The proliferation of generative models, such as Generative Adversarial Networks (GANs), Diffusion Models, and Variational Autoencoders (VAEs), has enabled the synthesis of high-quality multimedia data. However, these advancements have also raised significant concerns regarding adversarial attacks, unethical usage, and societal harm. Recognizing these challenges, researchers have increasingly focused on developing methodologies to detect synthesized data effectively, aiming to mitigate potential risks. Prior reviews have predominantly focused on deepfake detection and often overlook recent advancements in synthetic image forensics, particularly approaches that incorporate multimodal frameworks, reasoning-based detection, and training-free methodologies. To bridge this gap, this survey provides a comprehensive and up-to-date review of state-of-the-art techniques for detecting and classifying synthetic images generated by advanced generative AI models. The review systematically examines core detection paradigms, categorizes them into spatial-domain, frequency-domain, fingerprint-based, patch-based, training-free, and multimodal reasoning-based frameworks, and offers concise descriptions of their underlying principles. We further provide detailed comparative analyses of these methods on publicly available datasets to assess their generalizability, robustness, and interpretability. Finally, the survey highlights open challenges and future directions, emphasizing the potential of hybrid frameworks that combine the efficiency of training-free approaches with the semantic reasoning of multimodal models to advance trustworthy and explainable synthetic image forensics.

Clouds remain a major obstacle in optical satellite imaging, limiting accurate environmental and climate analysis. To address the strong spectral variability and the large scale differences among cloud types, we propose MSCloudCAM, a novel multi-scale context adapter network with convolution based cross-attention tailored for multispectral and multi-sensor cloud segmentation. A key contribution of MSCloudCAM is the explicit modeling of multiple complementary multi-scale context extractors. And also, rather than simply stacking or concatenating their outputs, our formulation uses one extractor's fine-resolution features and the other extractor's global contextual representations enabling dynamic, scale-aware feature selection. Building on this idea, we design a new convolution-based cross attention adapter that effectively fuses localized, detailed information with broader multi-scale context. Integrated with a hierarchical vision backbone and refined through channel and spatial attention mechanisms, MSCloudCAM achieves strong spectral-spatial discrimination. Experiments on various multisensor datatsets e.g. CloudSEN12 (Sentinel-2) and L8Biome (Landsat-8) show that MSCloudCAM outperforms recent state-of-the-art models while maintaining competitive model complexity, highlighting the novelty and effectiveness of the proposed design for large-scale Earth observation.

Image-to-Image translation in Generative Artificial Intelligence (Generative AI) has been a central focus of research, with applications spanning healthcare, remote sensing, physics, chemistry, photography, and more. Among the numerous methodologies, Generative Adversarial Networks (GANs) with contrastive learning have been particularly successful. This study aims to demonstrate that the Kolmogorov-Arnold Network (KAN) can effectively replace the Multi-layer Perceptron (MLP) method in generative AI, particularly in the subdomain of image-to-image translation, to achieve better generative quality. Our novel approach replaces the two-layer MLP with a two-layer KAN in the existing Contrastive Unpaired Image-to-Image Translation (CUT) model, developing the KAN-CUT model. This substitution favors the generation of more informative features in low-dimensional vector representations, which contrastive learning can utilize more effectively to produce high-quality images in the target domain. Extensive experiments, detailed in the results section, demonstrate the applicability of KAN in conjunction with contrastive learning and GANs in Generative AI, particularly for image-to-image translation. This work suggests that KAN could be a valuable component in the broader generative AI domain.

Road extraction is a sub-domain of remote sensing applications; it is a subject of extensive and ongoing research. The procedure of automatically extracting roads from satellite imagery encounters significant challenges due to the multi-scale and diverse structures of roads; improvement in this field is needed. Convolutional neural networks (CNNs), especially the DeepLab series known for its proficiency in semantic segmentation due to its efficiency in interpreting multi-scale objects’ features, address some of these challenges caused by the varying nature of roads. The present work proposes the utilization of DeepLabV3+, the latest version of the DeepLab series, by introducing an innovative Dense Depthwise Dilated Separable Spatial Pyramid Pooling (DenseDDSSPP) module and integrating it in the place of the conventional Atrous Spatial Pyramid Pooling (ASPP) module. This modification enhances the extraction of complex road structures from satellite images. This study hypothesizes that the integration of DenseDDSSPP with a CNN backbone network and a Squeeze-and-Excitation block will generate an efficient dense feature map by focusing on relevant features, leading to more precise and accurate road extraction from remote sensing images. The Results Section presents a comparison of our model’s performance against state-of-the-art models, demonstrating better results that highlight the effectiveness and success of the proposed approach.

One of the persistent challenges in multispectral image analysis is the interference caused by dense cloud cover and its resulting shadows, which can significantly obscure surface features. This becomes especially problematic when attempting to monitor surface changes over time using satellite imagery, such as from Landsat-8. In this study, rather than simply masking visual obstructions, we aimed to investigate the role and influence of clouds within the spectral data itself. To achieve this, we employed Independent Component Analysis (ICA), a statistical method capable of decomposing mixed signals into independent source components. By applying ICA to selected Landsat-8 bands and analyzing each component individually, we assessed the extent to which cloud signatures are entangled with surface data. This process revealed that clouds contribute to multiple ICA components simultaneously, indicating their broad spectral influence. With this influence on multiple wavebands, we managed to configure a set of components that could perfectly delineate the extent and location of clouds. Moreover, because Landsat-8 lacks cloud-penetrating wavebands, such as those in the microwave range (e.g., SAR), the surface information beneath dense cloud cover is not captured at all, making it physically impossible for ICA to recover what is not sensed in the first place. Despite these limitations, ICA proved effective in isolating and delineating cloud structures, allowing us to selectively suppress them in reconstructed images. Additionally, the technique successfully highlighted features such as water bodies, vegetation, and color-based land cover differences. These findings suggest that while ICA is a powerful tool for signal separation and cloud-related artifact suppression, its performance is ultimately constrained by the spectral and spatial properties of the input data. Future improvements could be realized by integrating data from complementary sensors—especially those operating in cloud-penetrating wavelengths—or by using higher spectral resolution imagery with narrower bands.

Latent Diffusion Models (LDM), a subclass of diffusion models, mitigate the computational complexity of pixel-space diffusion by operating within a compressed latent space constructed by Variational Autoencoders (VAEs), demonstrating significant advantages in Remote Sensing (RS) applications. Though numerous studies enhancing LDMs have been conducted, investigations explicitly targeting improvements within the intrinsic latent space remain scarce. This paper proposes an innovative perspective, utilizing the Discrete Wavelet Transform (DWT) to enhance the VAE's latent space representation, designed for satellite imagery. The proposed method, ExpDWT-VAE, introduces dual branches: one processes spatial domain input through convolutional operations, while the other extracts and processes frequency-domain features via 2D Haar wavelet decomposition, convolutional operation, and inverse DWT reconstruction. These branches merge to create an integrated spatial-frequency representation, further refined through convolutional and diagonal Gaussian mapping into a robust latent representation. We utilize a new satellite imagery dataset housed by the TerraFly mapping system to validate our method. Experimental results across several performance metrics highlight the efficacy of the proposed method at enhancing latent space representation.

Environmental monitoring data often have irregularities, like missing values or unevenly spaced observations. These irregularities are not just data anomalies; rather, they mirror the complexity and unpredictability of natural ecosystems and the challenges of collecting data in such settings. Preserving these irregularities is essential because artificial smoothing or interpolation can obscure critical insights and misrepresent natural processes. In this work, we propose a novel method that incorporates Neural Ordinary Differential Equations (Neural ODEs) with Long Short-Term Memory (LSTM)-based models, as well as feature engineering, to address these irregularities in water quality metrics forecasting. To ensure the authenticity of real-world dynamics, our methodology retains the dataset's inherent irregularities. While Long Short-Term Memory (LSTM)-based models excel in capturing sequential patterns and dependencies, Neural ODEs excel at handling uneven sampling by offering a continuous-time perspective. To make the models more adaptable, we incorporate features such as time differences between observations, timestamps, and other relevant water quality metrics. For this study, we use 30 years of data from five monitoring stations along Florida's Saint Lucie River, carefully maintaining the irregularities inherent in the dataset. This study shows that our proposed model significantly outperforms state-of-the-art Neural ODEs and recurrent architectures. This framework provides a reliable way to forecast water quality metrics while handling irregularities in time series forecasting, offering new insights and better predictions for environmental monitoring and management.

Traditional perimeter-based security models struggle to secure legacy systems against evolving threats posed by remote work, IoT adoption, and cloud migration. Yet most Zero Trust (ZT) roadmaps demand disruptive refactoring that many organizations cannot afford. We present a lightweight, identity-centric transition model that overlays, rather than replaces, existing networks, relying on just three open-source components: Identity & Access Management, Public Key Infrastructure, and Continuous Diagnostics & Mitigation. A three-node Azure prototype using StrongSwan mutual TLS tunnels demonstrates that the full control-plane bundle idles at approximately 8% CPU and 320 MB RAM, each endpoint agent under 1% CPU and 30 MB RAM, and encrypted throughput remains within 2.5% of underlay performance while certificate revocation propagates in 8 min 14 s. These results show that meaningful ZT protections can be deployed immediately-no new hardware, rewiring, or licensing-offering a practical path to deeper ZT maturity.

Generative models have recently gained popularity in remote sensing, offering substantial benefits for interpreting and utilizing satellite imagery across diverse applications such as climate monitoring, urban planning, and wildfire detection. These models are particularly adept at addressing the challenges posed by satellite images, which often exhibit domain variability due to seasonal changes, sensor characteristics, and, especially, variations in spectral bands. Such variability can significantly impact model performance across various tasks. In response to these challenges, our work introduces an adaptive approach that harnesses the capabilities of generative adversarial networks (GANs), augmented with contrastive learning, to generate target domain images that account for multispectral band variations effectively. By maximizing mutual information between corresponding patches and leveraging the power of GANs, our model aims to generate realistic-looking images across different multispectral domains. We present a comparative analysis of our model against other well-established generative models, demonstrating its efficacy in generating high-quality satellite images while effectively managing domain variations inherent to multispectral diversity.

Motivated by the challenge of navigating the complex landscape of cybersecurity compliance, this study critically examines and evaluates seven major cybersecurity frameworks: SOC 2, GDPR, PCI DSS, HIPAA, CIS Controls V8, NIST CSF, and CMMC 2.0. Our research focuses on understanding their distinct features and operational nuances, addressing a significant gap in current compliance strategies. We contribute a novel set of risk management-based evaluation criteria, offering a comprehensive analysis of these frameworks. The study further explores the Secure Controls Framework (SCF) and its effective integration with these frameworks, summarizing a unified mapping approach. This mapping facilitates streamlined compliance across multiple standards, providing a strategic tool for organizations. Our findings offer pivotal insights into the efficacy of each framework in managing cybersecurity risks, underlining the necessity for an integrated, risk-focused approach to compliance in the digital era.

We consider the task of classifying trajectories of boat activities as a proxy for assessing maritime threats. Previous approaches have considered entropy-based metrics for clustering boat activity into three broad categories: random walk, following, and chasing. Here, we comprehensively assess the accuracy of neural network-based approaches as alternatives to entropy-based clustering. We train four neural network models and compare them to shallow learning using synthetic data. We also investigate the accuracy of models as time steps increase and with and without rotated data. To improve test-time robustness, we normalize trajectories and perform rotation-based data augmentation. Our results show that deep networks can achieve a test-set accuracy of up to 100% on a full trajectory, with graceful degradation as the number of time steps decreases, outperforming entropy-based clustering.

Content caching is vital for enhancing web server efficiency and reducing network congestion, particularly in platforms predicting user actions. Despite many studies conducted toimprove cache replacement strategies, there remains space for improvement. This paper introduces STRCacheML, a Machine Learning (ML) assisted Content Caching Policy. STRCacheML leverages available attributes within a platform to make intelligent cache replacement decisions offline. We have tested various Machine Learning and Deep Learning algorithms to adapt the one with the highest accuracy; we have integrated that algorithm into our cache replacement policy. This selected ML algorithm was employed to estimate the likelihood of cache objects being requested again, an essential factor in cache eviction scenarios. The IMDb dataset, constituting numerous videos with corresponding attributes, was utilized to conduct our experiment. The experimental section highlights our model’s efficacy, presenting comparative results compared to the established approaches based on raw cache hits and cache hit rates.

Extensive prior work has provided methods for the optimization of routing based on the criteria of travel time and/or the cost of travel and/or the distance traveled. A typical method of routing involves building a graph comprised of street segments, assigning a normalized weighted value to each segment, and then applying the weighted-shorted path algorithm to the graph to find the best route. Some users desire that the routing suggestion include consideration pertaining to the reduction of risk of encountering violent crime. For example, a user desires a leisurely walk via a safe route from her hotel in an unknown city. Here, we present a method to quantify such user preferences and the risks of encountering crime and to augment the standard routing methods by assigning weights to safety considerations. The proposed method’s advantages, in comparison to other crimeavoidance routing algorithms, include weighting crime types with respect to their potential detrimental value to the user, with temporal qualification and quantification of crime and its statistical aggregation at the geographic resolution down to a city block.

In the dynamic field of urban planning and the context of unprecedented natural events, such as hurricanes, the fast generation of accurate maps from satellite imagery is paramount. While several studies have utilized Generative Adversarial Networks (GANs) for map generation from satellite images, the present work introduces a new approach by integrating contrastive learning into the GAN framework for enhanced map synthesis. Our methodology distinctively employs positive sampling by aligning similar features (e.g., roads) in both satellite images and their corresponding map outputs, and contrasts this with negative samples for disparate elements. This approach effectively replaces the conventional cyclic process in GANs with a more streamlined, unidirectional procedure, leading to improvements in both the quality of the synthesized maps and computational efficiency. We show the effectiveness of our proposed model, offering an advancement in map generation for remote sensing applications.

This study introduces a methodology integrating Zero Trust Architecture (ZTA) principles and Transparent Shaping into an AWS-hosted Online File Manager (OFM) application, enhancing security without substantial code modifications. We evaluate our approach with the Mozilla Observatory, highlighting significant security improvements and outlining a promising direction for applying Transparent Shaping and ZTA in cloud environments.

Dynamic urban development and natural disasters often lead to discrepancies between maps and satellite images, posing challenges for autonomous navigation and smart city planning. Automatic map generation from satellite imagery is crucial for addressing these discrepancies and ensuring accurate geospatial data. While Generative Adversarial Networks (GANs) favored with contrastive learning have been used for this task, limitations such as randomness in feature selection result in suboptimal performance, and MLP-based projection heads lack the expressiveness needed for effective feature generation. To address these limitations, we propose a novel model, KAN-Attn, which incorporates an attention-based query selection mechanism for selecting relevant features in contrastive learning. Also, KAN-Attn is the first application of Kolmogorov-Arnold Network (KAN) to improve feature generation with enhanced expressiveness in map generation. Our innovations allow KAN-Attn to achieve optimal performance in map generation, with experiments on publicly available datasets demonstrating state-of-the-art results based on relevant metrics on perceptual quality.

Over the past several years, due to the progression toward data-driven scientific disciplines, the field of Big Data has gained significant importance. These developments pose certain challenges in the area of efficient, effective, and secure management and transmission of digital information. This paper presents and evaluates a novel Distributed Ledger Technology (DLT) system, Fibereum, in a variety of use-cases, including a DLT-based system for Big Data exchange, as well as the fungible and non-fungible exchange of artwork, goods, commodities, and digital currency. Fibereum’s innovations include the application of non-linear data structures and a new concept of Lazy Verification. We demonstrate the benefits of these novel features for DLT system applications’ cost performance and their added resilience towards cyber-attacks via the consideration of several use cases.

Deep Learning is becoming a very popular tool for generating and reconstructing images. Research has shown that deep learning algorithms can perform cutting-edge restoration tasks for various types of images. The performance of these algorithms can be achieved by training Deep Convolutional Neural Networks (DCNNs) with data from a large sample size. The processing of high-resolution satellite imagery becomes difficult when there are only a few images in a dataset. An approach based on the intrinsic properties of Deep Convolutional Neural Networks (DCNNs) is presented in this paper for the detection and removal of clouds from remote sensing images without any prior training. Our results demonstrated that the algorithm we used performed well when compared to trained algorithms.

Abstract Hippocampus segmentation in brain MRI is a critical task for diagnosis, prognosis, and treatment planning of several neurological disorders. However, automated hippocampus segmentation methods have some limitations. More precisely, hippocampus is hard to visualize through MRI due to the low contrast of the surrounding tissue, also it is a relatively small region with highly variable shape. In this study, we propose a two-stage architecture to first locate the hippocampus and then segment it. We combine a transformer design with CNN based architecture and a customized loss function to segment the hippocampus via an end-to-end pipeline. In the encoding path, the image is passed through a CNN model to generate a feature map. This feature map is then divided into small patches which are passed to a transformer for extracting global contexts. The encoder used here is identical to that of the Vision Transformer image classification model. In the decoding path, the transformer outputs are combined with their corresponding feature maps to enable a precise segmentation of the hippocampus. The proposed architecture was trained and tested on a dataset containing 195 brain MRI from the Decathlon Challenge. The proposed network achieved a Dice value of 0.90±0.200, and 89% mean Jaccard value in this segmentation task. The mean volume difference between generated mask and ground truth is 5% with a standard deviation of 3%. Deploying the proposed method over our in-house data, consisting of 326 MRIs, showed a mean volume difference of 4.4 % with a standard deviation of 3.24%.

Geographic datasets are usually accompanied by spatial non-stationarity – a phenomenon that the relationship between features varies across space. Naturally, nonstationarity can be interpreted as the underlying rule that decides how data are generated and alters over space. Therefore, traditional machine learning algorithms are not suitable for handling non-stationary geographic datasets, as they only render a single global model. To solve this problem, researchers often adopt the multiple-local-model approach, which uses different models to account for different sub-regions of space. This approach has been proven efficient but not optimal, as it is inherently difficult to decide the size of subregions. Additionally, the fact that local models are only trained on a subset of data also limits their potential. This paper proposes an entirely different strategy that interprets nonstationarity as a lack of data and addresses it by introducing latent variables to the original dataset. Backpropagation is then used to find the best values for these latent variables. Experiments show that this method is at least as efficient as multiple-local-model-based approaches and has even greater potential.