Abstract In the past few years, the popularity and wide use of video conferencing software enjoyed exponential growth in market size. This technology enables participants in different geographic regions to have a virtual face-to-face meeting. Additionally, it allows participants to utilize virtual backgrounds to hide their real environment with privacy concerns or to reduce distractions, particularly in professional settings. In scenarios where the users should not hide their actual locations, they may mislead other participants into assuming that the displayed virtual backgrounds are real. In this paper, we propose a new publicly-available dataset of virtual and real backgrounds in video conferencing software (e.g., Zoom, Google Meet, Microsoft Teams). The presented archive was evaluated by an exhaustive series of tests and scenarios using two well-known features extraction methods: CRSPAM1372 and six co-mat. The first verification scenario considers the case where the detector is unaware of manipulated frames (i.e., the forensically-edited frames are not part of the training set). A model trained on zoom frames that were tested with Google Meet frames can detect real background images from virtual ones in video conferencing software with 99.80% detection accuracy. Furthermore, it is possible to distinguish virtual from real backgrounds in videos created for videoconferencing software at a high detection rate of approximately 99.80%. According to our conclusions, the proposed method greatly enhanced the detection accuracy and resistance against diverse adversarial conditions, making it a reliable technique for classifying actual as opposed to virtual backgrounds in video communications. Given the described dataset provided and some preliminary experiments that we performed, we expect that it will lead to more future research in this domain.

Federated Learning is an approach that enables multiple devices to collectively train a shared model without sharing raw data, thereby preserving data privacy. However, federated learning systems are vulnerable to data-poisoning attacks during the training and updating stages. Three data-poisoning attacks–label flipping, feature poisoning, and VagueGAN–are tested on FL models across one out of ten clients using the CIC and UNSW datasets. For label flipping, we randomly modify labels of benign data; for feature poisoning, we alter highly influential features identified by the Random Forest technique; and for VagueGAN, we generate adversarial examples using Generative Adversarial Networks. Adversarial samples constitute a small portion of each dataset. In this study, we vary the percentages by which adversaries can modify datasets to observe their impact on the Client and Server sides. Experimental findings indicate that label flipping and VagueGAN attacks do not significantly affect server accuracy, as they are easily detectable by the Server. In contrast, feature poisoning attacks subtly undermine model performance while maintaining high accuracy and attack success rates, highlighting their subtlety and effectiveness. Therefore, feature poisoning attacks manipulate the server without causing a significant decrease in model accuracy, underscoring the vulnerability of federated learning systems to such sophisticated attacks. To mitigate these vulnerabilities, we explore a recent defensive approach known as Random Deep Feature Selection, which randomizes server features with varying sizes (e.g., 50 and 400) during training. This strategy has proven highly effective in minimizing the impact of such attacks, particularly on feature poisoning.

Machine learning and deep learning are transformative forces reshaping our networks, industries, services, and ways of life. However, the susceptibility of these intelligent systems to adversarial attacks remains a significant issue. On the one hand, recent studies have demonstrated the potential transferability of adversarial attacks across diverse models. On the other hand, existing defense mechanisms are vulnerable to advanced attacks or are often limited to certain attack types. This study proposes a random deep feature selection approach to mitigate such transferability and improve the robustness of models against adversarial manipulations. Our approach is designed to strengthen deep models against poisoning (e.g., label flipping) and exploratory (e.g., DeepFool, BIM, FGSM, I-FGSM, L-BFGS, C&W, JSMA, and PGD) attacks that are applied in both the training and testing stages, and Transfer Learning-Based Adversarial Attacks. We consider scenarios involving perfect and semi-knowledgeable attackers. The performance of our approach is evaluated through extensive experiments on the renowned UNSW-NB15 dataset, including both real-world and synthetic data, covering a wide range of modern attack behaviors and benign activities. The results indicate that our approach boosts the effectiveness of the target network to over 80% against labelflipping poisoning attacks and over 60% against all major types of exploratory attacks.

In smart electric grid systems, various sensors and Internet of Things (IoT) devices are used to collect electrical data at substations. In a traditional system, a multitude of energy-related data from substations needs to be migrated to central storage, such as Cloud or edge devices, for knowledge extraction that might impose severe data misuse, data manipulation, or privacy leakage. This motivates to propose anomaly detection system to detect threats and Federated Learning to resolve the issues of data silos and privacy of data. In this article, we present a framework to identify anomalies in industrial data that are gathered from the remote terminal devices deployed at the substations in the smart electric grid system. The anomaly detection system is based on Long Short-Term Memory (LSTM) and autoencoders that employs Mean Standard Deviation (MSD) and Median Absolute Deviation (MAD) approaches for detecting anomalies. We deploy Federated Learning (FL) to preserve the privacy of the data generated by the substations. FL enables energy providers to train shared AI models cooperatively without disclosing the data to the server. In order to further enhance the security and privacy properties of the proposed framework, we implemented homomorphic encryption based on the Paillier algorithm for preserving data privacy. The proposed security model performs better with MSD approach using HE-128 bit key providing 97% F1-score and 98% accuracy for K=5 with low computation overhead as compared with HE-256 bit key.

Malicious URLs present significant threats to businesses, such as transportation and banking, causing disruptions in business operations. It is essential to identify these URLs; however, existing Machine Learning models are vulnerable to backdoor attacks. These attacks involve manipulating a small portion of the training data labels, such as Label Flipping, which can lead to misclassification. Therefore, it is crucial to incorporate defense mechanisms into machine-learning models to protect against such attacks. The focus of this study is on backdoor attacks in the context of URL detection using ensemble trees. By illuminating the motivations behind such attacks, highlighting the roles of attackers, and emphasizing the critical importance of effective defense strategies, this paper contributes to the ongoing efforts to fortify machine-learning models against adversarial threats within the machine-learning domain in network security. We propose an innovative alarm system that detects the presence of poisoned labels and a defense mechanism designed to uncover the original class labels with the aim of mitigating backdoor attacks on ensemble tree classifiers. We conducted a case study using the Alexa and Phishing Site URL datasets and showed that label-flipping attacks can be addressed using our proposed defense mechanism. Our experimental results prove that the Label Flipping attack achieved an Attack Success Rate between 50-65% within 2-5%, and the innovative defense method successfully detected poisoned labels with an accuracy of up to 100%.

In the past decades, the rise of artificial intelligence has given us the capabilities to solve the most challenging problems in our day-to-day lives, such as cancer prediction and autonomous navigation. However, these applications might not be reliable if not secured against adversarial attacks. In addition, recent works demonstrated that some adversarial examples are transferable across different models. Therefore, it is crucial to avoid such transferability via robust models that resist adversarial manipulations. In this paper, we propose a feature randomization-based approach that resists eight adversarial attacks targeting deep learning models in the testing phase. Our novel approach consists of changing the training strategy in the target network classifier and selecting random feature samples. We consider the attacker with a Limited-Knowledge and Semi-Knowledge conditions to undertake the most prevalent types of adversarial attacks. We evaluate the robustness of our approach using the well-known UNSW-NB15 datasets that include realistic and synthetic attacks. Afterward, we demonstrate that our strategy outperforms the existing state-of-the-art approach, such as the Most Powerful Attack, which consists of fine-tuning the network model against specific adversarial attacks. Further, we demonstrate the practicality of our approach using the VIPPrint dataset through a comprehensive set of experiments. Finally, our experimental results show that our methodology can secure the target network and resists adversarial attack transferability by over 60%.

Context: Speech Emotion Recognition (SER) is a valuable technology that identifies human emotions from spoken language, enabling the development of context-aware and personalized intelligent systems.To protect user privacy, Federated Learning (FL) has been introduced, enabling local training of models on user devices.However, FL raises concerns about the potential exposure of sensitive information from local model parameters, which is especially critical in applications like SER that involve personal voice data.Local Differential Privacy (LDP) has prevented privacy leaks in image and video data.However, it encounters notable accuracy degradation when applied to speech data, especially in the presence of high noise levels.In this paper, we propose an approach called LDP-FL with CSS, which combines LDP with a novel client selection strategy (CSS).By leveraging CSS, we aim to improve the representatives of updates and mitigate the adverse effects of noise on SER accuracy while ensuring client privacy through LDP.Furthermore, we conducted model inversion attacks to evaluate the robustness of LDP-FL in preserving privacy.These attacks involved an adversary attempting to reconstruct individuals' voice samples using the output labels provided by the SER model.The evaluation results reveal that LDP-FL with CSS achieved an accuracy of 65-70%, which is 4% lower than the initial SER model accuracy.Furthermore, LDP-FL demonstrated exceptional resilience against model inversion attacks, outperforming the non-LDP method by a factor of 10.Overall, our analysis emphasizes the importance of achieving a balance between privacy and accuracy in accordance with the requirements of the SER application.

Malicious advertisement URLs pose a security risk since they are the source of cyber-attacks, and the need to address this issue is growing in both industry and academia. Several attempts have been made in recent years for malicious URL detection using machine learning (ML). The most widely used techniques extract linguistic features of URL string to extract features like bag-of-words (BoW) before applying ML model. Existing malicious URL detection techniques require effective manual feature engineering that can handle unseen features and generalise to test data. In this study, we extract a novel set of lexical and Web-scrapped features and employ ML techniques for fraudulent advertisement URL detection. The combination set of six different kinds of features precisely overcomes the obfuscation in fraudulent URL classification. Based on distinct statistical properties, we use twelve differently formatted datasets for detection, prediction and classification task. We extend our prediction analysis for mismatched and unlabelled datasets. For this framework, we analyze the performance of four ML techniques: Random Forest, Gradient Boost, XGBoost and AdaBoost in the detection part. With our proposed method, we achieve a false negative rate up to 0.0037 while maintaining high detection accuracy of 99.63%. Moreover, we employ an unsupervised learning technique for data clustering using the${K}$-Means algorithm for the visual analysis. This paper analyses the vulnerability of decision tree-based models using the limited knowledge attack scenario. We considered the exploratory attack during the test phase and implemented Zeroth Order Optimization adversarial attack on the detection models.

In the past few years, Convolutional Neural Networks (CNN) have demonstrated promising performance in various real-world cybersecurity applications, such as network and multimedia security. However, the underlying fragility of CNN structures poses major security problems, making them inappropriate for use in security-oriented applications, including computer networks. Protecting these architectures from adversarial attacks necessitates using security-wise architectures that are challenging to attack. In this study, we present a novel architecture based on an ensemble classifier that combines the enhanced security of 1-Class classification (known as 1C) with the high performance of conventional 2-Class classification (known as 2C) in the absence of attacks. Our architecture is referred to as the 1.5-Class (cmb-classifier) classifier and is constructed using a final dense classifier, one 2C classifier (i.e., CNNs), and two parallel 1C classifiers (i.e., auto-encoders). In our experiments, we evaluated the robustness of our proposed architecture by considering eight possible adversarial attacks in various scenarios. We performed these attacks on the 2C and cmb-classifier architectures separately. The experimental results of our study showed that the Attack Success Rate (ASR) of the I-FGSM attack against a 2C classifier trained with the N-BaIoT dataset is 0.9900. In contrast, the ASR is 0.0000 for the cmb-classifier.

Convolutional Neural Networks (CNNs) models are one of the most frequently used deep learning networks, and extensively used in both academia and industry. Recent studies demonstrated that adversarial attacks against such models can maintain their effectiveness even when used on models other than the one targeted by the attacker. This major property is known as transferability, and makes CNNs ill-suited for security applications. In this paper, we provide the first comprehensive study which assesses the robustness of CNN-based models for computer networks against adversarial transferability. Furthermore, we investigate whether the transferability property issue holds in computer networks applications. In our experiments, we first consider five different attacks: the Iterative Fast Gradient Method (I-FGSM), the Jacobian-based Saliency Map (JSMA), the Limited-memory Broyden Fletcher Goldfarb Shanno BFGS (L-BFGS), the Projected Gradient Descent (PGD), and the DeepFool attack. Then, we perform these attacks against three well-known datasets: the Network-based Detection of IoT (N-BaIoT) dataset, the Domain Generating Algorithms (DGA) dataset, and the RIPE Atlas dataset. Our experimental results show clearly that the transferability happens in specific use cases for the I-FGSM, the JSMA, and the LBFGS attack. In such scenarios, the attack success rate on the target network range from 63.00% to 100%. Finally, we suggest two shielding strategies to hinder the attack transferability, by considering the Most Powerful Attacks (MPAs), and the mismatch LSTM architecture.

We investigate if the random feature selection approach proposed in [1] to improve the robustness of forensic detectors to targeted attacks, can be extended to detectors based on deep learning features. In particular, we study the transferability of adversarial examples targeting an original CNN image manipulation detector to other detectors (a fully connected neural network and a linear SVM) that rely on a random subset of the features extracted from the flatten layer of the original network. The results we got by considering three image manipulation detection tasks (resizing, median filtering and adaptive histogram equalization), two original network architectures and three classes of attacks, show that feature randomization helps to hinder attack transferability, even if, in some cases, simply changing the architecture of the detector, or even retraining the detector is enough to prevent the transferability of the attacks.

Recent studies have shown that Convolutional Neural Networks (CNN) are relatively easy to attack through the generation of so called adversarial examples. Such vulnerability also affects CNN-based image forensic tools. Research in deep learning has shown that adversarial examples exhibit a certain degree of transferability, i.e., they maintain part of their effectiveness even against CNN models other than the one targeted by the attack. This is a very strong property undermining the usability of CNN's in security-oriented applications. In this paper, we investigate if attack transferability also holds in image forensics applications. With specific reference to the case of manipulation detection, we analyse the results of several experiments considering different sources of mismatch between the CNN used to build the adversarial examples and the one adopted by the forensic analyst. The analysis ranges from cases in which the mismatch involves only the training dataset, to cases in which the attacker and the forensic analyst adopt different architectures. The results of our experiments show that, in the majority of the cases, the attacks are not transferable, thus easing the design of proper countermeasures at least when the attacker does not have a perfect knowledge of the target detector.

Detection of contrast adjustments in the presence of JPEG post processing is known to be a challenging task. JPEG post processing is often applied innocently, as JPEG is the most common image format, or it may correspond to a laundering attack, when it is purposely applied to erase the traces of manipulation. In this paper, we propose a CNN-based detector for generic contrast adjustment, which is robust to JPEG compression. The proposed system relies on a patch-based Convolutional Neural Network (CNN), trained to distinguish pristine images from contrast adjusted images, for some selected adjustment operators of different nature. Robustness to JPEG compression is achieved by training the CNN with JPEG examples, compressed over a range of Quality Factors (QFs). Experimental results show that the detector works very well and scales well with respect to the adjustment type, yielding very good performance under a large variety of unseen tonal adjustments.