixel distances from the original input. Enforcing per-neuron distribution quantile bounds allow…
Conventional adversarial training methods using attacks that manipulate the pixel value directly and individually, leading to models that are less robust in face of spatial transformation-based attacks. In this paper, we propose a joint adversarial training method that incorporates both spatial transformation-based and pixel-value based attacks for improving model robustness. We introduce a spatial transformation-based attack with an explicit notion of budget and develop an algorithm for spatial attack generation. We further integrate both pixel and spatial attacks into one generation model an…
Neural networks have been proven to be vulnerable to a variety of adversarial attacks. From a safety perspective, highly sparse adversarial attacks are particularly dangerous. On the other hand the pixelwise perturbations of sparse attacks are typically large and thus can be potentially detected. We propose a new black-box technique to craft adversarial examples aiming at minimizing $l_0$-distance to the original image. Extensive experiments show that our attack is better or competitive to the state of the art. Moreover, we can integrate additional bounds on the componentwise perturbation. All…
Adversarial patch attacks mislead neural networks by injecting adversarial pixels within a local region. Patch attacks can be highly effective in a variety of tasks and physically realizable via attachment (e.g. a sticker) to the real-world objects. Despite the diversity in attack patterns, adversarial patches tend to be highly textured and different in appearance from natural images. We exploit this property and present PatchZero, a general defense pipeline against white-box adversarial patches without retraining the downstream classifier or detector. Specifically, our defense detects adversa…
Improving Accuracy-robustness Trade-off via Pixel Reweighted Adversarial Training
School of Computing and Information Systems, The University of Melbourne（墨尔本大学计算与信息系统学院）；State Key Laboratory of Integrated Services Networks, Xidian University（西安电子科技大学）；School of Computer Science, The University of Auckland（奥克兰大学计算机科学学院）；RIKEN AIP（理化学研究所人工智能项目）；Sydney AI Centre, The University of Sydney（悉尼大学悉尼人工智能中心）
East China Normal University（华东师范大学）；Nanyang Technological University（南洋理工大学）；Alibaba Group（阿里巴巴集团）；University of Alberta（阿尔伯塔大学）；Zhejiang Sci-Tech University（浙江理工大学）
High-level representation-guided pixel denoising and adversarial training are independent solutions to enhance the robustness of CNNs against adversarial attacks by pre-processing input data and re-training models, respectively. Most recently, adversarial training techniques have been widely studied and improved while the pixel denoising-based method is getting less attractive. However, it is still questionable whether there exists a more advanced pixel denoising-based method and whether the combination of the two solutions benefits each other. To this end, we first comprehensively investigate…
Kyushu University（九州大学）；Advanced Telecommunications Research Institute International（先进电信研究国际研究所）
Recent research has revealed that the output of Deep Neural Networks (DNN) can be easily altered by adding relatively small perturbations to the input vector. In this paper, we analyze an attack in an extremely limited scenario where only one pixel can be modified. For that we propose a novel method for generating one-pixel adversarial perturbations based on differential evolution (DE). It requires less adversarial information (a black-box attack) and can fool more types of networks due to the inherent features of DE. The results show that 67.97% of the natural images in Kaggle CIFAR-10 test d…
University of Toronto（多伦多大学）；Carnegie Mellon University（卡内基梅隆大学）；McGill University（麦吉尔大学）
Many machine learning image classifiers are vulnerable to adversarial attacks, inputs with perturbations designed to intentionally trigger misclassification. Current adversarial methods directly alter pixel colors and evaluate against pixel norm-balls: pixel perturbations smaller than a specified magnitude, according to a measurement norm. This evaluation, however, has limited practical utility since perturbations in the pixel space do not correspond to underlying real-world phenomena of image formation that lead to them and has no security motivation attached. Pixels in natural images are mea…
Stony Brook University（石溪大学）；University of Michigan（密歇根大学）
Deep Neural Networks (DNNs) are known to be susceptible to adversarial examples. Adversarial examples are maliciously crafted inputs that are designed to fool a model, but appear normal to human beings. Recent work has shown that pixel discretization can be used to make classifiers for MNIST highly robust to adversarial examples. However, pixel discretization fails to provide significant protection on more complex datasets. In this paper, we take the first step towards reconciling these contrary findings. Focusing on the observation that discrete pixelization in MNIST makes the background comp…
New York University（纽约大学）；Central South University（中南大学）；The University of North Carolina at Chapel Hill（北卡罗来纳大学教堂山分校）；East China Normal University（华东师范大学）
Deep neural networks (DNNs) have achieved great success in image classification, but can be very vulnerable to adversarial attacks with small perturbations to images. To improve adversarial image generation for DNNs, we develop a novel method, called mFI-PSO, which utilizes a Manifold-based First-order Influence measure for vulnerable image and pixel selection and the Particle Swarm Optimization for various objective functions. Our mFI-PSO can thus effectively design adversarial images with flexible, customized options on the number of perturbed pixels, the misclassification probability, and t…
Injecting adversarial examples during training, known as adversarial training, can improve robustness against one-step attacks, but not for unknown iterative attacks. To address this challenge, we first show iteratively generated adversarial images easily transfer between networks trained with the same strategy. Inspired by this observation, we propose cascade adversarial training, which transfers the knowledge of the end results of adversarial training. We train a network from scratch by injecting iteratively generated adversarial images crafted from already defended networks in addition to o…
Shanghai Jiao Tong University（上海交通大学）；University of California, Santa Cruz（加州大学圣克鲁兹分校）
Unlearnable examples (ULEs) aim to protect data from unauthorized usage for training DNNs. Existing work adds $\ell_\infty$-bounded perturbations to the original sample so that the trained model generalizes poorly. Such perturbations, however, are easy to eliminate by adversarial training and data augmentations. In this paper, we resolve this problem from a novel perspective by perturbing only one pixel in each image. Interestingly, such a small modification could effectively degrade model accuracy to almost an untrained counterpart. Moreover, our produced \emph{One-Pixel Shortcut (OPS)} could…
South China Agricultural University（华南农业大学）；Pazhou Lab（琶洲实验室）
Since the Lipschitz properties of convolutional neural networks (CNNs) are widely considered to be related to adversarial robustness, we theoretically characterize the $\ell_1$ norm and $\ell_\infty$ norm of 2D multi-channel convolutional layers and provide efficient methods to compute the exact $\ell_1$ norm and $\ell_\infty$ norm. Based on our theorem, we propose a novel regularization method termed norm decay, which can effectively reduce the norms of convolutional layers and fully-connected layers. Experiments show that norm-regularization methods, including norm decay, weight decay, and s…
Although deep learning has made remarkable progress in processing various types of data such as images, text and speech, they are known to be susceptible to adversarial perturbations: perturbations specifically designed and added to the input to make the target model produce erroneous output. Most of the existing studies on generating adversarial perturbations attempt to perturb the entire input indiscriminately. In this paper, we propose ExploreADV, a general and flexible adversarial attack system that is capable of modeling regional and imperceptible attacks, allowing users to explore variou…
We propose a new adversarial attack to Deep Neural Networks for image classification. Different from most existing attacks that directly perturb input pixels, our attack focuses on perturbing abstract features, more specifically, features that denote styles, including interpretable styles such as vivid colors and sharp outlines, and uninterpretable ones. It induces model misclassfication by injecting imperceptible style changes through an optimization procedure. We show that our attack can generate adversarial samples that are more natural-looking than the state-of-the-art unbounded attacks. T…
With the excellent accuracy and feasibility, the Neural Networks have been widely applied into the novel intelligent applications and systems. However, with the appearance of the Adversarial Attack, the NN based system performance becomes extremely vulnerable:the image classification results can be arbitrarily misled by the adversarial examples, which are crafted images with human unperceivable pixel-level perturbation. As this raised a significant system security issue, we implemented a series of investigations on the adversarial attack in this work: We first identify an image's pixel vulnera…
We establish a theoretical link between adversarial training and operator norm regularization for deep neural networks. Specifically, we prove that $\ell_p$-norm constrained projected gradient ascent based adversarial training with an $\ell_q$-norm loss on the logits of clean and perturbed inputs is equivalent to data-dependent (p, q) operator norm regularization. This fundamental connection confirms the long-standing argument that a network's sensitivity to adversarial examples is tied to its spectral properties and hints at novel ways to robustify and defend against adversarial attacks. We p…
Carnegie Mellon University（卡内基梅隆大学）；Bosch Center for Artificial Intelligence（博世人工智能中心）
A rapidly growing area of work has studied the existence of adversarial examples, datapoints which have been perturbed to fool a classifier, but the vast majority of these works have focused primarily on threat models defined by $\ell_p$ norm-bounded perturbations. In this paper, we propose a new threat model for adversarial attacks based on the Wasserstein distance. In the image classification setting, such distances measure the cost of moving pixel mass, which naturally cover "standard" image manipulations such as scaling, rotation, translation, and distortion (and can potentially be applied…
Drexel University（德雷塞尔大学）；Conestoga College（康尼斯托加学院）；New York University（纽约大学）
Numerous recent studies have demonstrated how Deep Neural Network (DNN) classifiers can be fooled by adversarial examples, in which an attacker adds perturbations to an original sample, causing the classifier to misclassify the sample. Adversarial attacks that render DNNs vulnerable in real life represent a serious threat in autonomous vehicles, malware filters, or biometric authentication systems. In this paper, we apply Fast Gradient Sign Method to introduce perturbations to a facial image dataset and then test the output on a different classifier that we trained ourselves, to analyze transf…
Max Planck Institute for Informatics（马克斯·普朗克信息学研究所）
Modern deep learning architecture utilize batch normalization (BN) to stabilize training and improve accuracy. It has been shown that the BN layers alone are surprisingly expressive. In the context of robustness against adversarial examples, however, BN is argued to increase vulnerability. That is, BN helps to learn fragile features. Nevertheless, BN is still used in adversarial training, which is the de-facto standard to learn robust features. In order to shed light on the role of BN in adversarial training, we investigate to what extent the expressiveness of BN can be used to robustify fragi…
Deep neural networks (DNNs) have set benchmarks on a wide array of supervised learning tasks. Trained DNNs, however, often lack robustness to minor adversarial perturbations to the input, which undermines their true practicality. Recent works have increased the robustness of DNNs by fitting networks using adversarially-perturbed training samples, but the improved performance can still be far below the performance seen in non-adversarial settings. A significant portion of this gap can be attributed to the decrease in generalization performance due to adversarial training. In this work, we exten…
Indian Institute of Technology, Hyderabad（印度理工学院海得拉巴分校）
Current SOTA adversarially robust models are mostly based on adversarial training (AT) and differ only by some regularizers either at inner maximization or outer minimization steps. Being repetitive in nature during the inner maximization step, they take a huge time to train. We propose a non-iterative method that enforces the following ideas during training. Attribution maps are more aligned to the actual object in the image for adversarially robust models compared to naturally trained models. Also, the allowed set of pixels to perturb an image (that changes model decision) should be restrict…
Wuhan University（武汉大学）；Pennsylvania State University（宾夕法尼亚州立大学）
Adversarial examples are well-designed input samples, in which perturbations are imperceptible to the human eyes, but easily mislead the output of deep neural networks (DNNs). Existing works synthesize adversarial examples by leveraging simple metrics to penalize perturbations, that lack sufficient consideration of the human visual system (HVS), which produces noticeable artifacts. To explore why the perturbations are visible, this paper summarizes four primary factors affecting the perceptibility of human eyes. Based on this investigation, we design a multi-factor metric MulFactorLoss for mea…
National University of Singapore（新加坡国立大学）；University of California, Berkeley（加州大学伯克利分校）
Adversarial examples have been shown to exist for a variety of deep learning architectures. Deep reinforcement learning has shown promising results on training agent policies directly on raw inputs such as image pixels. In this paper we present a novel study into adversarial attacks on deep reinforcement learning polices. We compare the effectiveness of the attacks using adversarial examples vs. random noise. We present a novel method for reducing the number of times adversarial examples need to be injected for a successful attack, based on the value function. We further explore how re-trainin…
Deep neural networks perform well on real world data but are prone to adversarial perturbations: small changes in the input easily lead to misclassification. In this work, we propose an attack methodology not only for cases where the perturbations are measured by $\ell_p$ norms, but in fact any adversarial dissimilarity metric with a closed proximal form. This includes, but is not limited to, $\ell_1, \ell_2$, and $\ell_\infty$ perturbations; the $\ell_0$ counting "norm" (i.e. true sparseness); and the total variation seminorm, which is a (non-$\ell_p$) convolutional dissimilarity measuring lo…
Batch Normalization (BatchNorm) is effective for improving the performance and accelerating the training of deep neural networks. However, it has also shown to be a cause of adversarial vulnerability, i.e., networks without it are more robust to adversarial attacks. In this paper, we investigate how BatchNorm causes this vulnerability and proposed new normalization that is robust to adversarial attacks. We first observe that adversarial images tend to shift the distribution of BatchNorm input, and this shift makes train-time estimated population statistics inaccurate. We hypothesize that these…
There exists a vast number of adversarial attacks and defences for machine learning algorithms of various types which makes assessing the robustness of algorithms a daunting task. To make matters worse, there is an intrinsic bias in these adversarial algorithms. Here, we organise the problems faced: a) Model Dependence, b) Insufficient Evaluation, c) False Adversarial Samples, and d) Perturbation Dependent Results). Based on this, we propose a model agnostic dual quality assessment method, together with the concept of robustness levels to tackle them. We validate the dual quality assessment on…
UC San Diego（加州大学圣地亚哥分校）；SPAWAR（SPAWAR（太空与海战系统司令部））
We present a neural network architecture based upon the Autoencoder (AE) and Generative Adversarial Network (GAN) that promotes a convex latent distribution by training adversarially on latent space interpolations. By using an AE as both the generator and discriminator of a GAN, we pass a pixel-wise error function across the discriminator, yielding an AE which produces non-blurry samples that match both high- and low-level features of the original images. Interpolations between images in this space remain within the latent-space distribution of real images as trained by the discriminator, and …
We propose a novel method for creating adversarial examples. Instead of perturbing pixels, we use an encoder-decoder representation of the input image and perturb intermediate layers in the decoder. This changes the high-level features provided by the generative model. Therefore, our perturbation possesses semantic meaning, such as a longer beak or green tints. We formulate this task as an optimization problem by minimizing the Wasserstein distance between the adversarial and initial images under a misclassification constraint. We employ the projected gradient method with a simple inexact proj…
Indian Institute of Technology, Hyderabad（海得拉巴印度理工学院）；Microsoft Research India（微软研究院印度分院）；Chennai Mathematical Institue, Chennai（钦奈数学研究所）
(Non-)robustness of neural networks to small, adversarial pixel-wise perturbations, and as more recently shown, to even random spatial transformations (e.g., translations, rotations) entreats both theoretical and empirical understanding. Spatial robustness to random translations and rotations is commonly attained via equivariant models (e.g., StdCNNs, GCNNs) and training augmentation, whereas adversarial robustness is typically achieved by adversarial training. In this paper, we prove a quantitative trade-off between spatial and adversarial robustness in a simple statistical setting. We comple…
Image normalization is a critical step in medical imaging. This step is often done on a per-dataset basis, preventing current segmentation algorithms from the full potential of exploiting jointly normalized information across multiple datasets. To solve this problem, we propose an adversarial normalization approach for image segmentation which learns common normalizing functions across multiple datasets while retaining image realism. The adversarial training provides an optimal normalizer that improves both the segmentation accuracy and the discrimination of unrealistic normalizing functions. …
King Abdullah University of Science and Technology (KAUST)（阿卜杜拉国王科技大学）
One major factor impeding more widespread adoption of deep neural networks (DNNs) is their lack of robustness, which is essential for safety-critical applications such as autonomous driving. This has motivated much recent work on adversarial attacks for DNNs, which mostly focus on pixel-level perturbations void of semantic meaning. In contrast, we present a general framework for adversarial attacks on trained agents, which covers semantic perturbations to the environment of the agent performing the task as well as pixel-level attacks. To do this, we re-frame the adversarial attack problem as l…
University of Central Florida（中佛罗里达大学）；École Normale Supérieure（巴黎高等师范学院）；INRIA（法国国家信息与自动化研究所）
White box adversarial perturbations are sought via iterative optimization algorithms most often minimizing an adversarial loss on a $l_p$ neighborhood of the original image, the so-called distortion set. Constraining the adversarial search with different norms results in disparately structured adversarial examples. Here we explore several distortion sets with structure-enhancing algorithms. These new structures for adversarial examples, yet pervasive in optimization, are for instance a challenge for adversarial theoretical certification which again provides only $l_p$ certificates. Because adv…
We propose functional adversarial attacks, a novel class of threat models for crafting adversarial examples to fool machine learning models. Unlike a standard $\ell_p$-ball threat model, a functional adversarial threat model allows only a single function to be used to perturb input features to produce an adversarial example. For example, a functional adversarial attack applied on colors of an image can change all red pixels simultaneously to light red. Such global uniform changes in images can be less perceptible than perturbing pixels of the image individually. For simplicity, we refer to fun…
Obtaining accurate pixel-level localization from class labels is a crucial process in weakly supervised semantic segmentation and object localization. Attribution maps from a trained classifier are widely used to provide pixel-level localization, but their focus tends to be restricted to a small discriminative region of the target object. An AdvCAM is an attribution map of an image that is manipulated to increase the classification score produced by a classifier before the final softmax or sigmoid layer. This manipulation is realized in an anti-adversarial manner, so that the original image is…
University of California San Diego（加州大学圣迭戈分校）；Politecnico di Milano（米兰理工大学）；SRI International（斯坦福国际研究院）
We study black-box adversarial attacks for image classifiers in a constrained threat model, where adversaries can only modify a small fraction of pixels in the form of scratches on an image. We show that it is possible for adversaries to generate localized \textit{adversarial scratches} that cover less than $5\%$ of the pixels in an image and achieve targeted success rates of $98.77\%$ and $97.20\%$ on ImageNet and CIFAR-10 trained ResNet-50 models, respectively. We demonstrate that our scratches are effective under diverse shapes, such as straight lines or parabolic B\a'ezier curves, with sin…
Tsinghua Shenzhen International Graduate School, Tsinghua University（清华大学深圳国际研究生院）；PCL Research Center of Networks and Communications, Peng Cheng Laboratory（鹏城实验室网络通信PCL研究中心）
Recently, deep neural networks (DNNs) have been widely and successfully used in Object Detection, e.g. Faster RCNN, YOLO, CenterNet. However, recent studies have shown that DNNs are vulnerable to adversarial attacks. Adversarial attacks against object detection can be divided into two categories, whole-pixel attacks and patch attacks. While these attacks add perturbations to a large number of pixels in images, we proposed a diffused patch attack (\textbf{DPAttack}) to successfully fool object detectors by diffused patches of asteroid-shaped or grid-shape, which only change a small number of pi…
Current adversarial attacks for motion estimation (optical flow) optimize small per-pixel perturbations, which are unlikely to appear in the real world. In contrast, we exploit a real-world weather phenomenon for a novel attack with adversarially optimized snow. At the core of our attack is a differentiable renderer that consistently integrates photorealistic snowflakes with realistic motion into the 3D scene. Through optimization we obtain adversarial snow that significantly impacts the optical flow while being indistinguishable from ordinary snow. Surprisingly, the impact of our novel attack…
Zhejiang University（浙江大学）；AZFT Joint Lab for Knowledge Engine（AZFT知识引擎联合实验室）；Alibaba Group（阿里巴巴集团）
Recent neural-based relation extraction approaches, though achieving promising improvement on benchmark datasets, have reported their vulnerability towards adversarial attacks. Thus far, efforts mostly focused on generating adversarial samples or defending adversarial attacks, but little is known about the difference between normal and adversarial samples. In this work, we take the first step to leverage the salience-based method to analyze those adversarial samples. We observe that salience tokens have a direct correlation with adversarial perturbations. We further find the adversarial pertur…
This paper aims to explain adversarial attacks in terms of how adversarial perturbations contribute to the attacking task. We estimate attributions of different image regions to the decrease of the attacking cost based on the Shapley value. We define and quantify interactions among adversarial perturbation pixels, and decompose the entire perturbation map into relatively independent perturbation components. The decomposition of the perturbation map shows that adversarially-trained DNNs have more perturbation components in the foreground than normally-trained DNNs. Moreover, compared to the nor…
