Search Results (6,113)

Deep learning has emerged as an effective tool for automatic waste classification, supporting cleaner cities and more sustainable recycling systems. Because environmental protection is central to the United Nations Sustainable Development Goals (SDGs), improving the sorting and processing of everyday waste is a practical step toward this broader objective. In many real-world settings, however, waste is still sorted manually, which is slow, labor-intensive, and prone to human error. Although convolutional neural networks (CNNs) can automate this task with high accuracy, many state-of-the-art models remain too large and computationally demanding for low-cost edge devices intended for deployment in homes, schools, and small recycling facilities. In this work, we investigate lightweight waste-classification models suitable for TinyML deployment while preserving competitive accuracy. We first benchmark multiple CNN architectures to establish a strong baseline, then apply complementary compression strategies including quantization, pruning, singular value decomposition (SVD) low-rank approximation, and knowledge distillation. In addition, we evaluate an RL-guided multi-teacher selection benchmark that adaptively chooses one teacher per minibatch during distillation to improve student training stability, achieving up to 85% accuracy with only 0.496 M parameters (FP32 ≈ 1.89 MB; INT8 ≈ 0.47 MB). Across all experiments, the best accuracy–size trade-off is obtained by combining knowledge distillation with post-training quantization, reducing the model footprint from approximately 16 MB to 281 KB while maintaining 82% accuracy. The resulting model is feasible for deployment on mobile applications and resource-constrained embedded devices based on model size and TensorFlow Lite Micro compatibility. Full article

This paper investigates the evolution of the microstructure, mechanical performances, and corrosion resistance of selective laser melting (SLM) 304 steel under different intermittent stretching deformation step sizes, revealing the underlying evolution patterns. The results indicate that the intermittent deformation step size significantly affects the microstructure and performance of SLM 304 steel. Larger step sizes result in more complete molten pool contours, less deformation of grain and cellular structures, and a lower martensite volume fraction; smaller step sizes lead to distorted molten pools, fragmented grains, exacerbated cellular structure distortion, and increased martensite content. In terms of mechanical performances, tensile strength, nano-hardness, and elastic modulus decrease with increasing step size, while elongation increases accordingly. Corrosion resistance improves with larger step sizes, with specimens exhibiting more complete and thicker oxide films on the surface and superior pitting resistance; continuous stretching specimens exhibit the worst corrosion resistance, while the original specimens are the best. Intermittent deformation optimizes properties by regulating microstructure, providing a basis for the design of high-performance SLM 304 steel. This study provides theoretical support for the design and application of additive manufacturing stainless steel components, facilitating the engineering and industrial application of SLM technology in high-end equipment manufacturing. Full article

Metaheuristic optimization algorithms have attracted considerable research interest for solving complex optimization problems, yet many existing algorithms suffer from premature convergence and an inadequate balance between exploration and exploitation. The Coati Optimization Algorithm (COA) is a recently proposed nature-inspired metaheuristic that models the hunting and escape behaviors of coatis; however, it exhibits limited search diversity and tends to stagnate in local optima on high-dimensional, multimodal landscapes. This paper proposes an Improved Coati Optimization Algorithm (ICOA) that integrates four complementary enhancement strategies: (1) a Dynamic Adaptive Step-Size strategy that combines Lévy flights with Student’s t-distribution perturbations for heavy-tailed exploration; (2) a Population-Adaptive Dynamic Perturbation strategy that incorporates differential evolution operators with fitness-proportional scaling; (3) an Iterative-Cyclic Differential Perturbation strategy that employs sinusoidal scheduling and population-differential guidance; and (4) a Cosine-Adaptive Gaussian Perturbation strategy for refined exploitation with time-decaying intensity. ICOA is evaluated on 29 CEC2017, 10 CEC2020, and 12 CEC2022 benchmark functions across dimensions ranging from 10 to 100, compared against seven state-of-the-art algorithms in each benchmark suite. A statistical analysis using the Friedman test and the Wilcoxon rank-sum test confirms that ICOA achieves overall rank 1 on all three benchmark suites, with Friedman mean ranks of 1.207 (CEC2017, $D=100$), 1.000 (CEC2020, $D=10$), and 2.208 (CEC2022, $D=10$); the CEC2020 result should be interpreted in the context of its low dimensionality. A scalability analysis across four dimensionalities (10D, 30D, 50D, 100D) demonstrates consistent first-place rankings with mean ranks between 1.000 and 1.207. An ablation study and a sensitivity analysis of the strategy activation probability validate the contribution of each individual strategy and the optimality of the 50% activation setting. Furthermore, ICOA achieves the best results on all six constrained engineering design problems tested, with all improvements confirmed as statistically significant ($p<0.05$). Full article

Background/Objectives: Delayed graft function (DGF) remains a common early complication after deceased donor kidney transplantation and is challenging to anticipate using routine pre-implant clinical variables alone. We investigated whether high-throughput Fourier transform infrared (FTIR) spectroscopy of static cold storage preservation fluid (not machine perfusion perfusate) captures biochemical information associated with DGF and warrants further evaluation alongside routine pre-implant clinical predictors. Methods: In this single-center retrospective cohort, we analyzed preservation fluid samples from 56 kidney transplants originating from 49 deceased donors (7 donors contributed two kidneys); DGF occurred in 14/56 (25.0%). Dried-film FTIR spectra were acquired using a plate-based high-throughput accessory, and analyses focused on the fingerprint region (900–1800 cm⁻¹) with prespecified preprocessing and quality control. We developed and compared clinical-only, FTIR-only, and combined predictive models and estimated performance using donor-blinded 5-fold StratifiedGroupKFold cross-validation (grouped by donor code) to prevent leakage across paired kidneys. Results: Donor-blinded discrimination (pooled out-of-fold ROC-AUC) was 0.775 for the clinical-only model, 0.814 for the FTIR-only model, and 0.796 for the combined model; probabilistic accuracy (Brier score; lower is better) was 0.162, 0.194, and 0.177, respectively. Calibration intercepts were negative and slopes were <1, indicating overly extreme risk estimates under strict donor-blinded validation and supporting recalibration prior to deployment. Decision curve analysis suggested a positive net benefit for clinically plausible thresholds. Conclusions: These findings support the feasibility of rapid, low-cost FTIR profiling of routinely available preservation fluid as a proof-of-concept approach for exploratory DGF risk stratification, rather than as a clinically deployable prediction tool. Given the small sample size and the instability of subgroup estimates, the main next steps are external validation in larger multicenter cohorts, prospective workflow studies, and model updating/recalibration. Full article

Foodborne pathogens of Enterobacteriaceae are becoming an increasing global concern, with multidrug-resistant strains posing significant risks to food safety and public health, especially in high-risk products like dairy. This research focused on isolating, biologically characterizing, and genomically profiling new bacteriophages that target key Enterobacteriaceae members as potential biocontrol agents. Eight phages were isolated from wastewater using four bacterial hosts and analyzed through transmission electron microscopy, one-step growth analysis, adsorption kinetics, host range evaluation, whole-genome sequencing, comparative genomics, phylogenetic analysis, proteomic profiling, and virion assembly pathway characterization. All eight isolates exhibited icosahedral heads with contractile tails typical of Myoviridae morphology, demonstrated broad-spectrum lytic activity against 21 bacterial strains (infectivity: 47.6–95.2%), showed high adsorption efficiencies (84.75–99.98%), and had burst sizes ranging from 11 to 166 particles per cell. Genome sizes varied from 103 to 170 kb with coding densities between 92–96%. Importantly, none contained antimicrobial resistance genes, virulence factors, or lysogeny-associated elements, confirming their strictly lytic lifestyles and favorable biosafety profiles. Phylogenetic and comparative analyses indicated mosaic genomic structures influenced by horizontal gene transfer rather than host phylogeny. These findings provide a robust biological and genomic basis for evaluating these phages as potentially safe and effective alternatives to antibiotics in controlling foodborne Enterobacteriaceae, pending further in situ validation. Full article

Autonomous guided vehicle (AGV) navigation in high-density warehouses faces significant challenges due to narrow aisles and complex U-shaped traps. In such environments, traditional sampling-based path planning algorithms often converge slowly and produce suboptimal paths. To solve these issues, a novel hierarchical hybrid planning framework named MEG-RRT* (Multi-objective Evolutionary Guided RRT*) is proposed in this study. The proposed MEG-RRT* integrates an optimization engine based on NSGA-II into the sampling process. It guides exploration direction away from local minima by jointly optimizing convergence efficiency and safety-related objectives. Furthermore, a geometry-aware execution layer is introduced to improve motion through narrow passages and to refine the path structure. This layer includes radar-guided steering, adaptive step-size control, and ancestor shortcut operations. Comparative experiments were conducted in simulated scenarios of complex narrow passages and high-density warehouses to verify the superiority of the proposed MEG-RRT*. In complex narrow passages, the proposed algorithm achieves a 100% success rate; it also reduces convergence time by 43.5% compared to standard RRT* and by 44.9% compared to Informed-RRT*. In warehouse environments, it generates smooth, kinematically favorable paths that are 39% shorter than those produced by RRT-Connect. These results demonstrate that MEG-RRT* balances exploration efficiency and solution optimality, making it well suited for automated logistics applications. Full article

Deep learning has become central to semantic segmentation of three-dimensional medical images. However—despite many published models—their adoption in practice remains limited, as deployment often requires advanced programming skills and familiarity with specific machine learning frameworks. Thus, technical barriers restrict its use to specialized users. To address this, we present SegMed (version 1.0), an open-source, standalone desktop application that provides an end-to-end workflow for deep learning-based medical image segmentation. SegMed supports the loading and inspection of common medical image formats, as well as array-based formats. The application integrates standard preprocessing operations often used in the field and directly supports loading of pretrained segmentation models implemented in both PyTorch (version 2.X) and Keras (version 2.X) and those created using the Medical Open Network for AI framework (version 1.X). Models are automatically inspected to infer required configurations, such as input size and post-processing steps, enabling segmentation with minimal user intervention. Results can be exported as volumetric images or 3D surface meshes for downstream analysis, visualization, or special applications such as virtual reality. SegMed was tested using multiple publicly available pretrained models, demonstrating robustness and flexibility across diverse segmentation tasks. By abstracting low-level implementation details, SegMed lowers technical barriers, promotes reproducibility, and facilitates the integration of AI-assisted segmentation into medical imaging workflows. Full article

Current accounting standards do not allow recognition of intangible assets for indigenously created properties, resulting in a discrepancy between the book value and market value of firms operating within digital economies, where investments like cybersecurity and data governance are grossed up immediately on the statement of financial position as they are considered to be expensed under IFRS. This paper investigates whether voluntary Digital Reputation Risk Disclosure (DRRD) rectifies this valuation gap for the non-financial firms listed on the Saudi Exchange. Based on an automated bilingual dictionary-based textual analysis of 891 corporate documents and a two-step System GMM estimator run on an unbalanced panel of 619 firm-year observations from a sample of 132 firms for the period 2020–2024, we show that DRRD is statistically significantly negatively related to firm value at conventional levels, implying that investors perceive such disclosures as indications of higher risk exposure rather than stronger governance capabilities. While statistically insignificant, the moderating effect of firm size shows that negative valuation effects are concentrated on large firms according to sub-sample analysis. These findings are confirmed across several alternative specifications in the robustness checks. The findings demonstrate that voluntary digital risk disclosure, in the absence of standards-based frameworks, is not effective at bridging this valuation gap, and may instead activate functional fixation among investors. These findings highlight the importance of IASB’s standardization agenda regarding intangible assets and present relevant empirical data for developing capital markets. Full article

Real-time motion planning in cluttered and dynamically evolving environments remains challenging due to the need to ensure rapid convergence, collision avoidance, computational efficiency, and robustness against local minima under frequent changes. Although sampling-based planners such as RRTX* and ABIT* provide strong theoretical guarantees, their practical deployment in dense dynamic scenarios is often limited by high sampling overhead and computational latency. This paper proposes a Fast Converging Adaptive Algorithm (FCAA), a deterministic sampling-based framework integrating adaptive sampling density, temperature-controlled exploration, and dynamic step-size regulation within a unified heating and annealing mechanism. The temperature parameter governs both the spatial sampling band and incremental expansion radius, enabling controlled transitions between goal-directed expansion and stochastic exploration when stagnation occurs. The algorithm is evaluated using a two-stage protocol comprising intrinsic validation and benchmarking. Across 36 environments with obstacle densities ranging from 3% to 20% and velocities between −30 and +30 m/s, FCAA achieved a 100% success rate within the defined experimental design while maintaining path quality comparable to or better than RRTX* and ABIT*. Unlike the reference planners, which typically required tens of thousands of samples and seconds of computation, FCAA operated with substantially reduced sampling effort, typically tens of nodes, and planning times from 0.1 to 320 ms depending on scenario complexity. Within the simulation framework, the results indicate that the proposed temperature-regulated strategy enables fast and computationally efficient motion planning under dynamic constraints, making FCAA suitable for time-critical robotic navigation scenarios. Full article

Aiming at the inherent defects of the Horned Lizard Optimization Algorithm (HLOA), such as insufficient global exploration capability, premature convergence to local optima, and inadequate balance between exploration and exploitation, this paper proposes an enhanced Improved Horned Lizard Optimization Algorithm (IHLOA) integrated with multi-strategy improvements. Firstly, the Fuch chaotic mapping is introduced for population initialization, which enhances the ergodicity and diversity of the initial population by leveraging the pseudo-random and aperiodic characteristics of chaotic sequences, laying a high-quality foundation for subsequent optimization searches. Secondly, the golden sine strategy is embedded into the iterative update process to dynamically adjust the search step size and direction. This strategy utilizes the periodic amplitude variation in the sine function and the golden section coefficient to balance the global exploration for potential optimal regions and local exploitation for refined optimization, thereby accelerating convergence speed while avoiding local stagnation. Finally, the orthogonal crossover strategy is incorporated in the late iteration stage to promote effective information interaction between parent and offspring populations. By means of chromosome segment exchange and elitist retention mechanisms, this strategy reduces dimensional search blind spots and further enhances the algorithm’s ability to capture high-quality solutions. Comprehensive experimental evaluations are conducted based on classical benchmark test functions and eight state-of-the-art meta-heuristic algorithms. The results demonstrate that the IHLOA outperforms comparative algorithms in terms of optimization accuracy, convergence speed, and stability across 30-D, 50-D, and 80-D scenarios. For practical path planning applications, the IHLOA achieves remarkable performance improvements: in single-goal path planning, it reduces the path length by 2.54–87.64% compared with benchmark algorithms; in multi-goal path planning, it realizes a 1.24–7.99% reduction in path length and an 11.91% average reduction in the number of turning points relative to the original HLOA. Additionally, the IHLOA exhibits excellent robustness and adaptability in dynamic obstacle environments, effectively shortening the path length and reducing robot stuck times. This research not only enriches the improvement framework of meta-heuristic algorithms but also provides a high-efficiency optimization solution for mobile robot path planning in complex environments. Full article

Statistical inference in experimental neuroscience is routinely detached from the experimental record: analytic choices are reported in prose summaries that do not expose the code, assumptions, or decision pathways that produced the results. This detachment limits reproducibility and impairs peer review. Here, we describe the Companion Statistical Report (CSR), a structured, versioned document format designed to accompany empirical neuroscience manuscripts as peer-reviewed as a peer-reviewed resource. The CSR integrates data provenance, preprocessing decisions, exploratory analyses, model specifications, assumption diagnostics, inference with effect sizes, and sensitivity analyses into a single executable document, authored in Quarto 1.8 and supporting both R and Python workflows. We provide an open template hosted at on GitHub that implements this format with institutional branding, parameterization, and version tracking. The template was developed by the Bertrand Russell Research Excellence Group (NEC) at the School of Medicine, Rio de Janeiro State University. By making analytic choices auditable and reproducible by design, CSRs are designed to reduce the gap between what neuroscience experiments measure and what published statistics claim, offering a tractable and immediately implementable step toward greater transparency. Full article

The object of the study is a single heated carbonaceous particle of relatively small size, 0.003 to 0.01 m. Main hypothesis: The formation of soot particles and black carbon particles is caused by the thermochemical destruction of dry organic matter of forest fuel and the mechanical fragmentation of coke residue. The aim of the study is to conduct numerical simulations of heat and mass transfer in a single heated carbonaceous particle, taking into account the soot formation process and assessing its fragmentation with regard to heat exchange with the external environment in a 2D setting. As part of this study, a new model of heat and mass transfer in a pyrolyzed carbonaceous particle was developed, taking into account its step-by-step fragmentation (fragmentation tree model with four secondary particle formations from the initial particle). The calculations resulted in the distributions of temperature and volume fractions of phases in the carbonaceous particle across various scenarios. Scenarios of surface fires (initial temperatures of 900 K and 1000 K), crown fires (1100 K), and a firestorm (1200 K) for typical vegetation (pine, spruce, birch) are considered. Cubic carbonaceous particles are considered in the approximation of a 2D mathematical model. To describe heat and mass transfer in the structure of the carbonaceous particle, a differential equation of thermal conductivity with corresponding initial and boundary conditions of the third type is used, taking into account the gross reaction in the kinetic scheme of pyrolysis and soot formation. Differential analogues of partial differential equations are solved using the finite difference method of second-order approximation. Options for using the developed mathematical model and probabilistic fragmentation criterion for assessing aerosol emissions are proposed. Recommendations: The suggested mathematical model must be incorporated with mathematical models of forest fire plume and aerosol transport in the upper layers of the atmosphere. Moreover, probabilistic criteria for health assessment must be developed for the practical use of the suggested mathematical model. Full article

Carbon molecular sieve membranes (CMSMs) provide a means to greatly improve gas separations. CMSMs have a pore size distribution comprising micropores and ultramicropores which provide high flux and high selectivity, enabling them to outperform polymeric membranes. CMSMs, however, suffer from physical aging that results from the collapse of the pores that severely reduces their gas permeability. Therefore, reducing physical aging of CMSMs is an important step in the development of these types of materials. In this work, a method for reducing physical aging through the incorporation of metal nanoparticles that serve as a structural scaffold for the membrane pore structure is presented. The pore structure in CMSMs is dependent upon the polymeric precursor, and thus the support system incorporated must be tailored. The copper nanoparticles were formed in situ from soluble, copper-based metal–organic polyhedra 18 (MOP-18) dispersed into Matrimid^® 5218, a low free volume polymer. The size (2 to 20 nm) and shape (sphere, rods) of the copper particles were refined by adjusting MOP-18 loading, pyrolysis temperature, and soaking time. The Cu-pillared Matrimid^® 5218 CMSMs from this work showed no decline in permeability or selectivity for methane (107 Barrer) and carbon dioxide (1785 Barrer) over a period of 21 d. The results suggest that tailored metal pillars can suppress physical aging in CMSMs, thereby enhancing their long-term stability and applicability in gas separations. Full article

Red Amaranth (RA) Azo dye is a persistent pollutant in wastewater and stands as a toxicological risk, which has led to the development of effective methods for its removal and photocatalytic degradation. Therefore, CeO₂ nanoparticles were synthesized by a controlled precipitation method, and Ultraviolet-Visible (UV–Vis) analysis and Tauc plots yielded a band gap of ~3.24 eV. The CeO₂ nanoparticles showed the fluorite cubic phase, and nearly spherical particles with an average size of ~10 nm. Nitrogen physisorption revealed a type IV isotherm with a Brunauer–Emmett–Teller (BET) surface area of 85.27 m²·g⁻¹ and a total pore volume of 0.27 cm³·g⁻¹, indicating a mesoporous structure and high surface accessibility. The chemical behavior showed Ce and O, consistent with phase purity. Photocatalytic performance was evaluated in 20 ppm aqueous solution of RA under 365 nm UV irradiation (LED 100 W), with a temperature of ~20 °C and a 15 min dark adsorption step. Concentration decay was followed at λ_max = 520 nm by Lambert–Beer. The degradation efficiency η and pseudo-first-order kinetic were obtained from ln(C₀/C_t) vs. time. In addition, chemical oxygen demand (COD) tests were performed on RA solution before and after photodegradation, showing a COD reduction of ~85% (from 19.8 to 3 mg O₂·L⁻¹), which corroborates mineralization beyond chromophore bleaching. Under [C₀ = 20 mg·L⁻¹] and [m_cat = 1.0 g·L⁻¹], CeO₂ achieved [RA = 90% at 180 min, k = 0.0125 min⁻¹]. These results demonstrate that CeO₂ is an effective photocatalyst for RA degradation under UV-A irradiation, integrating adsorption, kinetic behavior, and mineralization performance into a coherent structure–property relationship. Full article

Fungal pigments have gained attention as eco-friendly and versatile materials for green nanotechnology because of their varied chemical structures, inherent redox properties, and strong metal ion-binding capabilities. These pigments, such as polyketides, azaphilones, melanins, and carotenoids, can function simultaneously as reducing, capping, and surface-functionalizing agents, facilitating the environmentally friendly production of metallic nanoparticles without the use of harmful chemicals. This review provides a critical overview of recent progress in the production, extraction, and application of fungal pigments for nanoparticle synthesis, focusing on the mechanistic roles of pigment functional groups in metal ion reduction, nanoparticle nucleation, growth, and stabilization. The impact of pigment chemistry and reaction conditions on the nanoparticle size, shape, crystallinity, and colloidal stability was thoroughly examined. Additionally, this review highlights the emerging biomedical, environmental, and industrial applications of pigment-mediated nanoparticles, emphasizing their biocompatibility and functional adaptability. Key challenges, such as variability in pigment yield and composition, limited mechanistic validation, lack of standardized synthesis protocols, and insufficient toxicity assessment, are critically analyzed in this review. Finally, future directions are outlined, emphasizing the importance of process optimization, omics-guided pigment discovery, and comprehensive safety evaluations as crucial steps toward the scalable and reliable use of fungal pigment-mediated nanoparticle synthesis in sustainable nanotechnology. Full article