Zhuoran Liu

Zhuoran Liu, Zhengyu Zhao, Martha Larson

Perturbative availability poisons (PAPs) add small changes to images to prevent their use for model training. Current research adopts the belief that practical and effective approaches to countering PAPs do not exist. In this paper, we argue that it is time to abandon this belief. We present extensive experiments showing that 12 state-of-the-art PAP methods are vulnerable to Image Shortcut Squeezing (ISS), which is based on simple compression. For example, on average, ISS restores the CIFAR-10 model accuracy to $81.73\%$, surpassing the previous best preprocessing-based countermeasures by $37.97\%$ absolute. ISS also (slightly) outperforms adversarial training and has higher generalizability to unseen perturbation norms and also higher efficiency. Our investigation reveals that the property of PAP perturbations depends on the type of surrogate model used for poison generation, and it explains why a specific ISS compression yields the best performance for a specific type of PAP perturbation. We further test stronger, adaptive poisoning, and show it falls short of being an ideal defense against ISS. Overall, our results demonstrate the importance of considering various (simple) countermeasures to ensure the meaningfulness of analysis carried out during the development of PAP methods.

Dirren van Vlijmen, Alex Kolmus, Zhuoran Liu et al.

We introduce ShortcutGen, a new data poisoning attack that generates sample-dependent, error-minimizing perturbations by learning a generator. The key novelty of ShortcutGen is the use of a randomly-initialized discriminator, which provides spurious shortcuts needed for generating poisons. Different from recent, iterative methods, our ShortcutGen can generate perturbations with only one forward pass in a label-free manner, and compared to the only existing generative method, DeepConfuse, our ShortcutGen is faster and simpler to train while remaining competitive. We also demonstrate that integrating a simple augmentation strategy can further boost the robustness of ShortcutGen against early stopping, and combining augmentation and non-augmentation leads to new state-of-the-art results in terms of final validation accuracy, especially in the challenging, transfer scenario. Lastly, we speculate, through uncovering its working mechanism, that learning a more general representation space could allow ShortcutGen to work for unseen data.

Hamid Bostani, Zhengyu Zhao, Zhuoran Liu et al.

Machine Learning (ML) promises to enhance the efficacy of Android Malware Detection (AMD); however, ML models are vulnerable to realistic evasion attacks--crafting realizable Adversarial Examples (AEs) that satisfy Android malware domain constraints. To eliminate ML vulnerabilities, defenders aim to identify susceptible regions in the feature space where ML models are prone to deception. The primary approach to identifying vulnerable regions involves investigating realizable AEs, but generating these feasible apps poses a challenge. For instance, previous work has relied on generating either feature-space norm-bounded AEs or problem-space realizable AEs in adversarial hardening. The former is efficient but lacks full coverage of vulnerable regions while the latter can uncover these regions by satisfying domain constraints but is known to be time-consuming. To address these limitations, we propose an approach to facilitate the identification of vulnerable regions. Specifically, we introduce a new interpretation of Android domain constraints in the feature space, followed by a novel technique that learns them. Our empirical evaluations across various evasion attacks indicate effective detection of AEs using learned domain constraints, with an average of 89.6%. Furthermore, extensive experiments on different Android malware detectors demonstrate that utilizing our learned domain constraints in Adversarial Training (AT) outperforms other AT-based defenses that rely on norm-bounded AEs or state-of-the-art non-uniform perturbations. Finally, we show that retraining a malware detector with a wide variety of feature-space realizable AEs results in a 77.9% robustness improvement against realizable AEs generated by unknown problem-space transformations, with up to 70x faster training than using problem-space realizable AEs.

Danpei Zhao, Zhuoran Liu, Bo Yuan

Improving search efficiency serves as one of the crucial objectives of Neural Architecture Search (NAS). However, many current approaches ignore the universality of the search strategy and fail to reduce the computational redundancy during the search process, especially in one-shot NAS architectures. Besides, current NAS methods show invalid reparameterization in non-linear search space, leading to poor efficiency in common search spaces like DARTS. In this paper, we propose TopoNAS, a model-agnostic approach for gradient-based one-shot NAS that significantly reduces searching time and memory usage by topological simplification of searchable paths. Firstly, we model the non-linearity in search spaces to reveal the parameterization difficulties. To improve the search efficiency, we present a topological simplification method and iteratively apply module-sharing strategies to simplify the topological structure of searchable paths. In addition, a kernel normalization technique is also proposed to preserve the search accuracy. Experimental results on the NASBench201 benchmark with various search spaces demonstrate the effectiveness of our method. It proves the proposed TopoNAS enhances the performance of various architectures in terms of search efficiency while maintaining a high level of accuracy. The project page is available at https://xdedss.github.io/topo_simplification.

Loes van Bemmel, Zhuoran Liu, Nik Vaessen et al.

Recent research has proposed approaches that modify speech to defend against gender inference attacks. The goal of these protection algorithms is to control the availability of information about a speaker's gender, a privacy-sensitive attribute. Currently, the common practice for developing and testing gender protection algorithms is "neural-on-neural", i.e., perturbations are generated and tested with a neural network. In this paper, we propose to go beyond this practice to strengthen the study of gender protection. First, we demonstrate the importance of testing gender inference attacks that are based on speech features historically developed by speech scientists, alongside the conventionally used neural classifiers. Next, we argue that researchers should use speech features to gain insight into how protective modifications change the speech signal. Finally, we point out that gender-protection algorithms should be compared with novel "vocal adversaries", human-executed voice adaptations, in order to improve interpretability and enable before-the-mic protection.

Zhengyu Zhao, Zhuoran Liu, Martha Larson

Achieving transferability of targeted attacks is reputed to be remarkably difficult. Currently, state-of-the-art approaches are resource-intensive because they necessitate training model(s) for each target class with additional data. In our investigation, we find, however, that simple transferable attacks which require neither additional data nor model training can achieve surprisingly high targeted transferability. This insight has been overlooked until now, mainly due to the widespread practice of unreasonably restricting attack optimization to a limited number of iterations. In particular, we, for the first time, identify that a simple logit loss can yield competitive results with the state of the arts. Our analysis spans a variety of transfer settings, especially including three new, realistic settings: an ensemble transfer setting with little model similarity, a worse-case setting with low-ranked target classes, and also a real-world attack against the Google Cloud Vision API. Results in these new settings demonstrate that the commonly adopted, easy settings cannot fully reveal the actual properties of different attacks and may cause misleading comparisons. We also show the usefulness of the simple logit loss for generating targeted universal adversarial perturbations in a data-free and training-free manner. Overall, the aim of our analysis is to inspire a more meaningful evaluation on targeted transferability. Code is available at https://github.com/ZhengyuZhao/Targeted-Tansfer

Zhengyu Zhao, Zhuoran Liu, Martha Larson

We introduce an approach that enhances images using a color filter in order to create adversarial effects, which fool neural networks into misclassification. Our approach, Adversarial Color Enhancement (ACE), generates unrestricted adversarial images by optimizing the color filter via gradient descent. The novelty of ACE is its incorporation of established practice for image enhancement in a transparent manner. Experimental results validate the white-box adversarial strength and black-box transferability of ACE. A range of examples demonstrates the perceptual quality of images that ACE produces. ACE makes an important contribution to recent work that moves beyond $L_p$ imperceptibility and focuses on unrestricted adversarial modifications that yield large perceptible perturbations, but remain non-suspicious, to the human eye. The future potential of filter-based adversaries is also explored in two directions: guiding ACE with common enhancement practices (e.g., Instagram filters) towards specific attractive image styles and adapting ACE to image semantics. Code is available at https://github.com/ZhengyuZhao/ACE.

Xu-Hui Zhou, Zhuo-Ran Liu, Heng Xiao

Bayesian inference offers a robust framework for updating prior beliefs based on new data using Bayes' theorem, but exact inference is often computationally infeasible, necessitating approximate methods. Though widely used, these methods struggle to estimate marginal likelihoods accurately, particularly due to the rigid functional structures of deterministic models like Gaussian processes and the limitations of small sample sizes in stochastic models like the ensemble Kalman method. In this work, we introduce BI-EqNO, an equivariant neural operator framework for generalized approximate Bayesian inference, designed to enhance both deterministic and stochastic approaches. BI-EqNO transforms priors into posteriors conditioned on observation data through data-driven training. The framework is flexible, supporting diverse prior and posterior representations with arbitrary discretizations and varying numbers of observations. Crucially, BI-EqNO's architecture ensures (1) permutation equivariance between prior and posterior representations, and (2) permutation invariance with respect to observational data. We demonstrate BI-EqNO's utility through two examples: (1) as a generalized Gaussian process (gGP) for regression, and (2) as an ensemble neural filter (EnNF) for sequential data assimilation. Results show that gGP outperforms traditional Gaussian processes by offering a more flexible representation of covariance functions. Additionally, EnNF not only outperforms the ensemble Kalman filter in small-ensemble settings but also has the potential to function as a "super" ensemble filter, capable of representing and integrating multiple ensemble filters for enhanced assimilation performance. This study highlights BI-EqNO's versatility and effectiveness, improving Bayesian inference through data-driven training while reducing computational costs across various applications.

Xiaoyun Xu, Zhuoran Liu, Stefanos Koffas et al.

Recent research on backdoor stealthiness focuses mainly on indistinguishable triggers in input space and inseparable backdoor representations in feature space, aiming to circumvent backdoor defenses that examine these respective spaces. However, existing backdoor attacks are typically designed to resist a specific type of backdoor defense without considering the diverse range of defense mechanisms. Based on this observation, we pose a natural question: Are current backdoor attacks truly a real-world threat when facing diverse practical defenses? To answer this question, we examine 12 common backdoor attacks that focus on input-space or feature-space stealthiness and 17 diverse representative defenses. Surprisingly, we reveal a critical blind spot: Backdoor attacks designed to be stealthy in input and feature spaces can be mitigated by examining backdoored models in parameter space. To investigate the underlying causes behind this common vulnerability, we study the characteristics of backdoor attacks in the parameter space. Notably, we find that input- and feature-space attacks introduce prominent backdoor-related neurons in parameter space, which are not thoroughly considered by current backdoor attacks. Taking comprehensive stealthiness into account, we propose a novel supply-chain attack called Grond. Grond limits the parameter changes by a simple yet effective module, Adversarial Backdoor Injection (ABI), which adaptively increases the parameter-space stealthiness during the backdoor injection. Extensive experiments demonstrate that Grond outperforms all 12 backdoor attacks against state-of-the-art (including adaptive) defenses on CIFAR-10, GTSRB, and a subset of ImageNet. In addition, we show that ABI consistently improves the effectiveness of common backdoor attacks.

Gorka Abad, Marina Krček, Stefanos Koffas et al.

Backdoor attacks pose a significant threat to deep learning models by implanting hidden vulnerabilities that can be activated by malicious inputs. While numerous defenses have been proposed to mitigate these attacks, the heterogeneous landscape of evaluation methodologies hinders fair comparison between defenses. This work presents a systematic (meta-)analysis of backdoor defenses through a comprehensive literature review and empirical evaluation. We analyzed 183 backdoor defense papers published between 2018 and 2025 across major AI and security venues, examining the properties and evaluation methodologies of these defenses. Our analysis reveals significant inconsistencies in experimental setups, evaluation metrics, and threat model assumptions in the literature. Through extensive experiments involving three datasets (MNIST, CIFAR-100, ImageNet-1K), four model architectures (ResNet-18, VGG-19, ViT-B/16, DenseNet-121), 16 representative defenses, and five commonly used attacks, totaling over 3\,000 experiments, we demonstrate that defense effectiveness varies substantially across different evaluation setups. We identify critical gaps in current evaluation practices, including insufficient reporting of computational overhead and behavior under benign conditions, bias in hyperparameter selection, and incomplete experimentation. Based on our findings, we provide concrete challenges and well-motivated recommendations to standardize and improve future defense evaluations. Our work aims to equip researchers and industry practitioners with actionable insights for developing, assessing, and deploying defenses to different systems.

Zhuoran Liu, Zhengyu Zhao, Alex Kolmus et al.

Recent work has shown that imperceptible perturbations can be applied to craft unlearnable examples (ULEs), i.e. images whose content cannot be used to improve a classifier during training. In this paper, we reveal the road that researchers should follow for understanding ULEs and improving ULEs as they were originally formulated (ULEOs). The paper makes four contributions. First, we show that ULEOs exploit color and, consequently, their effects can be mitigated by simple grayscale pre-filtering, without resorting to adversarial training. Second, we propose an extension to ULEOs, which is called ULEO-GrayAugs, that forces the generated ULEs away from channel-wise color perturbations by making use of grayscale knowledge and data augmentations during optimization. Third, we show that ULEOs generated using Multi-Layer Perceptrons (MLPs) are effective in the case of complex Convolutional Neural Network (CNN) classifiers, suggesting that CNNs suffer specific vulnerability to ULEs. Fourth, we demonstrate that when a classifier is trained on ULEOs, adversarial training will prevent a drop in accuracy measured both on clean images and on adversarial images. Taken together, our contributions represent a substantial advance in the state of art of unlearnable examples, but also reveal important characteristics of their behavior that must be better understood in order to achieve further improvements.

Zhuoran Liu, Niels Samwel, Léo Weissbart et al.

We introduce screen gleaning, a TEMPEST attack in which the screen of a mobile device is read without a visual line of sight, revealing sensitive information displayed on the phone screen. The screen gleaning attack uses an antenna and a software-defined radio (SDR) to pick up the electromagnetic signal that the device sends to the screen to display, e.g., a message with a security code. This special equipment makes it possible to recreate the signal as a gray-scale image, which we refer to as an emage. Here, we show that it can be used to read a security code. The screen gleaning attack is challenging because it is often impossible for a human viewer to interpret the emage directly. We show that this challenge can be addressed with machine learning, specifically, a deep learning classifier. Screen gleaning will become increasingly serious as SDRs and deep learning continue to rapidly advance. In this paper, we demonstrate the security code attack and we propose a testbed that provides a standard setup in which screen gleaning could be tested with different attacker models. Finally, we analyze the dimensions of screen gleaning attacker models and discuss possible countermeasures with the potential to address them.

Zhengyu Zhao, Zhuoran Liu, Martha Larson

Deep Neural Networks have been shown to be vulnerable to adversarial images. Conventional attacks strive for indistinguishable adversarial images with strictly restricted perturbations. Recently, researchers have moved to explore distinguishable yet non-suspicious adversarial images and demonstrated that color transformation attacks are effective. In this work, we propose Adversarial Color Filter (AdvCF), a novel color transformation attack that is optimized with gradient information in the parameter space of a simple color filter. In particular, our color filter space is explicitly specified so that we are able to provide a systematic analysis of model robustness against adversarial color transformations, from both the attack and defense perspectives. In contrast, existing color transformation attacks do not offer the opportunity for systematic analysis due to the lack of such an explicit space. We further demonstrate the effectiveness of our AdvCF in fooling image classifiers and also compare it with other color transformation attacks regarding their robustness to defenses and image acceptability through an extensive user study. We also highlight the human-interpretability of AdvCF and show its superiority over the state-of-the-art human-interpretable color transformation attack on both image acceptability and efficiency. Additional results provide interesting new insights into model robustness against AdvCF in another three visual tasks.

Zhuoran Liu, Martha Larson

E-commerce platforms provide their customers with ranked lists of recommended items matching the customers' preferences. Merchants on e-commerce platforms would like their items to appear as high as possible in the top-N of these ranked lists. In this paper, we demonstrate how unscrupulous merchants can create item images that artificially promote their products, improving their rankings. Recommender systems that use images to address the cold start problem are vulnerable to this security risk. We describe a new type of attack, Adversarial Item Promotion (AIP), that strikes directly at the core of Top-N recommenders: the ranking mechanism itself. Existing work on adversarial images in recommender systems investigates the implications of conventional attacks, which target deep learning classifiers. In contrast, our AIP attacks are embedding attacks that seek to push features representations in a way that fools the ranker (not a classifier) and directly lead to item promotion. We introduce three AIP attacks insider attack, expert attack, and semantic attack, which are defined with respect to three successively more realistic attack models. Our experiments evaluate the danger of these attacks when mounted against three representative visually-aware recommender algorithms in a framework that uses images to address cold start. We also evaluate potential defenses, including adversarial training and find that common, currently-existing, techniques do not eliminate the danger of AIP attacks. In sum, we show that using images to address cold start opens recommender systems to potential threats with clear practical implications.

Zhengyu Zhao, Zhuoran Liu, Martha Larson

The success of image perturbations that are designed to fool image classifier is assessed in terms of both adversarial effect and visual imperceptibility. The conventional assumption on imperceptibility is that perturbations should strive for tight $L_p$-norm bounds in RGB space. In this work, we drop this assumption by pursuing an approach that exploits human color perception, and more specifically, minimizing perturbation size with respect to perceptual color distance. Our first approach, Perceptual Color distance C&W (PerC-C&W), extends the widely-used C&W approach and produces larger RGB perturbations. PerC-C&W is able to maintain adversarial strength, while contributing to imperceptibility. Our second approach, Perceptual Color distance Alternating Loss (PerC-AL), achieves the same outcome, but does so more efficiently by alternating between the classification loss and perceptual color difference when updating perturbations. Experimental evaluation shows PerC approaches outperform conventional $L_p$ approaches in terms of robustness and transferability, and also demonstrates that the PerC distance can provide added value on top of existing structure-based methods to creating image perturbations.

Daanish Ali Khan, Linhong Li, Ninghao Sha et al.

Recent breakthroughs in the field of deep learning have led to advancements in a broad spectrum of tasks in computer vision, audio processing, natural language processing and other areas. In most instances where these tasks are deployed in real-world scenarios, the models used in them have been shown to be susceptible to adversarial attacks, making it imperative for us to address the challenge of their adversarial robustness. Existing techniques for adversarial robustness fall into three broad categories: defensive distillation techniques, adversarial training techniques, and randomized or non-deterministic model based techniques. In this paper, we propose a novel neural network paradigm that falls under the category of randomized models for adversarial robustness, but differs from all existing techniques under this category in that it models each parameter of the network as a statistical distribution with learnable parameters. We show experimentally that this framework is highly robust to a variety of white-box and black-box adversarial attacks, while preserving the task-specific performance of the traditional neural network model.

Zhuoran Liu, Zhengyu Zhao, Martha Larson

An adversarial query is an image that has been modified to disrupt content-based image retrieval (CBIR) while appearing nearly untouched to the human eye. This paper presents an analysis of adversarial queries for CBIR based on neural, local, and global features. We introduce an innovative neural image perturbation approach, called Perturbations for Image Retrieval Error (PIRE), that is capable of blocking neural-feature-based CBIR. PIRE differs significantly from existing approaches that create images adversarial with respect to CNN classifiers because it is unsupervised, i.e., it needs no labelled data from the data set to which it is applied. Our experimental analysis demonstrates the surprising effectiveness of PIRE in blocking CBIR, and also covers aspects of PIRE that must be taken into account in practical settings, including saving images, image quality and leaking adversarial queries into the background collection. Our experiments also compare PIRE (a neural approach) with existing keypoint removal and injection approaches (which modify local features). Finally, we discuss the challenges that face multimedia researchers in the future study of adversarial queries.

Zhuoran Liu, Yang Liu

Identifying and correcting grammatical errors in the text written by non-native writers has received increasing attention in recent years. Although a number of annotated corpora have been established to facilitate data-driven grammatical error detection and correction approaches, they are still limited in terms of quantity and coverage because human annotation is labor-intensive, time-consuming, and expensive. In this work, we propose to utilize unlabeled data to train neural network based grammatical error detection models. The basic idea is to cast error detection as a binary classification problem and derive positive and negative training examples from unlabeled data. We introduce an attention-based neural network to capture long-distance dependencies that influence the word being detected. Experiments show that the proposed approach significantly outperforms SVMs and convolutional networks with fixed-size context window.