ZDBERTa: Advancing Zero-Day Cyberattack Detection in Internet of Vehicle with Zero-Shot Learning

We investigate learning feature-to-feature translator networks by alternating back-propagation as a general-purpose solution to zero-shot learning (ZSL) problems. It is a generative model-based ZSL framework. In contrast to models based on generative adversarial networks (GAN) or variational autoencoders (VAE) that require auxiliary networks to assist the training, our model consists of a single conditional generator that maps class-level semantic features and Gaussian white noise vectors accounting for instance-level latent factors to visual features, and is trained by maximum likelihood estimation. The training process is a simple yet effective alternating back-propagation process that iterates the following two steps: (i) the inferential back-propagation to infer the latent noise vector of each observed example, and (ii) the learning back-propagation to update the model parameters. We show that, with slight modifications, our model is capable of learning from incomplete visual features for ZSL. We conduct extensive comparisons with existing generative ZSL methods on five benchmarks, demonstrating the superiority of our method in not only ZSL performance but also convergence speed and computational cost. Specifically, our model outperforms the existing state-of-the-art methods by a remarkable margin up to 3.1% and 4.0% in ZSL and generalized ZSL settings, respectively.

Zero-shot learning (ZSL) aims to recognize image instances of unseen classes solely based on the semantic descriptions of the unseen classes. In this field, Generalized Zero-Shot Learning (GZSL) is a challenging problem in which the images of both seen and unseen classes are mixed in the testing phase of learning. Existing methods formulate GZSL as a semantic-visual correspondence problem and apply generative models such as Generative Adversarial Networks and Variational Autoencoders to solve the problem. However, these methods suffer from the bias problem since the images of unseen classes are often misclassified into seen classes. In this work, a novel model named the Dual Projective model for Zero-Shot Learning (DPZSL) is proposed using text descriptions. In order to alleviate the bias problem, we leverage two autoencoders to project the visual and semantic features into a latent space and evaluate the embeddings by a visual-semantic correspondence loss function. An additional novel classifier is also introduced to ensure the discriminability of the embedded features. Our method focuses on a more challenging inductive ZSL setting in which only the labeled data from seen classes are used in the training phase. The experimental results, obtained from two popular datasets—Caltech-UCSD Birds-200-2011 (CUB) and North America Birds (NAB)—show that the proposed DPZSL model significantly outperforms both the inductive ZSL and GZSL settings. Particularly in the GZSL setting, our model yields an improvement up to 15.2% in comparison with state-of-the-art CANZSL on datasets CUB and NAB with two splittings.

With the increasing popularity of artificial intelligence applications, machine learning is also playing an increasingly important role in the Internet of Things (IoT) and the Internet of Vehicles (IoV). As an essential part of the IoV, smart transportation relies heavily on information obtained from images. However, inclement weather, such as snowy weather, negatively impacts the process and can hinder the regular operation of imaging equipment and the acquisition of conventional image information. Not only that, but the snow also makes intelligent transportation systems make the wrong judgment of road conditions and the entire system of the Internet of Vehicles adverse. This paper describes the single image snow removal task and the use of a vision transformer to generate adversarial networks. The residual structure is used in the algorithm, and the Transformer structure is used in the network structure of the generator in the generative adversarial networks, which improves the accuracy of the snow removal task. Moreover, the vision transformer has good scalability and versatility for larger models and has a more vital fitting ability than the previously popular convolutional neural networks. The Snow100K dataset is used for training, testing and comparison, and the peak signal-to-noise ratio and structural similarity are used as evaluation indicators. The experimental results show that the improved snow removal algorithm performs well and can obtain high-quality snow removal images.

Zero-shot Learning (ZSL), which aims to predict for those classes that have never appeared in the training data, has arisen hot research interests. The key of implementing ZSL is to leverage the prior knowledge of classes which builds the semantic relationship between classes and enables the transfer of the learned models (e.g., features) from training classes (i.e., seen classes) to unseen classes. However, the priors adopted by the existing methods are relatively limited with incomplete semantics. In this paper, we explore richer and more competitive prior knowledge to model the inter-class relationship for ZSL via ontology-based knowledge representation and semantic embedding. Meanwhile, to address the data imbalance between seen classes and unseen classes, we developed a generative ZSL framework with Generative Adversarial Networks (GANs). Our main findings include: (i) an ontology-enhanced ZSL framework that can be applied to different domains, such as image classification (IMGC) and knowledge graph completion (KGC); (ii) a comprehensive evaluation with multiple zero-shot datasets from different domains, where our method often achieves better performance than the state-of-the-art models. In particular, on four representative ZSL baselines of IMGC, the ontology-based class semantics outperform the previous priors e.g., the word embeddings of classes by an average of 12.4 accuracy points in the standard ZSL across two example datasets (see Figure 4).

Most zero-shot learning (ZSL) methods aim to learn a mapping from visual feature space to semantic feature space or from both visual and semantic feature spaces to a common joint space and align them. However, in these methods the visual and semantic information are not utilized sufficiently and the useless information is not excluded. Moreover, there exists a strong bias problem that the instances from unseen classes always tend to be predicted as some seen classes in most ZSL methods. In this paper, combining the advantages of generative adversarial networks (GANs), a method based on bidirectional projections between the visual and semantic feature spaces is proposed. GANs are used to perform bidirectional generations and alignments between the visual and semantic features. In addition, cycle mapping structure ensures that the important information are kept in the alignments. Furthermore, in order to better solve the bias problem, pseudo-labels are generated for unseen instances and the model is adjusted along with them iteratively. We conduct extensive experiments at traditional ZSL and generalized ZSL settings, respectively. Experiment results confirm that our method achieves the state-of-the-art performances on the popular datasets AWA2, aPY and SUN.

Most zero-shot learning (ZSL) methods based on generative adversarial networks (GANs) utilize random noise and semantic descriptions to synthesize image features of unseen classes, which alleviates the problem of training data imbalance between seen and unseen classes. However, these methods usually only learn the distributions of seen classes in the training stage, ignoring the unseen ones. Due to the different distributions of seen and unseen samples, i.e., image features, these methods cannot generate unseen features of sufficient quality, so the performances are also limited, especially for the generalized zero-shot learning (GZSL) setting. In this article, we propose a general transductive method based on GANs, called GT-GAN, which can improve the quality of generated unseen image features and therefore benefit the classification. A new loss function is introduced to make the relative positions between each unseen image and its <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$k$ </tex-math></inline-formula> nearest neighbors in the feature space as consistent as possible with their relative positions in the semantic space; this loss function may be easily applied in most existing GAN-based models. Experimental results on five benchmark datasets show a significant improvement in accuracy compared with that of original models, especially in the GZSL setting.

Generative Adversarial Networks (GANs) have shown promise in enabling intelligent and privacy-preserving data synthesis within decentralized environments like the Internet of Vehicles (IoV). However, applying GANs in such dynamic, resource-constrained, and latency-sensitive settings poses significant challenges, particularly due to intermittent connectivity, mobility-induced disruptions, and communication overhead. In this paper, we propose a novel collaborative Multi-Discriminator GAN (MD-GAN) approach tailored for IoV, where Road Side Units (RSUs) act as generators and vehicles function as mobile discriminators providing distributed feedback. Unlike traditional centralized or synchronous GAN setups, our proposed approach leverages an early update strategy that allows generators to proceed once a minimum threshold of feedback is received, along with a feedback selection mechanism that considers only discriminators offering improved loss scores. We adopt the Wasserstein GAN (WGAN) formulation to ensure stable convergence under sparse and asynchronous feedback conditions. The proposed approach is implemented through a hybrid simulation architecture combining NS3 and PyTorch to model both networklevel interactions and learning dynamics. Experimental results demonstrate that the proposed approach significantly reduces training latency and network overhead compared to the baselines, while maintaining competitive generator accuracy. Our findings highlight the effectiveness of communication-efficient GAN training in highly dynamic vehicular networks.

Network traffic data are critical for network research. With the help of synthetic traffic, researchers can readily generate data for network simulation and performance evaluation. However, the state-of-the-art traffic generators are either too simple to generate realistic traffic or require the implementation of original applications and user operations. We propose Synthetic PAcket Traffic Generative Adversarial Networks (SPATGAN) that are capable of generating synthetic traffic. The framework includes a server agent and a client agent, which transmit synthetic packets to each other and take the opponent's synthetic packets as conditional labels for the built-in Timing Synthesis Generative Adversarial Networks (TSynGAN) and a Packet Synthesis Generative Adversarial Networks (PSynGAN) to generate synthetic traffic. The evaluations demonstrate that the proposed framework can generate traffic whose distribution resembles real traffic distribution.

The Internet of Vehicles (IoV) enables technology that allows IoV and vehicles to connect everything. IoV has become an essential component of modern life. This exponential growth of IoV technology has introduced significant security and privacy issues, which pose potential threats to different types of attacks and cause different threats to the normal operation of vehicles. To prevent intelligent vehicle accidents and identify malicious attacks within IoV networks, various researchers have focused on machine learning (ML)-based methods to detect attacks. Intrusion detection systems (IDS) are a prominent solution for cyber attacks in IoV using ensemble learning. To achieve higher accuracy and detection rate, designing an improved detection framework using ensemble learning is a challenging task. The design of an ensemble-based IDS depends on two main challenges: selecting base classifiers and their combination methods. Therefore, in this study, we propose a hybrid ML model to detect various attacks in IoV. We have used different ML algorithms to develop an enhanced algorithm that can efficiently detect attacks in IoV networks. To evaluate the performance of the proposed system, we have used two well-known datasets, (CIC-IDS2017) and (UNSW-NB15). The proposed algorithm shows outstanding performance from the performance results, with an average attack detection accuracy of 99.75% and 100% and an F1 score of 99.74% and 100%, respectively, for both datasets. Further performance scores, that is, recall, precision, and F1 score metrics, validate the exceptional effectiveness of the proposed framework.

With the development of the Internet of Things (IoT) and automobile industry in recent years, the Internet of Vehicle (IoV) has become a future direction of automobile development. Due to the large amount of vehicles, the opening of wireless media, the high-speed movement of vehicles and the impact of the environment, it is inevitable to produce abnormal data in IoVs including data tampering, loss, disorder and so on. However, there are few systematic research results for outlier detection of IoVs. The usability of the existing outlier detection schemes and their performances are not yet evaluated. To this issue, we select six applicable schemes and propose the outlier detection process for IoVs. Then we evaluate the comparison performances of the proposed schemes on real vehicle data collected by a Focus car.

Zero-shot learning (ZSL) is a challenging task due to the lack of unseen class data during training. Existing works attempt to establish a mapping between the visual and class spaces through a common intermediate semantic space. The main limitation of existing methods is the strong bias towards seen class, known as the domain shift problem, which leads to unsatisfactory performance in both conventional and generalized ZSL tasks. To tackle this challenge, we propose to convert ZSL to the conventional supervised learning by generating features for unseen classes. To this end, a joint generative model that couples variational autoencoder (VAE) and generative adversarial network (GAN), called Zero-VAE-GAN, is proposed to generate high-quality unseen features. To enhance the class-level discriminability, an adversarial categorization network is incorporated into the joint framework. Besides, we propose two self-training strategies to augment unlabeled unseen features for the transductive extension of our model, addressing the domain shift problem to a large extent. Experimental results on five standard benchmarks and a large-scale dataset demonstrate the superiority of our generative model over the state-of-the-art methods for conventional, especially generalized ZSL tasks. Moreover, the further improvement of the transductive setting demonstrates the effectiveness of the proposed self-training strategies.

Zero-shot learning (ZSL) is a challenging task due to the lack of data from unseen classes during training. Existing methods tend to have the strong bias towards seen classes, which is also known as the domain shift problem. To mitigate the gap between seen and unseen class data, we propose a joint generative model to synthesize features as the replacement for unseen data. Based on the generated features, the conventional ZSL problem can be tackled in a supervised way. Specifically, our framework integrates Variational Autoencoders (VAE) and Generative Adversarial Networks (GAN) conditioned on class-level semantic attributes for feature generation based on element-wise and holistic reconstruction. A categorization network acts as the additional guide to generate features beneficial for the subsequent classification task. Moreover, we propose a perceptual reconstruction loss to preserve semantic similarities. Experimental results on five benchmarks show the superiority of our framework over the state-of-the-art approaches in terms of both conventional ZSL and generalized ZSL settings.

Traffic forecasting is a crucial aspect of modern Intelligent Transportation Systems (ITS) and the Internet of Vehicles (IoV), playing a vital role in improving the safety and efficiency of daily transportation activities. Despite the valuable contributions of traditional machine learning (ML) models and advanced deep learning (DL) techniques, there persist challenges in capturing the intricate spatial and temporal dependencies inherent in traffic flow. In response to these challenges, we present GRAPHITE, an innovative framework that combines Graph Neural Networks (GNNs) and Generative Adversarial Networks (GANs) to leverage generative reasoning for efficient traffic management. Our model seamlessly integrates historical traffic volume data collected by road sensors with local spatial information encoded through knowledge graphs (KGs) associated with each sensor. These KGs offer a structured representation of relationships between traffic sensors and points of interest (POIs) in their neighborhood, thereby enhancing the comprehension of the urban context and leading to more accurate traffic predictions. Extensive experiments conducted on diverse datasets underscore the efficacy of GRAPHITE. Notably, we achieved a maximum decrease in RMSE of 31.05% compared to GAN-GRU and a maximum increase in R2 of 8.15% compared to GAN-RNN, positioning GRAPHITE as a standout solution among the current state-of-the-art approaches. Our code is available at: https://github.com/MODAL-UNINA/GRAPHITE.

The current security method of Internet-of-Vehicles (IoV) systems is rare, which makes it vulnerable to various attacks. The malicious and unauthorized nodes can easily invade the IoV systems to destroy the integrity, availability, and confidentiality of information resources shared among vehicles. Indeed, access control mechanism can remedy this. However, as a static method, it cannot timely response to these attacks. To solve this problem, we propose an intelligent edge-chain-enabled access control framework with vehicle nodes and roadside units (RSUs) in this study. In our scenario, vehicle nodes act as lightweight nodes, whereas RUSs serve as full and edge nodes to provide access control services. Considering the low accuracy of risk prediction due to limited training sets, we leverage a generative adversarial networks (GANs) to convert the risk prediction to a sequence generation. Moreover, aiming at the problems of gradient disappearance and mode collapse existed in the original GANs, we devise a Wasserstein combined GANs (WCGANs). Simulation results demonstrate that WCGAN has higher prediction accuracy than the original GANs. Additionally, it can also improve the accuracy of access control of risk prediction-based access control (RPBAC) model.

The popularization of the Internet of Vehicles (IoV) technology has not only brought many conveniences but also caused security issues such as distributed denial of service (DDoS) attacks. Using deep learning methods to deal with DDoS attacks has achieved remarkable results. However, to improve the effectiveness of the model, a large amount of high-quality data is needed. This data is often scattered among different vehicle operating companies or organizations, making it difficult to train effective attack detection models by centralizing the data. To solve this problem, this paper proposes a DDoS attack detection method based on reinforced federated learning in the IoV environment. This method applies reinforcement learning to the aggregation process of federated learning to optimize the DDoS attack detection algorithm. In this method, the fog node server of the IoV serves as a participant in learning, uses local data to conduct reinforcement learning, and uploads the trained model parameters to the edge server of the IoV for aggregation. By learning the adaptability of the fog node model to the environment, reinforced federated learning can improve the accuracy of the DDoS attack detection model in terms of unbalanced data distribution and model convergence efficiency. Experiments were carried out using the public DDoS attack dataset. The experimental results showed that the proposed method exhibited excellent performance in the IoV environment, providing new ideas and methods for the research of IoV security.