Search Results (4,305)

Early printed books, particularly incunabula, are invaluable archives of the beginnings of modern educational systems. However, their complex layouts, antique typefaces, and page degradation caused by bleed-through and ink fading pose significant challenges for automatic transcription. In this work, we present a modular pipeline that addresses these problems by combining modern layout analysis and language modeling techniques. The pipeline begins with historical layout-aware text segmentation using Kraken, a neural network-based tool tailored for early typographic structures. Initial optical character recognition (OCR) is then performed with Kraken’s recognition engine, followed by post-correction using a fine-tuned ByT5 transformer model trained on manually aligned line-level data. By learning to map noisy OCR outputs to verified transcriptions, the model substantially improves recognition quality. The pipeline also integrates a preprocessing stage based on our previous work on bleed-through removal using robust statistical filters, including non-local means, Gaussian mixtures, biweight estimation, and Gaussian blur. This step enhances the legibility of degraded pages prior to OCR. The entire solution is open, modular, and scalable, supporting long-term preservation and improved accessibility of cultural heritage materials. Experimental results on 15th-century incunabula show a reduction in the Character Error Rate (CER) from around 38% to around 15% and an increase in the Bilingual Evaluation Understudy (BLEU) score from 22 to 44, confirming the effectiveness of our approach. This work demonstrates the potential of integrating transformer-based correction with layout-aware segmentation to enhance OCR accuracy in digital humanities applications. Full article

Cancer remains one of the most significant global health challenges due to its high mortality rates and the limited understanding of its progression. Early diagnosis is critical to improving patient outcomes, especially in skin cancer, where timely detection can significantly enhance recovery rates. Histopathological analysis is a widely used diagnostic method, but it is a time-consuming process that heavily depends on the expertise of highly trained specialists. Recent advances in Artificial Intelligence have shown promising results in image classification, highlighting its potential as a supportive tool for medical diagnosis. In this study, we explore the application of hybrid Artificial Intelligence models for melanoma diagnosis using histopathological images. The dataset used consisted of 506 histopathological images, from which 313 curated images were selected after quality control and preprocessing. We propose a two-step framework that employs an Autoencoder for dimensionality reduction and feature extraction of the images, followed by a classification algorithm to distinguish between melanoma and nevus, trained on the extracted feature vectors from the bottleneck of the Autoencoder. We evaluated Support Vector Machines, Random Forest, Multilayer Perceptron, and K-Nearest Neighbours as classifiers. Among these, the combinations of Autoencoder with K-Nearest Neighbours achieved the best performance and inference time, reaching an average accuracy of approximately 97.95% on the test set and requiring 3.44 min per diagnosis. The baseline comparison results were consistent, demonstrating strong generalisation and outperforming the other models by 2 to 13 percentage points. Full article

Alperujo, a solid by-product from the two-phase olive oil extraction process, poses significant environmental challenges due to its high organic load, phytotoxicity, and phenolic content. At the same time, it represents a promising feedstock for recovering value-added compounds such as phenols and volatile fatty acids (VFAs). When used as a substrate for white rot fungi (WRF), it also produces ligninolytic enzymes. This study explores the use of two native WRF, Anthracophyllum discolor and Stereum hirsutum, for the biotransformation of alperujo under solid-state fermentation conditions, with and without supplementation of copper and manganese, two cofactors known to enhance fungal enzymatic activity. S. hirsutum stood out for its ability to release high concentrations of phenolic compounds (up to 6001 ± 236 mg gallic acid eq L⁻¹) and VFAs (up to 1627 ± 325 mg L⁻¹) into the aqueous extract, particularly with metal supplementation. In contrast, A. discolor was more effective in degrading phenolic compounds within the solid matrix, achieving a 41% reduction over a 30-day period. However, its ability to accumulate phenolics and VFAs in the extract was limited. Both WRF exhibited increased enzymatic activities (particularly Laccase and Manganese Peroxidase) with the addition of Cu-Mn, highlighting the potential of the aqueous extract as a natural source of biocatalysts. Phytotoxicity assays using Solanum lycopersicum seeds confirmed a partial detoxification of the treated alperujo. However, none of the fungi could entirely eliminate inhibitory effects on their own, suggesting the need for complementary stabilization steps before agricultural reuse. Overall, the results indicate that S. hirsutum, especially when combined with metal supplementation, is better suited for valorizing alperujo through the recovery of bioactive compounds. Meanwhile, A. discolor may be more suitable for detoxifying the solid phase strategies. These findings support the integration of fungal pretreatment into biorefinery schemes that valorize agroindustrial residues while mitigating environmental issues. Full article

The increasing need for sustainable valorization of fossil-based and waste-derived materials has gained interest in converting complex organic matrices such as kerogen into valuable chemicals. This study explores a two-step oxidative strategy to decompose and valorize kerogen-rich oil shale, aiming to develop a locally based source of aliphatic dicarboxylic acids (DCAs). The method combines air oxidation with subsequent nitric acid treatment to enable selective breakdown of the organic structure under milder conditions. Air oxidation was conducted at 165–175 °C using 1% KOH as an alkaline promoter and 40 bar oxygen pressure (or alternatively 185 °C at 30 bar), targeting 30–40% carbon conversion. The resulting material was then subjected to nitric acid oxidation using an 8% HNO₃ solution. This approach yielded up to 23% DCAs, with pre-oxidation allowing a twofold reduction in acid dosage while maintaining efficiency. However, two-step oxidation was still accompanied by substantial degradation of the structure, resulting in elevated CO₂ formation, highlighting the need to balance conversion and carbon retention. The process offers a possible route for transforming solid fossil residues into useful chemical precursors and supports the advancement of regionally sourced, sustainable DCA production from unconventional raw materials. Full article

Grapholita molesta is a globally significant fruit pest. Females achieve maximal reproductive output through efficient sperm utilization following a single copulation. Post-mating maturation of eupyrene sperm is a critical step in reproductive success. Here, we report that a male accessory gland-derived serine protease (named GmAGSP1) is essential for this process. GmAGSP1 was only distantly related to other identified sperm-activating SPs, and its transcript was highly expressed in the AG at 48 h after emergence. RNAi-mediated knockdown of GmAGSP1 in males did not affect courtship rate, copulation duration, or mating frequency, whereas male fertility decreased significantly. Mating with GmAGSP1-knockdown males markedly impaired eupyrene sperm maturation in the spermatophores, with phenotypes including failure of eupyrene sperm bundles to dissociate normally and marked reduction in viability of the dissociated eupyrene sperm. Finally, untargeted metabolomic analysis preliminarily demonstrated marked alterations in multiple metabolic pathways within the spermatophore following mating with GmAGSP1-knockdown males. This study advances our understanding of the regulatory mechanism of “sperm activation in the spermatophore’s metabolic microenvironment mediated by male AG-derived SP” while providing critical insights for the development of novel genetic control strategies targeting G. molesta. Full article

This research presents a reinforcement learning framework for stable quadruped locomotion using Proximal Policy Optimization (PPO). We address critical challenges in articulated robot control—including mechanical complexity and trajectory instability by implementing a 12-degree-of-freedom model in PyBullet simulation. Our approach features three key innovations: (1) a hybrid reward function ($R_t=0.72 · e^{-\Delta {\mathrm{CoG}}_t} + 0.25 · v_t - 0.11 · ‖{\tau }_t‖$) explicitly prioritizing center-of-gravity (CoG) stabilization; (2) rigorous benchmarking demonstrating Adam’s superiority over SGD for policy convergence (68% lower reward variance); and (3) a four-metric evaluation protocol quantifying locomotion quality through reward progression, CoG deviation, policy loss, and KL-divergence penalties. Experimental results confirm an 87.5% reduction in vertical CoG oscillation (from 2.0″ to 0.25″) across 1 million training steps. Policy optimization achieved −6.2 × 10⁻⁴ loss with KL penalties converging to 0.13, indicating stable gait generation. The framework’s efficacy is further validated by consistent CoG stabilization during deployment, demonstrating potential for real-world applications requiring robust terrain adaptation. Full article

Hexavalent chromium (Cr(VI)), a common contaminant in industrial wastewater, poses severe health risks due to its carcinogenic and mutagenic properties. Consequently, the development of efficient and environmentally friendly methods to reduce Cr(VI) to the less toxic trivalent chromium (Cr(III)) is of great importance. In this study, we present a cost-effective photocatalytic approach using graphitic carbon nitride (g-C₃N₄) modified with 1,3,5-trihydroxybenzene via one-step thermal condensation. The modified photo-catalyst exhibited improved surface area, porosity, visible-light absorption, and a narrowed band gap, all of which contributed to enhanced charge separation. As a result, nearly complete reduction in Cr(VI) was achieved within 90 min under visible-light irradiation. Further optimization of catalyst dosage and EDTA concentration gave even higher reduction efficiency. This work offers a promising strategy for the design of high-performance photocatalysts for environmental remediation. Full article

The transition to sustainable energy sources has intensified interest in lignocellulosic biomass (LCB) as a feedstock for second-generation biofuels. However, the inherent structural recalcitrance of LCB requires the utilization of an effective pretreatment to enhance enzymatic hydrolysis and subsequent fermentation yields. This manuscript presents a novel, single-step, and optimized nitrogen explosive decompression system (NED 3.0) designed to address the critical limitations of earlier NED versions by enabling the in situ removal of inhibitory compounds from biomass slurry and fermentation inefficiency at elevated temperatures, thereby reducing or eliminating the need for post-treatment detoxification. Aspen wood (Populus tremula) was pretreated by NED 3.0 at 200 °C, followed by enzymatic hydrolysis and fermentation. The analytical results confirmed substantial reductions in common fermentation inhibitors, such as acetic acid (up to 2.18 g/100 g dry biomass) and furfural (0.18 g/100 g dry biomass), during early filtrate recovery. Hydrolysate analysis revealed a glucose yield of 26.41 g/100 g dry biomass, corresponding to a hydrolysis efficiency of 41.3%. Fermentation yielded up to 8.05 g ethanol/100 g dry biomass and achieved a fermentation efficiency of 59.8%. Inhibitor concentrations in both hydrolysate and fermentation broth remained within tolerable limits, allowing for effective glucose release and sustained fermentation performance. Compared with earlier NED configurations, the optimized system improved sugar recovery and ethanol production. These findings confirm the operational advantages of NED 3.0, including reduced inhibitory stress, simplified process integration, and chemical-free operation, underscoring its potential for scalability in line with the EU Green Deal for bioethanol production from woody biomass. Full article

Background/Objectives: This case study presents the first documented use of a low-cost, simulated, patient-specific three-dimensional (3D) printed model to support presurgical planning for an infant with Apert syndrome in a resource-limited setting. The primary objectives are to (1) demonstrate the value of 3D printing as a simulation tool for preoperative planning in low-resource environments and (2) identify opportunities for future AI-enhanced simulation models in craniofacial surgical planning. Methods: High-resolution CT data were segmented using InVesalius 3, with mesh refinement performed in ANSYS SpaceClaim (version 2021). The cranial model was fabricated using fused deposition modeling (FDM) on a Creality Ender-3 printer with Acrylonitrile Butadiene Styrene (ABS) filament. Results: The resulting 3D-printed simulated model enabled the surgical team to assess cranial anatomy, simulate incision placement, and rehearse osteotomies. These steps contributed to a reduction in operative time and fewer complications during surgery. Conclusions: This case demonstrates the value of accessible 3D printing as a simulation tool in surgical planning within low-resource settings. Building on this success, the study highlights potential points for AI integration, such as automated image segmentation and model reconstruction, to increase efficiency and scalability in future 3D-printed simulation models. Full article

Novel networking paradigms such as the Internet of Things (IoT) have expanded their usage and deployment to various application domains. Consequently, unseen critical security vulnerabilities such as zero-day attacks have emerged in such deployments. The design of intrusion detection systems for IoT networks is often challenged by a lack of labeled data, which complicates the development of robust defenses against adversarial attacks. As deep learning-based network intrusion detection systems, network intrusion detection systems (NIDS) have been used to counteract emerging security vulnerabilities. However, the deep learning models used in such NIDS are vulnerable to adversarial examples. Adversarial examples are specifically engineered samples tailored to a specific deep learning model; they are developed by minimal perturbation of network packet features, and are intended to cause misclassification. Such examples can bypass NIDS or enable the rejection of regular network traffic. Research in the adversarial example detection domain has yielded several prominent methods; however, most of those methods involve computationally expensive retraining steps and require access to labeled data, which are often lacking in IoT network deployments. In this paper, we propose an unsupervised method for detecting adversarial examples that performs early detection based on the intrinsic characteristics of the deep learning model. Our proposed method requires neither computationally expensive retraining nor extra hardware overhead for implementation. For the work in this paper, we first perform adversarial example generation on a deep learning model using autoencoders. After successful adversarial example generation, we perform adversarial example detection using the intrinsic characteristics of the layers in the deep learning model. A robustness analysis of our approach reveals that an attacker can easily bypass the detection mechanism by using low-magnitude log-normal Gaussian noise. Furthermore, we also test the robustness of our detection method against further compromise by the attacker. We tested our approach on the Kitsune datasets, which are state-of-the-art datasets obtained from deployed IoT network scenarios. Our experimental results show an average adversarial example generation time of 0.337 s and an average detection rate of almost 100%. The robustness analysis of our detection method reveals a reduction of almost 100% in adversarial example detection after compromise by the attacker. Full article

Background/Objectives: Elevated serum uric acid levels are associated with the occurrence, development, and adverse events of coronary heart disease (CHD) and CHD risk factors. However, the extent of any pathogenic effect of the serum uric acid on CHD and whether CHD risk factors play a confounding or mediating role are still unclear. Methods: The potential causal associations of serum uric acid with CHD were evaluated via cross-trait linkage disequilibrium score regression analysis and Mendelian randomization. The pleiotropy of genetic tools was analyzed via a Bayesian colocalization approach. Moreover, we utilized two-step MR to identify risk factors mediating the relationship between uric acid and CHD. Results: Mendelian randomization results derived from two genetic instrument selection strategies support that serum uric acid levels have a significant causal relationship with coronary artery disease, stable angina pectoris, and myocardial infarction. This causal relationship was partially mediated by diastolic blood pressure, mean arterial pressure, and serum triglycerides. Transcriptomic analysis revealed that serum uric acid may directly contribute to the development of atherosclerosis by inducing transcriptomic changes in macrophages. Conclusions: Our findings highlight that the control of serum urate concentration in the long-term management of CHD patients may be necessary. Well-designed clinical trials and foundational research are presently required to furnish conclusive proof regarding the specific clinical scenarios in which adequate reduction in urate concentrations can confer cardiovascular advantages. Full article

Molasses, a by-product of raw sugar production, is widely used as a cost-effective carbon and nutrient source for industrial fermentations, including the production of baker’s yeast (Saccharomyces cerevisiae). Due to the cost and limited availability of molasses, efforts have been made to replace molasses with cheaper and more readily available substrates such as corn syrup. However, the quality of dry yeast drops following the replacement of molasses with corn syrup, despite the same amount of total sugar being provided. Our understanding of how molasses replacement affects yeast physiology, especially during the dehydration step, is limited. Here, we examined changes in gene expression of a strain of baker’s yeast during fermentation with increasing corn syrup to molasses ratios at the transcriptomic level. Our findings revealed that the limited availability of the key metal ions copper, iron, and zinc, as well as sulfur from corn syrup (i) reduced their intracellular storage, (ii) impaired the synthesis of unsaturated fatty acids and ergosterol, as evidenced by the decreasing proportions of these important membrane components with higher proportions of corn syrup, and (iii) inactivated oxidative stress response enzymes. Taken together, the molecular and metabolic changes observed suggest a potential reduction in nutrient reserves for fermentation and a possible compromise in cell viability during the drying process, which may ultimately impact the quality of the final dry yeast product. These findings emphasize the importance of precise nutrient supplementation when substituting molasses with cheaper substrates. Full article

This research introduces an automated 3D virtual reconstruction system tailored for architectural heritage (AH) applications, contributing to the ongoing paradigm shift from traditional CAD-based workflows to artificial intelligence-driven methodologies. It reviews recent advancements in machine learning and deep learning—particularly neural radiance fields (NeRFs) and its successor, Gaussian splatting (GS)—as state-of-the-art techniques in the domain. The study advocates for replacing point cloud data in heritage building information modeling workflows with image-based inputs, proposing a novel “photo-to-BIM” pipeline. A proof-of-concept system is presented, capable of processing photographs or video footage of ancient ruins—specifically, Romanesque–Mudéjar churches—to automatically generate 3D mesh reconstructions. The system’s performance is assessed using both objective metrics and subjective evaluations of mesh quality. The results confirm the feasibility and promise of image-based reconstruction as a viable alternative to conventional methods. The study successfully developed a system for automated 3D mesh reconstruction of AH from images. It applied GS and Mip-splatting for NeRFs, proving superior in noise reduction for subsequent mesh extraction via surface-aligned Gaussian splatting for efficient 3D mesh reconstruction. This photo-to-mesh pipeline signifies a viable step towards HBIM. Full article

Aluminothermic reduction is an alternative processing route for the circular economy because Al is produced electrochemically in the Hall–Héroult process with minimal CO₂ emissions if the electricity input is sourced from non-fossil fuel energy sources. This circular processing option attracts increased research attention in the aluminothermic production of manganese and silicon alloys. The Al₂O₃ product must be recycled through hydrometallurgical processing, with leaching as the first step. Recent work has shown that the NaAlO₂ compound is easily leached in water. In this work, a suitable slag formulation is applied in the aluminothermic reduction of manganese ore to form a Na₂O-based slag of high Al₂O₃ solubility to effect good alloy–slag separation. The synergistic effect of carbon and silicon reductants with aluminium is illustrated and compared to the test result with only carbon reductant. The addition of small amounts of carbon reductant to MnO₂-containing ore ensures rapid pre-reduction to MnO, facilitating aluminothermic reduction. At 1350 °C, a loosely sintered mass formed when carbon was added alone. The alloy and slag chemical analyses are compared to the thermochemistry predicted phase chemistry. The alloy consists of 66% Mn, 22–28% Fe, 2–9% Si, 0.4–1.4% Al, and 2.2–3.5% C. The higher %Si alloy is formed by adding Si metal. Although the product slag has a higher Al₂O₃ content (52–55% Al₂O₃) compared to the target slag (39% Al₂O₃), the fluidity of the slags appears sufficient for good alloy separation. Full article

A novel macrodesign for a dental implant characterized by a non-monotonic variation in core diameter and thread shape has been described to produce lower stress levels during insertion as compared to conventional tapered implants. Two finite element models resembling the lower left molar region with preformed osteotomies were created based on a cone beam computed tomography (CBCT) scan. Insertion of both the novel and the conventional, tapered implant type were simulated using Standard for the Exchange of Product model data (STEP) files of both implant types. Von Mises equivalent stress, strain development, and amount of redistributed bone were recorded. The conventional implant demonstrated a continuous increase in strain values and reaction moment throughout the insertion process, with a brief decrease observed during the final stages. Stress levels in the cortical bone gradually increased, followed by a reduction when the implant was finally positioned subcrestally. The novel implant achieved the maximum magnitude of reaction moment and cortical bone strain values when the implant’s maximum core diameter passed the cortical bone layer at around 60% of the insertion process. Following a notable decrease, both the reaction moment and stress started to rise again as the implant penetrated further. The novel implant removed more bones in the trabecular region while the conventional implant predominantly interacted with cortical bone. Overall, the novel design seems to be less traumatic to alveolar bone during the insertion process and hence may lead to reduced levels of initial peri-implant bone loss. Full article