Genuine Ones Research Articles

Deep Convolutional Generative Adversarial Networks in Image-Based Android Malware Detection

The recent advancements in generative adversarial networks have showcased their remarkable ability to create images that are indistinguishable from real ones. This has prompted both the academic and industrial communities to tackle the challenge of distinguishing fake images from genuine ones. We introduce a method to assess whether images generated by generative adversarial networks, using a dataset of real-world Android malware applications, can be distinguished from actual images. Our experiments involved two types of deep convolutional generative adversarial networks, and utilize images derived from both static analysis (which does not require running the application) and dynamic analysis (which does require running the application). After generating the images, we trained several supervised machine learning models to determine if these classifiers can differentiate between real and generated malicious applications. Our results indicate that, despite being visually indistinguishable to the human eye, the generated images were correctly identified by a classifier with an F-measure of approximately 0.8. While most generated images were accurately recognized as fake, some were not, leading them to be considered as images produced by real applications.

Detecting Phishing Links Analysis using Machine Learning

This proposed strategy for identifying phishing websites employed the Gradient Boosting Classifier model, focusing on various aspects of URL significance. By meticulously extracting and comparing different characteristics between legitimate and phishing URLs, our approach leverages the Gradient Boosting Classifier to identify phishing URLs effectively. The study's findings underscore the successful application of our suggested approach in real-time, demonstrating its ability to distinguish between legitimate and bogus websites. Given the relentless evolution of phishing techniques facilitated by advancing technology, employing anti-phishing methods is imperative. Phishing attacks, which often rely on deceptive websites closely resembling genuine ones in appearance and language, pose a significant threat. Machine learning emerges as a robust tool in thwarting such assaults, offering the ability to discern subtle patterns indicative of malicious intent. Phishing remains a preferred tactic for attackers due to its effectiveness in bypassing traditional security measures. By duping unsuspecting users into clicking seemingly authentic yet malicious links, attackers exploit human vulnerability, highlighting the importance of proactive detection mechanisms. In this context, our utilization of the Gradient Boosting Classifier underscores the efficacy of machine learning in fortifying defenses against phishing attacks. As cyber threats evolve, embracing innovative approaches like machine learning becomes essential in safeguarding against emerging risks.

A Quality-Aware and Obfuscation-Based Data Collection Scheme for Cyber-Physical Metaverse Systems

In pursuit of an immersive virtual experience within the Cyber-Physical Metaverse Systems (CPMS), the construction of Avatars often requires a significant amount of real-world data. Mobile Crowd Sensing (MCS) has emerged as an efficient method for collecting data for CPMS. While progress has been made in protecting the privacy of workers, little attention has been given to safeguarding task privacy, potentially exposing the intentions of applications and posing risks to the development of the Metaverse. Additionally, existing privacy protection schemes hinder the exchange of information among entities, inadvertently compromising the quality of the collected data. To this end, we propose a Quality-aware and Obfuscation-based Task Privacy-Preserving (QOTPP) scheme, which protects task privacy and enhances data quality without third-party involvement. The QOTPP scheme initially employs the insight of “showing the fake, and hiding the real” by employing differential privacy techniques to create fake tasks and conceal genuine ones. Additionally, we introduce a two-tier truth discovery mechanism using Deep Matrix Factorization (DMF) to efficiently identify high-quality workers. Furthermore, we propose a Combinatorial Multi-Armed Bandit (CMAB)-based worker incentive and selection mechanism to improve the quality of data collection. Theoretical analysis confirms that our QOTPP scheme satisfies essential properties such as truthfulness, individual rationality, and ϵ-differential privacy. Extensive simulation experiments validate the state-of-the-art performance achieved by QOTPP.

Jordanian banknote data recognition: A CNN-based approach with attention mechanism

Identifying counterfeit banknotes is crucial in financial transactions, as the process of identification cannot be handled by ATMs or vending machines. The recent developments in technology, particularly the smart systems that are integrated with cameras and artificial intelligence (AI) tools, allow for the distinction of currency and the detection of counterfeit. In this study, we are suggesting an approach for identifying counterfeit Jordanian banknotes and differentiating them from genuine ones. The suggested approach collaborates deep learning through a convolutional neural network (CNN) and another attention mechanism which contributes to focusing on features of importance while avoiding features of less importance. The proposed model has proven its ability to recognize counterfeits with high performance and accuracy while focusing on the important features extracted. The study made use of a data set from Kaggle that includes a collection of Jordanian banknotes in five different denominations. Image processing techniques were employed to produce artificial images by boosting the brightness of real ones. Eight trained models and the suggested model were compared. It demonstrated excellence with encouraging outcomes, achieving 96% accuracy, 96.6% precision, 96.4% recall, and 94.5% f1-score. Also, we tested our approach on two datasets Indian dataset and DS1, DS2, and DS3 datasets, we obtained 88% and 99.9% accuracy, respectively. The achievement of detecting counterfeit Jordanian banknotes is proof that a well-established AI model contributes to dealing with security vulnerabilities in many institutions.

A deep learning fusion model for accurate classification of brain tumours in Magnetic Resonance images

AbstractDetecting brain tumours is complex due to the natural variation in their location, shape, and intensity in images. While having accurate detection and segmentation of brain tumours would be beneficial, current methods still need to solve this problem despite the numerous available approaches. Precise analysis of Magnetic Resonance Imaging (MRI) is crucial for detecting, segmenting, and classifying brain tumours in medical diagnostics. Magnetic Resonance Imaging is a vital component in medical diagnosis, and it requires precise, efficient, careful, efficient, and reliable image analysis techniques. The authors developed a Deep Learning (DL) fusion model to classify brain tumours reliably. Deep Learning models require large amounts of training data to achieve good results, so the researchers utilised data augmentation techniques to increase the dataset size for training models. VGG16, ResNet50, and convolutional deep belief networks networks extracted deep features from MRI images. Softmax was used as the classifier, and the training set was supplemented with intentionally created MRI images of brain tumours in addition to the genuine ones. The features of two DL models were combined in the proposed model to generate a fusion model, which significantly increased classification accuracy. An openly accessible dataset from the internet was used to test the model's performance, and the experimental results showed that the proposed fusion model achieved a classification accuracy of 98.98%. Finally, the results were compared with existing methods, and the proposed model outperformed them significantly.

DNN controlled adaptive front-end for replay attack detection systems

Developing robust countermeasures to protect automatic speaker verification systems against replay spoofing attacks is a well-recognized challenge. Current approaches to spoofing detection are generally based on a fixed front-end, typically a time-invariant filter bank, followed by a machine learning back-end. In this paper, we propose a novel approach whereby the front-end comprises an adaptive filter bank with a deep neural network-based controller, which is jointly trained along with a neural network back-end. Specifically, the deep neural network-based adaptive filter controller tunes the selectivity and sensitivity of the front-end filter bank at every frame to capture replay-related artefacts. We demonstrate the effectiveness of the proposed framework in spoofing attack detection on a synthesized dataset and ASVSpoof 2019 and ASVSpoof 2021 challenge datasets in terms of equal error rate and its ability to capture artefacts that differentiate replayed signals from genuine ones in comparison to conventional non-adaptive front-ends.

An explainable ensemble of multi-view deep learning model for fake review detection

Online reviews significantly impact consumers who are purchasing or seeking services via the Internet. Businesses and review platforms need to manage these online reviews to avoid misleading customers through fake ones. This necessitates developing intelligent solutions to detect these fake reviews and prevent their negative impact on businesses and customers. Therefore, many fake review detection models have been proposed to help distinguish fake reviews from genuine ones. However, these techniques depend on a limited perspective of features, mainly review content, to detect fake reviews, leading to poor performance in discovering the new patterns of fake review content and the dynamic behaviour of spammers. Therefore, there is still a need to develop new solutions to detect the new patterns of fake reviews. Hence, this paper proposes an explainable multi-view deep learning model to identify fake reviews based on different feature perspectives and classifiers. The proposed model can extract essential features from different perspectives, including review content, reviewer data, and product description. Moreover, we employ an ensemble approach that combines three popular deep learning algorithms: Bi-LSTM, CNN, and DNN, to enhance the performance of the fake review detection model. The results of two real-life datasets presented demonstrated the efficiency of our proposed model, where it outperformed the state-of-the-art methods with improvements ranging from 1% to 7% in terms of the AUC metric. To provide visibility into the outcomes of our proposed model and demonstrate the trust and transparency in the obtained results, we also offer a comprehensive explanation for our model results using Shapely Additive Explanations (SHAP) method and attention techniques. The experimental results prove that our proposed model can provide reasonable explanations that help users understand why specific reviews are classified as fake.

Graph-Based Soft-Balanced Fuzzy Clustering

Spectral clustering have attracted more and more attention due to their well-defined mathematical frameworks and superior performance. However, there still exist two limitations to be solved: 1) most spectral clustering methods consist of two independent stages, which may cause unpredictable deviation of obtained clustering results from the genuine ones and lead to severe information loss and performance degradation; 2) spectral clustering methods employ the hard clustering mode, which lacks interpretability for data points in the boundary area belonging to multiple clusters. To simultaneously address these challenging issues for spectral clustering, we propose a Graph based soft-Balanced Fuzzy Clustering (GBFC) model. Specifically, we explicitly preserve the nonnegative property of the clustering indicator matrix to enhance the interpretability of clustering results. Moreover, row normalization is imposed on the cluster indicator matrix to show the membership of each data point to different clusters. Additionally, a novel balanced constraint is designed to regularize the clustering results and constrain the size of clusters. We can directly obtain the clustering assignments without any post-processing, and the limitations of the previous two-stage clustering framework can be effectively addressed. Extensive experiments performed on both synthetic datasets and real world datasets demonstrate the superiority and effectiveness of the proposed algorithm compared with several state-of-the-art methods.

Generation and Detection of Face Morphing Attacks

Abstract: Face morphing attacks are a growing concern in the digital world, with the potential to compromise personal privacy and security. In this project, we investigate the application of deep learning methods for both generating and detecting face morphing attacks. For the generation of morphed, we use a combination of convolutional neural networks and auto encoders to learn the underlying facial features and generate realistic-looking images. On the detection side, we develop a framework based on facial feature consistency analysis to differentiate distorted images from real ones. The proposed framework achieves high accuracy in detecting face morphing attacks, even in cases where the morphed images are visually similar to the genuine ones. This project highlights the potential of deep-learning based approaches for addressing the problem of face morphing attacks and provides insights for further research and development in this area

Exploring Latent Characteristics of Fake Reviews and Their Intermediary Role in Persuading Buying Decisions

Online reviews play a significant role in shaping consumer purchase decisions. Accordingly, emergence of fake reviews has proliferated as an instrument to manipulate customers’ buying preferences. Such manifestation, however, lacks theoretical grounding and remains under researched due to two notable challenges: first, absence of conceptual underpinnings between consumers’ writing style and recommendation behavior. Second, little knowledge about the role of product characteristics underlying fake reviews and their influence on nudging product preferences. Through the lens of environmental psychology, this study uses an empirical investigation utilizing natural language processing (NLP) to uncover latent product-specific features underlying customer reviews and their impact on persuading buying preferences. As a major finding, we observe that characteristics underlying fake reviews, as opposed to genuine ones, fail to influence product recommendation or discouragement. Accordingly, we suggest firms permitting fake reviews on their portals to be aware of the limited economic advantages of such practices.

Fake News Detection using Machine Learning

The approach of the Internet and the quick reception of online entertainment stages (like Facebook and Twitter) prepared for data scattering that has never been seen in mankind's set of experiences previously. With the ongoing use of online entertainment stages, shoppers are making and sharing more data than at any other time in recent memory, some of which are deluding with no importance to the real world. Mechanized characterization of a text article as deception or disinformation is difficult. Indeed, even a specialist in a specific space needs to investigate various viewpoints before deciding on the honesty of an article. In this work, we propose to utilize an AI-gathering approach for robotized order of news stories. Our review investigates different printed properties that can be utilized to separate phony items from genuine ones. By utilizing those properties, we train a mix of various AI calculations utilizing different troupe strategies and assess their exhibition on 4 genuine world datasets. Exploratory assessment affirms the unrivaled presentation of our proposed outfit student approach in contrast with individual students. Counterfeit news discovery is bit by bit happening to principal significance to our society to keep away from the alleged reality dizziness and safeguard specifically the less instructed people. Different AI procedures have been proposed to resolve this issue. This article presents a thorough exhibition assessment of eight AI calculations for counterfeit news identification/characterization

Designed to abuse? Deepfakes and the non-consensual diffusion of intimate images

The illicit diffusion of intimate photographs or videos intended for private use is a troubling phenomenon known as the diffusion of Non-Consensual Intimate Images (NCII). Recently, it has been feared that the spread of deepfake technology, which allows users to fabricate fake intimate images or videos that are indistinguishable from genuine ones, may dramatically extend the scope of NCII. In the present essay, we counter this pessimistic view, arguing for qualified optimism instead. We hypothesize that the growing diffusion of deepfakes will end up disrupting the status that makes our visual experience of photographic images and videos epistemically and affectively special; and that once divested of this status, NCII will lose much of their allure in the eye of the perpetrators, probably resulting in diminished diffusion. We conclude by offering some caveats and drawing some implications to better understand, and ultimately better counter, this phenomenon.

LMD: A Learnable Mask Network to Detect Adversarial Examples for Speaker Verification

Although the security of automatic speaker verification (ASV) is seriously threatened by recently emerged adversarial attacks, there have been some countermeasures to alleviate the threat. However, many defense approaches not only require the prior knowledge of the attackers but also possess weak interpretability. To address this issue, in this paper, we propose an <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">attacker-independent</i> and <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">interpretable</i> method, named <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">learnable mask detector</i> (LMD), to separate adversarial examples from the genuine ones. It utilizes score variation as an indicator to detect adversarial examples, where the score variation is the absolute discrepancy between the ASV scores of an original audio recording and its transformed audio synthesized from its masked complex spectrogram. A core component of the score variation detector is to generate the masked spectrogram by a neural network. The neural network needs only genuine examples for training, which makes it an attacker-independent approach. Its interpretability lies that the neural network is trained to minimize the score variation of the targeted ASV, and maximize the number of the masked spectrogram bins of the genuine training examples. Its foundation is based on the observation that, masking out the vast majority of the spectrogram bins with little speaker information will inevitably introduce a large score variation to the adversarial example, and a small score variation to the genuine example. Experimental results with 12 attackers and two representative ASV systems show that our proposed method outperforms five state-of-the-art baselines. The extensive experimental results can also be a benchmark for the detection-based ASV defenses.

Insurance fraud detection: Pre-emptive analysis and prevention

Insurance companies have been facing the problem of false insurance claims, costing resources and money in detecting fraudulent claims among the many true ones. Many times, fraudulent claims are mistaken for genuine claims, and genuine ones mistaken for fraudulent claims. This paper aims to develop a Machine Learning algorithm to detect fraudulent claims and distinguish them with genuine claims. This project also aims to compare other algorithms and see which one is the most effective to deal with this growing problem that the insurance industry is facing.

A Novel Smart Deepfake Video Detection System

Rapid advancements in deep learning-based technologies have developed several synthetic video and audio generation methods producing incredibly hyper-realistic deepfakes. These deepfakes can be employed to impersonate the identity of a source person in videos by swapping the source's face with the target one. Deepfakes can also be used to clone the voice of a target person utilizing audio samples. Such deepfakes may pose a threat to societies if they are utilized maliciously. Consequently, distinguishing either one or both deepfake visual video frames and cloned voices from genuine ones has become an urgent issue. This work presents a novel smart deepfake video detection system. The video frames and audio are extracted from given videos. Two feature extraction methods are proposed, one for each modality; visual video frames, and audio. The first method is an upgraded XceptionNet model, which is utilized for extracting spatial features from video frames. It produces feature representation for visual video frames. The second one is a modified InceptionResNetV2 model based on the Constant-Q Transform (CQT) method. It is employed to extract deep time-frequency features from the audio modality. It produces feature representation for the audio. The corresponding extracted features of both modalities are fused at a mid-layer to produce a bimodal information-based feature representation for the whole video. These three representation levels are independently fed into the Gated Recurrent Unit (GRU) based attention mechanism helping to learn and extract deep and important temporal information per level. Then, the system checks whether the forgery is only applied to video frames, audio, or both, and produces the final decision about video authenticity. The newly suggested method has been evaluated on the FakeAVCeleb multimodal videos dataset. The experimental results analysis assures the superiority of the new method over the current–state-of-the-art methods.

Metaheuristic Optimized Ensemble Model for Classification of SMS Spam in Computer Networks

By use of electronic communication, we are able to communicate a message to the recipient. In this digital age, a collaboration between several people is possible thanks to a variety of digital technologies. This interaction may take place in a variety of media formats, including but not limited to text, images, sound, and language. Today, a person's primary means of communication is their smart gadget, most commonly a cell phone. Spam is another side effect of our increasingly text-based modes of communication. We received a bunch of spam texts on our phones, and we know they're not from anyone we know. The vast majority of businesses nowadays use spam texts to advertise their wares, even when recipients have explicitly requested not to receive such messages. As a rule, there are many more spam emails than genuine ones. We apply text classification approaches to define short messaging service (SMS) and spam filtering in this study, which effectively categorizes messages. In this paper, we use machine learning algorithms and metaheuristic optimization to determine what percentage of incoming SMS messages are spam. This is why we used the optimized models to evaluate and contrast many classification strategies for gathering data.

KurdFace Morph Dataset Creation Using OpenCV

Automated facial recognition is rapidly being used to reliably identify the identities of individuals for a variety of applications, from automated border control to unlocking mobile phones. The attack of Morphing has presented a significant risk to the face recognition system (FRS) at automated border control. Face morphing is a technique for blending the facial images of two or more people such that the outcome looks like both of them. For example, a morphing attack may be used to get a fake passport by using a morphed image. This passport can be used by both the modified image contributors while crossing the border. Due to the publicly available digital altering tools that criminals may use to carry out face morphing attacks. Morph Attack Detection (MAD) systems have received a lot of attention in recent years. In the absence of automated morphing detection, Face Recognition Systems (FRS) are extremely susceptible to morphing attacks. Due to the limited number of publicly available face morph datasets to investigate, especially to our knowledge, there is no Kurdish morph dataset. In this work, we decided to generate a new face dataset, including morphed images which we named as "KurdFace" dataset. OpenCV was used to generate morphed images. Then we study the susceptibility of biometric systems to such morphed face attacks by designing and creating a Morph Attack Detection model to distinguish morphed images from genuine ones. To evaluate the robustness of our dataset regarding morphing attack detection, we compare it with the AMSL dataset to determine the classification error rate on both datasets to see how our dataset is different from others. Local Binary Pattern and Uniform Local Binary Pattern are used as feature extraction techniques, and as a classifier, SVM is utilized. The experimental result shows that our dataset is suitable for research purposes.

LBPNet: Exploiting texture descriptor for deepfake detection

Recent AI advancements made it significantly easier to generate high-quality synthesized faces, referred as deepfakes. It can make people saying and doing things that never happened actually. Owing to the vast usage of social media, manipulated content is susceptible to spread unrest in the society. With the continuous evolvement of AI-enabled deepfake generators, the research on deepake detection has started progressing in the last few years. From analyzing eye-blinking in the previous generation deepfakes to deep learning based models for advanced deepfakes, the existing deepfake detectors have analyzed numerous artefacts. This paper proposes a novel detection approach, LBPNet, to distinguish deepfaked faces from genuine ones by means of exploiting inconsistencies in texture information. In particular, Local Binary Pattern (LBP) of deepfaked and pristine faces have been investigated through a CNN based model. Moreover, the proposed LBPNet technique is evaluated on more advanced and diverse deepfaked datasets such as Celeb-DF, DFDC, and DeeperForensics, which provided a detection accuracy of 92.38%, 80% and 86% respectively. Comprehensive analysis on different benchmark datasets and comparison with state-of-art endorse the superior performance of the proposed method. Thorough experimentations also reveal the robustness of LBPNet against different compression levels and tampering types.

Real-time DDoS Detection and Mitigation in Software Defined Networks using Machine Learning Techniques

Software Defined Network (SDN) is the new era of networking technology based on a centralized controller that separates the switch hardware from its operating software. The most important challenge is the security of SDN and the most prominent attack is the Distributed Denial of Service (DDoS) attack. Some of the research work done so far detects DDoS attacks using a threshold, which is usually assumed without proper scientific reason and hence may not be always accurate. The mitigation techniques used by some researchers block the host from sending the network traffic beyond a threshold, by installing drop rules in the flow table of the switch connected to that host. Doing so will not only block the attack traffic but also the genuine ones from other applications of that host. In this paper, we propose a model that calculates the threshold limit for the type of applications sending data to a particular switch, in real-time using a machine learning (ML) model, and determines whether that application traffic is DDoS traffic. After the detection, only application type sending DDoS traffic is blocked while other genuine applications are allowed to send the network traffic without any interruption. The use of a dynamic threshold, based on the current network traffic, will help in detecting DDoS efficiently.

Anti-phishing technique based on dynamic image captcha using multi secret sharing scheme

With the cutting-edge improvement of web, online abuses have been increasing rapidly. Phishing is the most widely recognized abuses performed by digital crooks nowadays. It is an activity to steal private data (for example, client names, passwords and Visa data) in an electronic correspondence. It is a sort of fraud with the end goal of monetary benefit and other fake exercises. It utilizes phony websites that resemble genuine ones. Phishing messages might contain links to sites that are contaminated with malware. In this paper, “an anti-phishing approach using multi secret sharing scheme” is implemented as an answer to this problem. Here, Dynamic Image CAPTCHA based verification using multi secret sharing is performed. Image CAPTCHA is divided into two pieces called shares. Multiple secret pictures are revealed by overlapping the same set of shares at different angles. In the proposed approach, shares are of different modes i.e., user’s share is imprinted on a physical transparency while server’s share is in digital mode. By using the proposed approach, websites and end clients can cross confirm their identity.