arXiv subjects

Stjepan Picek

Publications and source records attributed to Stjepan Picek.

At least 73 records · Page 4Linked to original sources

arXiv2023

A New Angle: On Evolving Rotation Symmetric Boolean Functions

Rotation symmetric Boolean functions represent an interesting class of Boolean functions as they are relatively rare compared to general Boolean functions. At the same time, the functions in this class can have excellent properties, making them interesting for various practical applications. The usage of metaheuristics to construct rotation symmetric Boolean functions is a direction that has been explored for almost twenty years. Despite that, there are very few results considering evolutionary computation methods. This paper uses several evolutionary algorithms to evolve rotation symmetric Boolean functions with different properties. Despite using generic metaheuristics, we obtain results that are competitive with prior work relying on customized heuristics. Surprisingly, we find that bitstring and floating point encodings work better than the tree encoding. Moreover, evolving highly nonlinear general Boolean functions is easier than rotation symmetric ones.

cs.NE↗

arXiv2023

Look into the Mirror: Evolving Self-Dual Bent Boolean Functions

Bent Boolean functions are important objects in cryptography and coding theory, and there are several general approaches for constructing such functions. Metaheuristics proved to be a strong choice as they can provide many bent functions, even when the size of the Boolean function is large (e.g., more than 20 inputs). While bent Boolean functions represent only a small part of all Boolean functions, there are several subclasses of bent functions providing specific properties and challenges. One of the most interesting subclasses comprises (anti-)self-dual bent Boolean functions. This paper provides a detailed experimentation with evolutionary algorithms with the goal of evolving (anti-)self-dual bent Boolean functions. We experiment with two encodings and two fitness functions to directly evolve self-dual bent Boolean functions. Our experiments consider Boolean functions with sizes of up to 16 inputs, and we successfully construct self-dual bent functions for each dimension. Moreover, when comparing with the evolution of bent Boolean functions, we notice that the difficulty for evolutionary algorithms is rather similar. Finally, we also tried evolving secondary constructions for self-dual bent functions, but this direction provided no successful results.

cs.NE↗

arXiv2023

MUDGUARD: Taming Malicious Majorities in Federated Learning using Privacy-Preserving Byzantine-Robust Clustering

Byzantine-robust Federated Learning (FL) aims to counter malicious clients and train an accurate global model while maintaining an extremely low attack success rate. Most existing systems, however, are only robust when most of the clients are honest. FLTrust (NDSS '21) and Zeno++ (ICML '20) do not make such an honest majority assumption but can only be applied to scenarios where the server is provided with an auxiliary dataset used to filter malicious updates. FLAME (USENIX '22) and EIFFeL (CCS '22) maintain the semi-honest majority assumption to guarantee robustness and the confidentiality of updates. It is therefore currently impossible to ensure Byzantine robustness and confidentiality of updates without assuming a semi-honest majority. To tackle this problem, we propose a novel Byzantine-robust and privacy-preserving FL system, called MUDGUARD, that can operate under malicious minority \emph{or majority} in both the server and client sides. Based on DBSCAN, we design a new method for extracting features from model updates via pairwise adjusted cosine similarity to boost the accuracy of the resulting clustering. To thwart attacks from a malicious majority, we develop a method called \textit{Model Segmentation}, that aggregates together only the updates from within a cluster, sending the corresponding model only to the clients of the corresponding cluster. The fundamental idea is that even if malicious clients are in their majority, their poisoned updates cannot harm benign clients if they are confined only within the malicious cluster. We also leverage multiple cryptographic tools to conduct clustering without sacrificing training correctness and updates confidentiality. We present a detailed security proof and empirical evaluation along with a convergence analysis for MUDGUARD.

cs.CR↗

arXiv2023

Momentum Gradient-based Untargeted Attack on Hypergraph Neural Networks

Hypergraph Neural Networks (HGNNs) have been successfully applied in various hypergraph-related tasks due to their excellent higher-order representation capabilities. Recent works have shown that deep learning models are vulnerable to adversarial attacks. Most studies on graph adversarial attacks have focused on Graph Neural Networks (GNNs), and the study of adversarial attacks on HGNNs remains largely unexplored. In this paper, we try to reduce this gap. We design a new HGNNs attack model for the untargeted attack, namely MGHGA, which focuses on modifying node features. We consider the process of HGNNs training and use a surrogate model to implement the attack before hypergraph modeling. Specifically, MGHGA consists of two parts: feature selection and feature modification. We use a momentum gradient mechanism to choose the attack node features in the feature selection module. In the feature modification module, we use two feature generation approaches (direct modification and sign gradient) to enable MGHGA to be employed on discrete and continuous datasets. We conduct extensive experiments on five benchmark datasets to validate the attack performance of MGHGA in the node and the visual object classification tasks. The results show that MGHGA improves performance by an average of 2% compared to the than the baselines.

cs.LG↗

arXiv2023

Invisible Threats: Backdoor Attack in OCR Systems

Optical Character Recognition (OCR) is a widely used tool to extract text from scanned documents. Today, the state-of-the-art is achieved by exploiting deep neural networks. However, the cost of this performance is paid at the price of system vulnerability. For instance, in backdoor attacks, attackers compromise the training phase by inserting a backdoor in the victim's model that will be activated at testing time by specific patterns while leaving the overall model performance intact. This work proposes a backdoor attack for OCR resulting in the injection of non-readable characters from malicious input images. This simple but effective attack exposes the state-of-the-art OCR weakness, making the extracted text correct to human eyes but simultaneously unusable for the NLP application that uses OCR as a preprocessing step. Experimental results show that the attacked models successfully output non-readable characters for around 90% of the poisoned instances without harming their performance for the remaining instances.

cs.CR↗

arXiv2023

Universal Soldier: Using Universal Adversarial Perturbations for Detecting Backdoor Attacks

Deep learning models achieve excellent performance in numerous machine learning tasks. Yet, they suffer from security-related issues such as adversarial examples and poisoning (backdoor) attacks. A deep learning model may be poisoned by training with backdoored data or by modifying inner network parameters. Then, a backdoored model performs as expected when receiving a clean input, but it misclassifies when receiving a backdoored input stamped with a pre-designed pattern called "trigger". Unfortunately, it is difficult to distinguish between clean and backdoored models without prior knowledge of the trigger. This paper proposes a backdoor detection method by utilizing a special type of adversarial attack, universal adversarial perturbation (UAP), and its similarities with a backdoor trigger. We observe an intuitive phenomenon: UAPs generated from backdoored models need fewer perturbations to mislead the model than UAPs from clean models. UAPs of backdoored models tend to exploit the shortcut from all classes to the target class, built by the backdoor trigger. We propose a novel method called Universal Soldier for Backdoor detection (USB) and reverse engineering potential backdoor triggers via UAPs. Experiments on 345 models trained on several datasets show that USB effectively detects the injected backdoor and provides comparable or better results than state-of-the-art methods.

cs.LG↗

arXiv2023

FLAIRS: FPGA-Accelerated Inference-Resistant & Secure Federated Learning

Federated Learning (FL) has become very popular since it enables clients to train a joint model collaboratively without sharing their private data. However, FL has been shown to be susceptible to backdoor and inference attacks. While in the former, the adversary injects manipulated updates into the aggregation process; the latter leverages clients' local models to deduce their private data. Contemporary solutions to address the security concerns of FL are either impractical for real-world deployment due to high-performance overheads or are tailored towards addressing specific threats, for instance, privacy-preserving aggregation or backdoor defenses. Given these limitations, our research delves into the advantages of harnessing the FPGA-based computing paradigm to overcome performance bottlenecks of software-only solutions while mitigating backdoor and inference attacks. We utilize FPGA-based enclaves to address inference attacks during the aggregation process of FL. We adopt an advanced backdoor-aware aggregation algorithm on the FPGA to counter backdoor attacks. We implemented and evaluated our method on Xilinx VMK-180, yielding a significant speed-up of around 300 times on the IoT-Traffic dataset and more than 506 times on the CIFAR-10 dataset.

cs.CR↗

arXiv2023

IB-RAR: Information Bottleneck as Regularizer for Adversarial Robustness

In this paper, we propose a novel method, IB-RAR, which uses Information Bottleneck (IB) to strengthen adversarial robustness for both adversarial training and non-adversarial-trained methods. We first use the IB theory to build regularizers as learning objectives in the loss function. Then, we filter out unnecessary features of intermediate representation according to their mutual information (MI) with labels, as the network trained with IB provides easily distinguishable MI for its features. Experimental results show that our method can be naturally combined with adversarial training and provides consistently better accuracy on new adversarial examples. Our method improves the accuracy by an average of 3.07% against five adversarial attacks for the VGG16 network, trained with three adversarial training benchmarks and the CIFAR-10 dataset. In addition, our method also provides good robustness for undefended methods, such as training with cross-entropy loss only. Finally, in the absence of adversarial training, the VGG16 network trained using our method and the CIFAR-10 dataset reaches an accuracy of 35.86% against PGD examples, while using all layers reaches 25.61% accuracy.

cs.LG↗

arXiv2023

On Feasibility of Server-side Backdoor Attacks on Split Learning

Split learning is a collaborative learning design that allows several participants (clients) to train a shared model while keeping their datasets private. Recent studies demonstrate that collaborative learning models, specifically federated learning, are vulnerable to security and privacy attacks such as model inference and backdoor attacks. Backdoor attacks are a group of poisoning attacks in which the attacker tries to control the model output by manipulating the model's training process. While there have been studies regarding inference attacks on split learning, it has not yet been tested for backdoor attacks. This paper performs a novel backdoor attack on split learning and studies its effectiveness. Despite traditional backdoor attacks done on the client side, we inject the backdoor trigger from the server side. For this purpose, we provide two attack methods: one using a surrogate client and another using an autoencoder to poison the model via incoming smashed data and its outgoing gradient toward the innocent participants. We did our experiments using three model architectures and three publicly available datasets in the image domain and ran a total of 761 experiments to evaluate our attack methods. The results show that despite using strong patterns and injection methods, split learning is highly robust and resistant to such poisoning attacks. While we get the attack success rate of 100% as our best result for the MNIST dataset, in most of the other cases, our attack shows little success when increasing the cut layer.

cs.CR↗

arXiv2023

Turning Privacy-preserving Mechanisms against Federated Learning

Recently, researchers have successfully employed Graph Neural Networks (GNNs) to build enhanced recommender systems due to their capability to learn patterns from the interaction between involved entities. In addition, previous studies have investigated federated learning as the main solution to enable a native privacy-preserving mechanism for the construction of global GNN models without collecting sensitive data into a single computation unit. Still, privacy issues may arise as the analysis of local model updates produced by the federated clients can return information related to sensitive local data. For this reason, experts proposed solutions that combine federated learning with Differential Privacy strategies and community-driven approaches, which involve combining data from neighbor clients to make the individual local updates less dependent on local sensitive data. In this paper, we identify a crucial security flaw in such a configuration, and we design an attack capable of deceiving state-of-the-art defenses for federated learning. The proposed attack includes two operating modes, the first one focusing on convergence inhibition (Adversarial Mode), and the second one aiming at building a deceptive rating injection on the global federated model (Backdoor Mode). The experimental results show the effectiveness of our attack in both its modes, returning on average 60% performance detriment in all the tests on Adversarial Mode and fully effective backdoors in 93% of cases for the tests performed on Backdoor Mode.

cs.LG↗

arXiv2023

Going In Style: Audio Backdoors Through Stylistic Transformations

This work explores stylistic triggers for backdoor attacks in the audio domain: dynamic transformations of malicious samples through guitar effects. We first formalize stylistic triggers - currently missing in the literature. Second, we explore how to develop stylistic triggers in the audio domain by proposing JingleBack. Our experiments confirm the effectiveness of the attack, achieving a 96% attack success rate. Our code is available in https://github.com/skoffas/going-in-style.

cs.CR↗

arXiv2023

SoK: A Systematic Evaluation of Backdoor Trigger Characteristics in Image Classification

Deep learning achieves outstanding results in many machine learning tasks. Nevertheless, it is vulnerable to backdoor attacks that modify the training set to embed a secret functionality in the trained model. The modified training samples have a secret property, i. e., a trigger. At inference time, the secret functionality is activated when the input contains the trigger, while the model functions correctly in other cases. While there are many known backdoor attacks (and defenses), deploying a stealthy attack is still far from trivial. Successfully creating backdoor triggers depends on numerous parameters. Unfortunately, research has not yet determined which parameters contribute most to the attack performance. This paper systematically analyzes the most relevant parameters for the backdoor attacks, i.e., trigger size, position, color, and poisoning rate. Using transfer learning, which is very common in computer vision, we evaluate the attack on state-of-the-art models (ResNet, VGG, AlexNet, and GoogLeNet) and datasets (MNIST, CIFAR10, and TinyImageNet). Our attacks cover the majority of backdoor settings in research, providing concrete directions for future works. Our code is publicly available to facilitate the reproducibility of our results.

cs.CV↗

arXiv2023

Rethinking the Trigger-injecting Position in Graph Backdoor Attack

Backdoor attacks have been demonstrated as a security threat for machine learning models. Traditional backdoor attacks intend to inject backdoor functionality into the model such that the backdoored model will perform abnormally on inputs with predefined backdoor triggers and still retain state-of-the-art performance on the clean inputs. While there are already some works on backdoor attacks on Graph Neural Networks (GNNs), the backdoor trigger in the graph domain is mostly injected into random positions of the sample. There is no work analyzing and explaining the backdoor attack performance when injecting triggers into the most important or least important area in the sample, which we refer to as trigger-injecting strategies MIAS and LIAS, respectively. Our results show that, generally, LIAS performs better, and the differences between the LIAS and MIAS performance can be significant. Furthermore, we explain these two strategies' similar (better) attack performance through explanation techniques, which results in a further understanding of backdoor attacks in GNNs.

cs.LG↗

arXiv2023

Sniper Backdoor: Single Client Targeted Backdoor Attack in Federated Learning

Federated Learning (FL) enables collaborative training of Deep Learning (DL) models where the data is retained locally. Like DL, FL has severe security weaknesses that the attackers can exploit, e.g., model inversion and backdoor attacks. Model inversion attacks reconstruct the data from the training datasets, whereas backdoors misclassify only classes containing specific properties, e.g., a pixel pattern. Backdoors are prominent in FL and aim to poison every client model, while model inversion attacks can target even a single client. This paper introduces a novel technique to allow backdoor attacks to be client-targeted, compromising a single client while the rest remain unchanged. The attack takes advantage of state-of-the-art model inversion and backdoor attacks. Precisely, we leverage a Generative Adversarial Network to perform the model inversion. Afterward, we shadow-train the FL network, in which, using a Siamese Neural Network, we can identify, target, and backdoor the victim's model. Our attack has been validated using the MNIST, F-MNIST, EMNIST, and CIFAR-100 datasets under different settings -- achieving up to 99\% accuracy on both source (clean) and target (backdoor) classes and against state-of-the-art defenses, e.g., Neural Cleanse, opening a novel threat model to be considered in the future.

cs.CR↗

arXiv2023

Digging Deeper: Operator Analysis for Optimizing Nonlinearity of Boolean Functions

Boolean functions are mathematical objects with numerous applications in domains like coding theory, cryptography, and telecommunications. Finding Boolean functions with specific properties is a complex combinatorial optimization problem where the search space grows super-exponentially with the number of input variables. One common property of interest is the nonlinearity of Boolean functions. Constructing highly nonlinear Boolean functions is difficult as it is not always known what nonlinearity values can be reached in practice. In this paper, we investigate the effects of the genetic operators for bit-string encoding in optimizing nonlinearity. While several mutation and crossover operators have commonly been used, the link between the genotype they operate on and the resulting phenotype changes is mostly obscure. By observing the range of possible changes an operator can provide, as well as relative probabilities of specific transitions in the objective space, one can use this information to design a more effective combination of genetic operators. The analysis reveals interesting insights into operator effectiveness and indicates how algorithm design may improve convergence compared to an operator-agnostic genetic algorithm.

cs.NE↗

arXiv2023

A Survey of Metaheuristic Algorithms for the Design of Cryptographic Boolean Functions

Boolean functions are mathematical objects used in diverse domains and have been actively researched for several decades already. One domain where Boolean functions play an important role is cryptography. There, the plethora of settings one should consider and cryptographic properties that need to be fulfilled makes the search for new Boolean functions still a very active domain. There are several options to construct appropriate Boolean functions: algebraic constructions, random search, and metaheuristics. In this work, we concentrate on metaheuristic approaches and examine the related works appearing in the last 25 years. To the best of our knowledge, this is the first survey work on this topic. Additionally, we provide a new taxonomy of related works and discuss the results obtained. Finally, we finish this survey with potential future research directions.

cs.CR↗

arXiv2022

On the Evolution of Boomerang Uniformity in Cryptographic S-boxes

S-boxes are an important primitive that help cryptographic algorithms to be resilient against various attacks. The resilience against specific attacks can be connected with a certain property of an S-box, and the better the property value, the more secure the algorithm. One example of such a property is called boomerang uniformity, which helps to be resilient against boomerang attacks. How to construct S-boxes with good boomerang uniformity is not always clear. There are algebraic techniques that can result in good boomerang uniformity, but the results are still rare. In this work, we explore the evolution of S-boxes with good values of boomerang uniformity. We consider three different encodings and five S-box sizes. For sizes $4\times 4$ and $5\times 5$, we manage to obtain optimal solutions. For $6\times 6$, we obtain optimal boomerang uniformity for the non-APN function. For larger sizes, the results indicate the problem to be very difficult (even more difficult than evolving differential uniformity, which can be considered a well-researched problem).

cs.NE↗

arXiv2022

Evolutionary Strategies for the Design of Binary Linear Codes

The design of binary error-correcting codes is a challenging optimization problem with several applications in telecommunications and storage, which has also been addressed with metaheuristic techniques and evolutionary algorithms. Still, all these efforts focused on optimizing the minimum distance of unrestricted binary codes, i.e., with no constraints on their linearity, which is a desirable property for efficient implementations. In this paper, we present an Evolutionary Strategy (ES) algorithm that explores only the subset of linear codes of a fixed length and dimension. To that end, we represent the candidate solutions as binary matrices and devise variation operators that preserve their ranks. Our experiments show that up to length $n=14$, our ES always converges to an optimal solution with a full success rate, and the evolved codes are all inequivalent to the Best-Known Linear Code (BKLC) given by MAGMA. On the other hand, for larger lengths, both the success rate of the ES as well as the diversity of the evolved codes start to drop, with the extreme case of $(16,8,5)$ codes which all turn out to be equivalent to MAGMA's BKLC.

cs.NE↗