Unlearnable datasets lead to a drastic drop in the generalization performance of models trained on them by introducing elaborate and imperceptible perturbations into clean training sets. Many existing defenses, e.g., JPEG compression and adversarial training, effectively counter UDs based on norm-constrained additive noise. However, a fire-new type of convolution-based UDs have been proposed and render existing defenses all ineffective, presenting a greater challenge to defenders. To address this, we express the convolution-based unlearnable sample as the result of multiplying a matrix by a clean sample in a simplified scenario, and formalize the intra-class matrix inconsistency as $\Theta_{imi}$, inter-class matrix consistency as $\Theta_{imc}$ to investigate the working mechanism of the convolution-based UDs. We conjecture that increasing both of these metrics will mitigate the unlearnability effect. Through validation experiments that commendably support our hypothesis, we further design a random matrix to boost both $\Theta_{imi}$ and $\Theta_{imc}$, achieving a notable degree of defense effect. Hence, by building upon and extending these facts, we first propose a brand-new image COrruption that employs randomly multiplicative transformation via INterpolation operation to successfully defend against convolution-based UDs. Our approach leverages global pixel random interpolations, effectively suppressing the impact of multiplicative noise in convolution-based UDs. Additionally, we have also designed two new forms of convolution-based UDs, and find that our defense is the most effective against them.
Transformer-based pre-trained models of code (PTMC) have been widely utilized and have achieved state-of-the-art performance in many mission-critical applications. However, they can be vulnerable to adversarial attacks through identifier substitution or coding style transformation, which can significantly degrade accuracy and may further incur security concerns. Although several approaches have been proposed to generate adversarial examples for PTMC, the effectiveness and efficiency of such approaches, especially on different code intelligence tasks, has not been well understood. To bridge this gap, this study systematically analyzes five state-of-the-art adversarial attack approaches from three perspectives: effectiveness, efficiency, and the quality of generated examples. The results show that none of the five approaches balances all these perspectives. Particularly, approaches with a high attack success rate tend to be time-consuming; the adversarial code they generate often lack naturalness, and vice versa. To address this limitation, we explore the impact of perturbing identifiers under different contexts and find that identifier substitution within for and if statements is the most effective. Based on these findings, we propose a new approach that prioritizes different types of statements for various tasks and further utilizes beam search to generate adversarial examples. Evaluation results show that it outperforms the state-of-the-art ALERT in terms of both effectiveness and efficiency while preserving the naturalness of the generated adversarial examples.
Self-supervised learning usually uses a large amount of unlabeled data to pre-train an encoder which can be used as a general-purpose feature extractor, such that downstream users only need to perform fine-tuning operations to enjoy the benefit of "large model". Despite this promising prospect, the security of pre-trained encoder has not been thoroughly investigated yet, especially when the pre-trained encoder is publicly available for commercial use. In this paper, we propose AdvEncoder, the first framework for generating downstream-agnostic universal adversarial examples based on the pre-trained encoder. AdvEncoder aims to construct a universal adversarial perturbation or patch for a set of natural images that can fool all the downstream tasks inheriting the victim pre-trained encoder. Unlike traditional adversarial example works, the pre-trained encoder only outputs feature vectors rather than classification labels. Therefore, we first exploit the high frequency component information of the image to guide the generation of adversarial examples. Then we design a generative attack framework to construct adversarial perturbations/patches by learning the distribution of the attack surrogate dataset to improve their attack success rates and transferability. Our results show that an attacker can successfully attack downstream tasks without knowing either the pre-training dataset or the downstream dataset. We also tailor four defenses for pre-trained encoders, the results of which further prove the attack ability of AdvEncoder.
Multimodal contrastive learning aims to train a general-purpose feature extractor, such as CLIP, on vast amounts of raw, unlabeled paired image-text data. This can greatly benefit various complex downstream tasks, including cross-modal image-text retrieval and image classification. Despite its promising prospect, the security issue of cross-modal pre-trained encoder has not been fully explored yet, especially when the pre-trained encoder is publicly available for commercial use. In this work, we propose AdvCLIP, the first attack framework for generating downstream-agnostic adversarial examples based on cross-modal pre-trained encoders. AdvCLIP aims to construct a universal adversarial patch for a set of natural images that can fool all the downstream tasks inheriting the victim cross-modal pre-trained encoder. To address the challenges of heterogeneity between different modalities and unknown downstream tasks, we first build a topological graph structure to capture the relevant positions between target samples and their neighbors. Then, we design a topology-deviation based generative adversarial network to generate a universal adversarial patch. By adding the patch to images, we minimize their embeddings similarity to different modality and perturb the sample distribution in the feature space, achieving unviersal non-targeted attacks. Our results demonstrate the excellent attack performance of AdvCLIP on two types of downstream tasks across eight datasets. We also tailor three popular defenses to mitigate AdvCLIP, highlighting the need for new defense mechanisms to defend cross-modal pre-trained encoders.
Adversarial examples (AEs) for DNNs have been shown to be transferable: AEs that successfully fool white-box surrogate models can also deceive other black-box models with different architectures. Although a bunch of empirical studies have provided guidance on generating highly transferable AEs, many of these findings lack explanations and even lead to inconsistent advice. In this paper, we take a further step towards understanding adversarial transferability, with a particular focus on surrogate aspects. Starting from the intriguing little robustness phenomenon, where models adversarially trained with mildly perturbed adversarial samples can serve as better surrogates, we attribute it to a trade-off between two predominant factors: model smoothness and gradient similarity. Our investigations focus on their joint effects, rather than their separate correlations with transferability. Through a series of theoretical and empirical analyses, we conjecture that the data distribution shift in adversarial training explains the degradation of gradient similarity. Building on these insights, we explore the impacts of data augmentation and gradient regularization on transferability and identify that the trade-off generally exists in the various training mechanisms, thus building a comprehensive blueprint for the regulation mechanism behind transferability. Finally, we provide a general route for constructing better surrogates to boost transferability which optimizes both model smoothness and gradient similarity simultaneously, e.g., the combination of input gradient regularization and sharpness-aware minimization (SAM), validated by extensive experiments. In summary, we call for attention to the united impacts of these two factors for launching effective transfer attacks, rather than optimizing one while ignoring the other, and emphasize the crucial role of manipulating surrogate models.
Optimal margin Distribution Machine (ODM) is a newly proposed statistical learning framework rooting in the novel margin theory, which demonstrates better generalization performance than the traditional large margin based counterparts. Nonetheless, it suffers from the ubiquitous scalability problem regarding both computation time and memory as other kernel methods. This paper proposes a scalable ODM, which can achieve nearly ten times speedup compared to the original ODM training method. For nonlinear kernels, we propose a novel distribution-aware partition method to make the local ODM trained on each partition be close and converge fast to the global one. When linear kernel is applied, we extend a communication efficient SVRG method to accelerate the training further. Extensive empirical studies validate that our proposed method is highly computational efficient and almost never worsen the generalization.
Deep neural networks are proven to be vulnerable to backdoor attacks. Detecting the trigger samples during the inference stage, i.e., the test-time trigger sample detection, can prevent the backdoor from being triggered. However, existing detection methods often require the defenders to have high accessibility to victim models, extra clean data, or knowledge about the appearance of backdoor triggers, limiting their practicality. In this paper, we propose the test-time corruption robustness consistency evaluation (TeCo), a novel test-time trigger sample detection method that only needs the hard-label outputs of the victim models without any extra information. Our journey begins with the intriguing observation that the backdoor-infected models have similar performance across different image corruptions for the clean images, but perform discrepantly for the trigger samples. Based on this phenomenon, we design TeCo to evaluate test-time robustness consistency by calculating the deviation of severity that leads to predictions' transition across different corruptions. Extensive experiments demonstrate that compared with state-of-the-art defenses, which even require either certain information about the trigger types or accessibility of clean data, TeCo outperforms them on different backdoor attacks, datasets, and model architectures, enjoying a higher AUROC by 10% and 5 times of stability.
Point cloud completion, as the upstream procedure of 3D recognition and segmentation, has become an essential part of many tasks such as navigation and scene understanding. While various point cloud completion models have demonstrated their powerful capabilities, their robustness against adversarial attacks, which have been proven to be fatally malicious towards deep neural networks, remains unknown. In addition, existing attack approaches towards point cloud classifiers cannot be applied to the completion models due to different output forms and attack purposes. In order to evaluate the robustness of the completion models, we propose PointCA, the first adversarial attack against 3D point cloud completion models. PointCA can generate adversarial point clouds that maintain high similarity with the original ones, while being completed as another object with totally different semantic information. Specifically, we minimize the representation discrepancy between the adversarial example and the target point set to jointly explore the adversarial point clouds in the geometry space and the feature space. Furthermore, to launch a stealthier attack, we innovatively employ the neighbourhood density information to tailor the perturbation constraint, leading to geometry-aware and distribution-adaptive modifications for each point. Extensive experiments against different premier point cloud completion networks show that PointCA can cause a performance degradation from 77.9% to 16.7%, with the structure chamfer distance kept below 0.01. We conclude that existing completion models are severely vulnerable to adversarial examples, and state-of-the-art defenses for point cloud classification will be partially invalid when applied to incomplete and uneven point cloud data.