This paper introduces the Global-Local Image Perceptual Score (GLIPS), an image metric designed to assess the photorealistic image quality of AI-generated images with a high degree of alignment to human visual perception. Traditional metrics such as FID and KID scores do not align closely with human evaluations. The proposed metric incorporates advanced transformer-based attention mechanisms to assess local similarity and Maximum Mean Discrepancy (MMD) to evaluate global distributional similarity. To evaluate the performance of GLIPS, we conducted a human study on photorealistic image quality. Comprehensive tests across various generative models demonstrate that GLIPS consistently outperforms existing metrics like FID, SSIM, and MS-SSIM in terms of correlation with human scores. Additionally, we introduce the Interpolative Binning Scale (IBS), a refined scaling method that enhances the interpretability of metric scores by aligning them more closely with human evaluative standards. The proposed metric and scaling approach not only provide more reliable assessments of AI-generated images but also suggest pathways for future enhancements in image generation technologies.

A novel antenna technology referred to as dynamic metasurface antennas (DMA) that employs reconfigurable metamaterials as its radiating/receiving elements has emerged as a potential alternative to traditional antennas. It specifically helps to simplify the hardware complexities of transceiver structures by reducing the number of radio frequency (RF) chains, thus lowering the operational cost of multiple-input multiple-output (MIMO) systems. To assess its usefulness, this study analyzes the capacity characterization of different DMA-based multiuser MIMO systems by considering uplink and downlink communications when DMA is deployed at the base station or the users. In each case, we formulate a capacity maximization problem. Further, using either reduced minimum mean square error or iterative waterfilling techniques, we propose efficient algorithms to optimize transmit precoders and DMA configurable weights alternately. The solution for each optimizable variable is designed to attain fast convergence that ensures the overall low complexity of the proposed scheme. The simulation results demonstrate the effectiveness of the proposed algorithms in achieving performance comparable to conventional MIMO while significantly reducing the number of RF chains.

We show the Optical-Butterfly-Phenomena: the small differences in the structures of slits lead to profoundly different patterns.

The clothing industry demands unending novelty in attire designs to meet developing client advantages. However, established forms of founding attire patterns are labor-exhaustive, high-priced, and frequently impassable to tiny designers and arising fashion brands. FashionFusion presents an AI-compelled answer that automates the production of singular and superior attire patterns, leveraging the capacity of deep education and fruitful models. By streamlining the design process, this method allows designers to speedily original creative patterns, lowering two together opportunity and cost. The manifesto democratizes approach to progressive design finishes, enabling consumers to forge embodied designs outside the need for thorough mechanics knowledge. With this resolution, the impediments to fashion artistry are considerably reduced, promoting a more all-embracing and vital design environment.

In this paper, we identify optical reconfigurable intelligent surfaces (ORISs) as key enablers of next-generation free-space optical (FSO) multiple access systems. By leveraging their beam steering and beam splitting capabilities, ORISs are able to effectively address line-of-sight (LoS) constraints, while enabling multiuser connectivity. We consider an ORIS-assisted non-orthogonal multiple access (NOMA) system model consisting of a single transmitter (Tx) and two receivers (Rxs). We derive novel analytical expressions to characterize the statistical particularities of the Tx-ORIS-Rx communication channel. Building upon the aforementioned expressions, we investigate the outage performance of the Rxs by deriving exact analytical expressions for the outage probability (OP) of each Rx. To provide deeper insights into the impact of various system parameters and physical conditions on the outage performance of each Rx, we conduct a high signal-to-noise ratio (SNR) analysis, that returns asymptotic expressions for the Rxs' OPs at the high-SNR regime. Monte Carlo simulations validate the analysis, demonstrate the effectiveness of ORIS-enabled NOMA under a variety of configurations and physical scenarios, and showcase its superiority over its orthogonal-based counterpart.

The advent of multi-unmanned aerial vehicle (multi-UAV) networks in mobile edge computing (MEC) introduces dynamic computational topologies where UAVs, acting as mobile edge servers, are tasked with processing data from ground-based user equipment (UE). This paper addresses the dual challenges of optimizing both UAV deployment and task offloading within such networks to minimize communication latency and efficiently utilize UAV resources, which are limited by battery life and processing capabilities. We propose a bi-level optimization framework that simultaneously tackles the placement of UAVs and the distribution of computational tasks among them. At the higher level, UAV deployment is optimized to ensure minimal distance to the UEs, thereby reducing latency and energy consumption during data transmission. At the lower level, task offloading is optimized to balance the computational load across the UAV network, considering each UAV's capacity and battery constraints. We demonstrate through extensive simulations the significant improvements in system efficiency, latency, and resilience. This approach not only enhances the performance of UAV-assisted MEC networks but also provides scalable solutions adaptable to various operational scenarios.

A nonlinear electromagnetic inversion scheme incorporating a deep-learning-based plug-andplay (PNP) regularization approach is proposed. Unlike conventional regularization techniques that rely on fixed priors, the PNP method decouples the data fidelity and regularization terms, enabling the integration of a learned prior via a state-of-the-art denoiser. The recently introduced Swin-Conv-UNet (SCUNet), known for its superior image denoising capabilities by leveraging swin transformer blocks and residual convolutional blocks in its architecture, is employed as the PNP denoiser within a multi-frequency Gauss-Newton nonlinear inversion framework. The resulting PNP-SCUNet inversion scheme is evaluated on both synthetic and experimental data, demonstrating superior performance compared to traditional inversion methods, including Tikhonov and total variation, and previously developed PNP regularization with block-matching and 3-D filtering (BM3D) denoiser.

Principal Component Analysis (PCA) is a widely adopted technique for dimensionality reduction and data exploration. In this paper we have implemented from scratch and applied to both a synthetically generated random dataset and the Abalone dataset to evaluate its effectiveness. PCA, a vital dimensionality reduction technique, simplifies complex datasets, enhances interpretability, and improves computational efficiency while preserving key patterns. The study involves constructing a random dataset to validate the robustness of the custom PCA algorithm via eigenvalue decomposition of the covariance matrix. This approach is then extended to the Abalone dataset to extract principal components, analyze variance contributions, and visualize the data in reduced dimensions. Results demonstrate that the handcrafted PCA algorithm effectively captures the underlying structure of the data, providing critical insights into feature importance and enabling efficient data visualization. This dual application highlights the versatility and efficacy of PCA in handling diverse datasets, confirming its utility in various data analysis scenarios.

Large Language Models (LLMs) have shown remarkable performance in automated code generation. However, existing approaches often rely heavily on pre-defined test cases, which become impractical in scenarios where such cases are unavailable. While prior works explore filtering techniques between programs and test cases, they overlook the refinement of test cases. To address this limitation, we introduce CoCoEvo, a novel LLM-based co-evolution framework that simultaneously evolves programs and test cases. CoCoEvo eliminates the dependency on pre-defined test cases by generating both programs and test cases directly from natural language problem descriptions and function headers. The framework employs specialized evolutionary operators, including LLM-based crossover and mutation operators for program evolution, along with a test case generation operator for test case evolution. Additionally, we propose optimization strategies such as a crossover rate scheduler to balance exploration and convergence, and a multi-objective optimization method for test case selection. Experimental results on multiple state-of-the-art LLMs demonstrate that CoCoEvo surpasses existing methods, achieving state-of-the-art performance in automated code generation and testing. These results underscore the potential of co-evolutionary techniques in advancing the field of automated programming.

We explored Naïve Bayes classification methods on the Page Blocks Classification Dataset, which categorizes page blocks into text, horizontal line, graphic, vertical line, and picture. Utilizing both Gaussian Naïve Bayes and Categorical Naïve Bayes classifiers, we evaluated their performance on this multi-class classification problem. The dataset, with 5473 instances and 10 attributes, presents a practical challenge for distinguishing various page block types in document images. We compared the accuracy, precision, recall, and F1-score of the classifiers across various scenarios. Results show that the Gaussian Naïve Bayes classifier slightly outperforms the Categorical Naïve Bayes in accuracy and F1-score, likely due to the continuous nature of the dataset's attributes. Additionally, the Gaussian classifier demonstrates faster processing times. Our findings suggest that while both models are viable, the Gaussian Naïve Bayes classifier may be more suitable for similar datasets. Future research directions include integrating feature selection techniques and exploring hybrid models.

This paper addresses the challenge of wireless power transfer to implantable endovascular devices, which have been used for pressure monitoring and tissue stimulation. However, a thorough analysis and optimization of the wireless link remain unexplored, leaving the maximum power efficiency and power delivery unknown. Methods: Theoretical analysis, computer simulations, and experimental measurements with biological tissues were combined. Results: Different coils were designed and optimized for endovascular devices and compared. With proper geometry, material, and frequency selection for coils, power efficiency over 10% and power delivery over 100 mW was achieved at 15 mm under human safety limits. Conclusion: This work demonstrates an optimized results and provides some key insights for endovascular coil design. Significance: The designed inductive link enables a minimally invasive, durable, and powerefficient solution for future endovascular applications.

In this paper, we present CHRONOS-a tool that enables the querying, exploration, and discovery of historical change events to source code. Unlike traditional Revision-Control-System tools, CHRONOS allows queries across any subset of the code, down to the line-level, which can potentially be contiguous or disparate, even among multiple files. In addition, CHRONOS provides change history across all historical versions (i.e., it is not limited to a pairwise "diff"). The tool implements a zoom-able user interface as a visualization of the history of the queried code to provide both a high-level view of the changes, which supports pattern recognition and discovery, and a lowlevel view that supports semantic comprehension for tasks such as reverse engineering and identifying design rationale. In this paper, we describe use cases in which CHRONOS may be helpful, provide a motivating example to demonstrate the benefits brought by CHRONOS, and describe its visualization in detail.

Solar energy is a ubiquitous and free-of-charge source of renewable energy. When combined with other sources and applied to drones, this source of energy can significantly enhance drone endurance. In this paper, a novel approach is used to model photovoltaic solar cells under various environmental conditions. Photovoltaic (PV) cells are modeled and simulated using double and three-diode models with the ant colony optimization (ACO) algorithm to study the behavior of PV cells at different temperatures. This research uses a CdTe thin-film cell whose experimental data are extracted from the SCAPS simulation program at temperatures of 250 – 310 K and irradiance of 1000 W/m2. In 30 runs and 200 iteration steps with a population of 30 ants, the best root-mean-square error (RMSE) value is 5.2936 × 10-5 and 1.711 × 10-5 in double and three-diode cases, respectively. This shows that efficiency and performance of photovoltaic solar cell embedded in drones has improved drastically compared to other similar researches, which indicates the considerable superiority of the proposed method over many other algorithms developed so far. In each iteration step, the average time is 2.304 and 2.612 s in double and three-diode models, respectively. The number of investigated points is between 83 – 93 points depending on the temperature of the experiment in descending order. Moreover, the corresponding optimization has been performed on a 30-core server with 32GB RAM.

In this paper Distribution of Relaxation Times (DRT) method is introduced for analyzing aging and failure conditions in lithium-ion (Li-ion) batteries, addressing challenges associated with its implementation. While Electrochemical Impedance Spectroscopy (EIS) and Equivalent Circuit Models (ECM) are commonly used to monitor battery performance, their interpretation is often complicated by overlapping semicircles in impedance spectra. The DRT technique resolves this issue by deconvolving relaxation times, enabling the separation of individual electrochemical processes and providing a clearer understanding of aging and failure conditions. The peaks of lower frequency components in DRT plots, specifically the charge transfer and diffusion processes, are key indicators of the battery failure point. When these two processes merge, it signals that the battery can no longer function, marking a critical failure point in Li-ion batteries. Identifying failure conditions and aging in Li-ion batteries using DRT offers a more advanced approach compared to ECM, as it delivers greater detail in the electrochemical processes that contribute to performance degradation. The analysis of two kinds of different Lithium-Ion battery cells based on the DRT reveals the specific aging and failure patterns, particularly in later battery life stages. The findings demonstrate the potential of DRT as a real-time indicator to monitor the status and the lifecycle of the battery.

With rapid digitization and digitalization, drawing a fine line between the digital and the physical world has become nearly impossible. It has become essential more than ever to integrate all spheres of life into a single Digital Thread to address pressing challenges of modern society – accessible and inclusive healthcare in terms of equality, and equity. The techno-social advancements and mutual acceptance have enabled the infusion of digital models to simulate social settings with minimum resource utilization to make effective decisions. However, there is a significant gap in feeding back the models with appropriate changes in real-time. In other words, an active behavioral modeling of modern society is lacking, that influences community healthcare as a “whole”. By creating virtual replicas of physical systems, digital twins can enable real-time monitoring, simulation, and optimization of urban dynamics. This paper explores the potential of digital twins to promote inclusive healthcare for evolving smart cities. We argue that digital twins can be used to:Identify and address disparities in access to healthcare servicesFacilitate community participationSimulate the impact of urban policies and interventions on different groups of peopleAid policy making bodies for better access to healthcareThis paper proposes several proposals for stitching the actual and virtual societies using digital twins. Several discussed concepts within this framework envision an active, integrated, and synchronized community aware of data privacy and security. The proposal also provides high- level step-wise transitions that will enable this transformation.

This work presents an analysis of the influence of connector repeatability in three different methods for estimating the propagation constant of transmission lines from twoport measurements. For this purpose, the repeatability of 16 transitions using 1.85 mm coaxial-to-microstrip end-launcher connectors has been tested. It has shown that using the same pair of connectors instead of the whole set significantly reduces the standard deviation of the transition S-parameters that affects the final estimation of the propagation constant, and especially the attenuation constant. In addition, the hypothesis that the measured data have a normal probability distribution has been validated by performing an Anderson-Darling test on the estimated S-parameters of the transition. The obtained standard deviation has been included in a sensitivity analysis, generating S-parameters from normal distribution and performing a Monte Carlo simulation. The objective is to study the standard deviation of the propagation constant obtained using the proposed methods when there are errors related to connector repeatability. In this case, unlike random errors of the analyzer, it has been found that all the compared strategies for the estimation of the propagation constant (traces, eigenvalues, and determinants) work in the same way concerning launcher repeatability errors. Furthermore, it has been seen that the propagation constant obtained also follows a normal distribution. Finally, to validate the presented theory, methods have been applied to several measurements of two lines in the 0.01-to 67-GHz frequency range, using the same kit and different combinations of different connectors. Results show that higher accuracy is obtained when using the same pair of connectors, considerably reducing attenuation constant ripple, which assesses the suitability of the proposed error analysis.

L3 autonomy represents a significant advancement in self-driving technology, paving the way for fully autonomous vehicles. The significance of this research is rooted in its identification of substantial opportunities for further exploration in the domain of autonomous vehicle technology. L3 autonomous driving requires a 'handover' of control from the vehicle to the driver under certain circumstances, which is critical for ensuring safety and reliability. The foremost challenge for L3 autonomous vehicles is ensuring a seamless transition of control between the human driver and the autonomous system. This requires developing a robust and reliable takeover request strategy that accurately accounts for driver performance and experience during all takeover phases. Enhancing safety and improving the driving experience would build trust in technology. This paper addresses current challenges and sets the foundation for future research by proposing an advanced personalized framework for driver takeovers. This framework aims to improve the reliability and safety of control transitions between human drivers and autonomous systems. Additionally, it provides a structured methodology for future studies to build upon, ultimately contributing to the broader adoption and enhancement of autonomous driving technologies.

We evaluated variable autonomy, i.e., who controls the shift between the levels of automation (LoA) for a teleoperated robot navigation task in a simulated office environment. Two variable autonomy control modes were defined-Human-Initiated (HI) and Robot-Initiated (RI) to examine if operators' cognitive workload, situational awareness (SA), and task performance vary between modes. Two levels of automation were used: remotely monitoring (high LoA) and teleoperating (low LoA) the robot while attending to secondary tasks. Task performance metrics included task completion time, collision rates, time spent in the low LoA, and secondary task performance. Operatorrelated measures included workload (NASA-TLX), trust, and perceived situation awareness (SA). Results indicate that the RI control mode outperformed HI in efficiency and reduced workload but compromised operators' SA. Conversely, the HI mode enhanced operator awareness, highlighting a tradeoff between efficiency and situational engagement. This work contributes to understanding the role of variable autonomy in human-robot collaboration and emphasizes the importance of evaluating operator performance in addition to task performance for designing effective teleoperation systems in habituated environments.