While convolutional neural networks (CNNs) have achieved excellent performances in various computer vision tasks, they often misclassify with malicious samples, a.k.a. adversarial examples. Adversarial training is a popular and straightforward technique to defend against the threat of adversarial examples. Unfortunately, CNNs must sacrifice the accuracy of standard samples to improve robustness against adversarial examples when adversarial training is used. In this work, we propose Masking and Mixing Adversarial Training (M2AT) to mitigate the trade-off between accuracy and robustness. We focus on creating diverse adversarial examples during training. Specifically, our approach consists of two processes: 1) masking a perturbation with a binary mask and 2) mixing two partially perturbed images. Experimental results on CIFAR-10 dataset demonstrate that our method achieves better robustness against several adversarial attacks than previous methods.
Adversarial examples provoke weak reliability and potential security issues in deep neural networks. Although adversarial training has been widely studied to improve adversarial robustness, it works in an over-parameterized regime and requires high computations and large memory budgets. To bridge adversarial robustness and model compression, we propose a novel adversarial pruning method, Masking Adversarial Damage (MAD) that employs second-order information of adversarial loss. By using it, we can accurately estimate adversarial saliency for model parameters and determine which parameters can …
Mohamed Bin Zayed University of Artificial Intelligence（穆罕默德·本·扎耶德人工智能大学）；Australian National University（澳大利亚国立大学）；Qualcomm（高通公司）；Linköping University（林雪平大学）
Adversarial training is an effective approach to make deep neural networks robust against adversarial attacks. Recently, different adversarial training defenses are proposed that not only maintain a high clean accuracy but also show significant robustness against popular and well studied adversarial attacks such as PGD. High adversarial robustness can also arise if an attack fails to find adversarial gradient directions, a phenomenon known as `gradient masking'. In this work, we analyse the effect of label smoothing on adversarial training as one of the potential causes of gradient masking. We…
Peking University（北京大学）；The University of Tokyo（东京大学）
Vision Transformers (ViTs) have recently achieved competitive performance in broad vision tasks. Unfortunately, on popular threat models, naturally trained ViTs are shown to provide no more adversarial robustness than convolutional neural networks (CNNs). Adversarial training is still required for ViTs to defend against such adversarial attacks. In this paper, we provide the first and comprehensive study on the adversarial training recipe of ViTs via extensive evaluation of various training techniques across benchmark datasets. We find that pre-training and SGD optimizer are necessary for ViTs…
Generating Transferrable Adversarial Examples via Local Mixing and Logits Optimization for Remote Sensing Object Recognition
State Key Laboratory of Spatial Datum, College of Remote Sensing and Geoinformatics Engineering, Henan University（空间基准国家重点实验室、遥感与地理信息工程学院、河南大学）；School of Computer and Information Engineering, Henan University（计算机与信息工程学院、河南大学）；School of Automation Science and Electrical Engineering, Beihang University（自动化科学与电气工程学院、北京航空航天大学）
The progress in the last decade has enabled machine learning models to achieve impressive performance across a wide range of tasks in Computer Vision. However, a plethora of works have demonstrated the susceptibility of these models to adversarial samples. Adversarial training procedure has been proposed to defend against such adversarial attacks. Adversarial training methods augment mini-batches with adversarial samples, and typically single-step (non-iterative) methods are used for generating these adversarial samples. However, models trained using single-step adversarial training converge t…
Naval Center for Applied Research in Artificial Intelligence（海军人工智能应用研究中心）；U.S. Naval Research Laboratory（美国海军研究实验室）
Adversarial attacks and defenses are currently active areas of research for the deep learning community. A recent review paper divided the defense approaches into three categories; gradient masking, robust optimization, and adversarial example detection. We divide gradient masking and robust optimization differently: (1) increasing intra-class compactness and inter-class separation of the feature vectors improves adversarial robustness, and (2) marginalization or removal of non-robust image features also improves adversarial robustness. By reframing these topics differently, we provide a fresh…
Interdisciplinary Centre for Security, Reliability and Trust, University of Luxembourg（卢森堡大学安全、可靠性与信任跨学科中心）
Deep Neural Networks (DNNs) are vulnerable to adversarial examples, which causes serious threats to security-critical applications. This motivated much research on providing mechanisms to make models more robust against adversarial attacks. Unfortunately, most of these defenses, such as gradient masking, are easily overcome through different attack means. In this paper, we propose MUTEN, a low-cost method to improve the success rate of well-known attacks against gradient-masking models. Our idea is to apply the attacks on an ensemble model which is built by mutating the original model elements…
To combat against adversarial attacks, autoencoder structure is widely used to perform denoising which is regarded as gradient masking. In this paper, we revisit the role of autoencoders in adversarial settings. Through the comprehensive experimental results and analysis, this paper presents the inherent property of adversarial robustness in the autoencoders. We also found that autoencoders may use robust features that cause inherent adversarial robustness. We believe that our discovery of the adversarial robustness of the autoencoders can provide clues to the future research and applications …
In this work we revisit gradient regularization for adversarial robustness with some new ingredients. First, we derive new per-image theoretical robustness bounds based on local gradient information. These bounds strongly motivate input gradient regularization. Second, we implement a scaleable version of input gradient regularization which avoids double backpropagation: adversarially robust ImageNet models are trained in 33 hours on four consumer grade GPUs. Finally, we show experimentally and through theoretical certification that input gradient regularization is competitive with adversarial …
In this paper, we study fast training of adversarially robust models. From the analyses of the state-of-the-art defense method, i.e., the multi-step adversarial training, we hypothesize that the gradient magnitude links to the model robustness. Motivated by this, we propose to perturb both the image and the label during training, which we call Bilateral Adversarial Training (BAT). To generate the adversarial label, we derive an closed-form heuristic solution. To generate the adversarial image, we use one-step targeted attack with the target label being the most confusing class. In the experime…
The evaluation of robustness against adversarial manipulation of neural networks-based classifiers is mainly tested with empirical attacks as methods for the exact computation, even when available, do not scale to large networks. We propose in this paper a new white-box adversarial attack wrt the $l_p$-norms for $p \in \{1,2,\infty\}$ aiming at finding the minimal perturbation necessary to change the class of a given input. It has an intuitive geometric meaning, yields quickly high quality results, minimizes the size of the perturbation (so that it returns the robust accuracy at every threshol…
Attention-Guided Patch-Wise Sparse Adversarial Attacks on Vision-Language-Action Models
Shenzhen International Graduate School, Tsinghua University, Shenzhen, China（清华大学深圳国际研究生院）；Huazhong University of Science and Technology, Wuhan, China（华中科技大学）；Ping An Technology, Shenzhen, China（平安科技）
Adversarial attacks hamper the functionality and accuracy of Deep Neural Networks (DNNs) by meddling with subtle perturbations to their inputs.In this work, we propose a new Mask-based Adversarial Defense scheme (MAD) for DNNs to mitigate the negative effect from adversarial attacks. To be precise, our method promotes the robustness of a DNN by randomly masking a portion of potential adversarial images, and as a result, the %classification result output of the DNN becomes more tolerant to minor input perturbations. Compared with existing adversarial defense techniques, our method does not need…
The security of models based on new architectures such as MLP-Mixer and ViTs needs to be studied urgently. However, most of the current researches are mainly aimed at the adversarial attack against ViTs, and there is still relatively little adversarial work on MLP-mixer. We propose an adversarial attack method against MLP-Mixer called Maxwell's demon Attack (MA). MA breaks the channel-mixing and token-mixing mechanism of MLP-Mixer by controlling the part input of MLP-Mixer's each Mixer layer, and disturbs MLP-Mixer to obtain the main information of images. Our method can mask the part input of…
Salesforce Research Asia（赛富时亚洲研究院）；National University of Singapore（新加坡国立大学）；Nanyang Technological University（南洋理工大学）
Multilingual models have demonstrated impressive cross-lingual transfer performance. However, test sets like XNLI are monolingual at the example level. In multilingual communities, it is common for polyglots to code-mix when conversing with each other. Inspired by this phenomenon, we present two strong black-box adversarial attacks (one word-level, one phrase-level) for multilingual models that push their ability to handle code-mixed sentences to the limit. The former uses bilingual dictionaries to propose perturbations and translations of the clean example for sense disambiguation. The latter…
Ghent University（根特大学）；Ghent University Global Campus（根特大学全球校区）
Detecting adversarial examples currently stands as one of the biggest challenges in the field of deep learning. Adversarial attacks, which produce adversarial examples, increase the prediction likelihood of a target class for a particular data point. During this process, the adversarial example can be further optimized, even when it has already been wrongly classified with 100% confidence, thus making the adversarial example even more difficult to detect. For this kind of adversarial examples, which we refer to as over-optimized adversarial examples, we discovered that the logits of the model …
The field of defense strategies against adversarial attacks has significantly grown over the last years, but progress is hampered as the evaluation of adversarial defenses is often insufficient and thus gives a wrong impression of robustness. Many promising defenses could be broken later on, making it difficult to identify the state-of-the-art. Frequent pitfalls in the evaluation are improper tuning of hyperparameters of the attacks, gradient obfuscation or masking. In this paper we first propose two extensions of the PGD-attack overcoming failures due to suboptimal step size and problems of t…
Novelty detection is the task of recognizing samples that do not belong to the distribution of the target class. During training, the novelty class is absent, preventing the use of traditional classification approaches. Deep autoencoders have been widely used as a base of many unsupervised novelty detection methods. In particular, context autoencoders have been successful in the novelty detection task because of the more effective representations they learn by reconstructing original images from randomly masked images. However, a significant drawback of context autoencoders is that random mask…
UnICLAM:Contrastive Representation Learning with Adversarial Masking for Unified and Interpretable Medical Vision Question Answering
Zhejiang University（浙江大学）；University of Illinois Urbana–Champaign Institute（伊利诺伊大学厄巴纳-香槟分校研究院）；Sixth Medical Center of Chinese PLA General Hospital（中国人民解放军总医院第六医学中心）
Carnegie Mellon University（卡内基梅隆大学）；Amazon Web Services（亚马逊云科技）；Amazon（亚马逊）
Recent methods in self-supervised learning have demonstrated that masking-based pretext tasks extend beyond NLP, serving as useful pretraining objectives in computer vision. However, existing approaches apply random or ad hoc masking strategies that limit the difficulty of the reconstruction task and, consequently, the strength of the learnt representations. We improve upon current state-of-the-art work in learning adversarial masks by proposing a new framework that generates masks in a sequential fashion with different constraints on the adversary. This leads to improvements in performance on…
Beijing University of Posts and Telecommunications（北京邮电大学）；Commonwealth Science and Industrial Research Organisation (CSIRO)（联邦科学与工业研究组织（CSIRO））；University of Manitoba（曼尼托巴大学）；Princeton University（普林斯顿大学）
Adversarial attacks and defenses in machine learning and deep neural network have been gaining significant attention due to the rapidly growing applications of deep learning in the Internet and relevant scenarios. This survey provides a comprehensive overview of the recent advancements in the field of adversarial attack and defense techniques, with a focus on deep neural network-based classification models. Specifically, we conduct a comprehensive classification of recent adversarial attack methods and state-of-the-art adversarial defense techniques based on attack principles, and present them…
Exploring Gradient-Guided Masked Language Model to Detect Textual Adversarial Attacks
School of Information and Communication Technology, Griffith University（格里菲斯大学信息与通信技术学院）；School of Computer Science, University of Technology Sydney（悉尼科技大学计算机科学学院）；College of Computer and Information Science, Southwest University（西南大学计算机与信息科学学院）；School of Cyber Science and Engineering, Huazhong University of Science and Technology（华中科技大学网络空间科学与工程学院）
Collecting large-scale datasets is crucial for training deep models, annotating the data, however, inevitably yields noisy labels, which poses challenges to deep learning algorithms. Previous efforts tend to mitigate this problem via identifying and removing noisy samples or correcting their labels according to the statistical properties (e.g., loss values) among training samples. In this paper, we aim to tackle this problem from a new perspective, delving into the deep feature maps, we empirically find that models trained with clean and mislabeled samples manifest distinguishable activation f…
University of Oxford（牛津大学）；The University of Edinburgh（爱丁堡大学）；The Alan Turing Institute（艾伦·图灵研究所）；DeepMind（深度思维）
We propose ADIOS, a masked image model (MIM) framework for self-supervised learning, which simultaneously learns a masking function and an image encoder using an adversarial objective. The image encoder is trained to minimise the distance between representations of the original and that of a masked image. The masking function, conversely, aims at maximising this distance. ADIOS consistently improves on state-of-the-art self-supervised learning (SSL) methods on a variety of tasks and datasets -- including classification on ImageNet100 and STL10, transfer learning on CIFAR10/100, Flowers102 and …
Recent work has put forth the hypothesis that adversarial vulnerabilities in neural networks are due to them overusing "non-robust features" inherent in the training data. We show empirically that for PGD-attacks, there is a training stage where neural networks start heavily relying on non-robust features to boost natural accuracy. We also propose a mechanism reducing vulnerability to PGD-style attacks consisting of mixing in a certain amount of images contain-ing mostly "robust features" into each training batch, and then show that robust accuracy is improved, while natural accuracy is not su…
We propose the Square Attack, a score-based black-box $l_2$- and $l_\infty$-adversarial attack that does not rely on local gradient information and thus is not affected by gradient masking. Square Attack is based on a randomized search scheme which selects localized square-shaped updates at random positions so that at each iteration the perturbation is situated approximately at the boundary of the feasible set. Our method is significantly more query efficient and achieves a higher success rate compared to the state-of-the-art methods, especially in the untargeted setting. In particular, on Ima…
Georgia Institute of Technology（佐治亚理工学院）；IBM Research（IBM研究院）
Discovering and exploiting the causality in deep neural networks (DNNs) are crucial challenges for understanding and reasoning causal effects (CE) on an explainable visual model. "Intervention" has been widely used for recognizing a causal relation ontologically. In this paper, we propose a causal inference framework for visual reasoning via do-calculus. To study the intervention effects on pixel-level features for causal reasoning, we introduce pixel-wise masking and adversarial perturbation. In our framework, CE is calculated using features in a latent space and perturbed prediction from a D…
Princeton University（普林斯顿大学）；University of Chicago（芝加哥大学）
Localized adversarial patches aim to induce misclassification in machine learning models by arbitrarily modifying pixels within a restricted region of an image. Such attacks can be realized in the physical world by attaching the adversarial patch to the object to be misclassified, and defending against such attacks is an unsolved/open problem. In this paper, we propose a general defense framework called PatchGuard that can achieve high provable robustness while maintaining high clean accuracy against localized adversarial patches. The cornerstone of PatchGuard involves the use of CNNs with sma…
Object detectors, which are widely deployed in security-critical systems such as autonomous vehicles, have been found vulnerable to patch hiding attacks. An attacker can use a single physically-realizable adversarial patch to make the object detector miss the detection of victim objects and undermine the functionality of object detection applications. In this paper, we propose ObjectSeeker for certifiably robust object detection against patch hiding attacks. The key insight in ObjectSeeker is patch-agnostic masking: we aim to mask out the entire adversarial patch without knowing the shape, siz…
Beyond Dropout: Robust Convolutional Neural Networks Based on Local Feature Masking
School of Informatics, Xiamen University（厦门大学信息学院）；School of Electronic Science and Engineering, Xiamen University（厦门大学电子科学与技术学院）；College of Computer and Cyber Security, Fujian Normal University（福建师范大学计算机与网络空间安全学院）
Adversarial Masking Contrastive Learning for vein recognition
Chongqing Technology and Business University（重庆工商大学）；Telecom SudParis（巴黎电信学院）；University Paris-Saclay（巴黎-萨克雷大学）；China University of Mining and Technology（中国矿业大学）
Neural network compression methods like pruning and quantization are very effective at efficiently deploying Deep Neural Networks (DNNs) on edge devices. However, DNNs remain vulnerable to adversarial examples-inconspicuous inputs that are specifically designed to fool these models. In particular, Universal Adversarial Perturbations (UAPs), are a powerful class of adversarial attacks which create adversarial perturbations that can generalize across a large set of inputs. In this work, we analyze the effect of various compression techniques to UAP attacks, including different forms of pruning a…
Deep Neural Networks (DNNs) have been shown to be vulnerable against adversarial examples, which are data points cleverly constructed to fool the classifier. Such attacks can be devastating in practice, especially as DNNs are being applied to ever increasing critical tasks like image recognition in autonomous driving. In this paper, we introduce a new perspective on the problem. We do so by first defining robustness of a classifier to adversarial exploitation. Next, we show that the problem of adversarial example generation can be posed as learning problem. We also categorize attacks in litera…
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）
Modern face recognition systems (FRS) still fall short when the subjects are wearing facial masks, a common theme in the age of respiratory pandemics. An intuitive partial remedy is to add a mask detector to flag any masked faces so that the FRS can act accordingly for those low-confidence masked faces. In this work, we set out to investigate the potential vulnerability of such FRS equipped with a mask detector, on large-scale masked faces, which might trigger a serious risk, e.g., letting a suspect evade the FRS where both facial identity and mask are undetected. As existing face recognizers …
