Multimodal hateful social media meme detection is an important and challenging problem in the vision-language domain. Recent studies show high accuracy for such multimodal tasks due to datasets that provide better joint multimodal embedding to narrow the semantic gap. Religiously hateful meme detection is not extensively explored among published datasets. While there is a need for higher accuracy on religiously hateful memes, deep learning–based models often suffer from inductive bias. This issue is addressed in this work with the following contributions. First, a religiously hateful memes dataset is created and published publicly to advance hateful religious memes detection research. Over 2000 meme images are collected with their corresponding text. The proposed approach compares and fine-tunes VisualBERT pre-trained on the Conceptual Caption (CC) dataset for the downstream classification task. We also extend the dataset with the Facebook hateful memes dataset. We extract visual features using ResNeXT-152 Aggregated Residual Transformations–based Masked Regions with Convolutional Neural Networks (R-CNN) and Bidirectional Encoder Representations from Transformers (BERT) uncased for textual encoding for the early fusion model. We use the primary evaluation metric of an Area Under the Operator Characters Curve (AUROC) to measure model separability. Results show that the proposed approach has a higher AUROC score of 78%, proving the model’s higher separability performance and an accuracy of 70%. It shows comparatively superior performance considering dataset size and against ensemble-based machine learning approaches.

The healthcare industry is one of the most vulnerable to cybercrime and privacy violations because health data is very sensitive and spread out in many places. Recent confidentiality trends and a rising number of infringements in different sectors make it crucial to implement new methods that protect data privacy while maintaining accuracy and sustainability. Moreover, the intermittent nature of remote clients with imbalanced datasets poses a significant obstacle for decentralized healthcare systems. Federated learning (FL) is a decentralized and privacy-protecting approach to deep learning and machine learning models. In this article, we implement a scalable FL framework for interactive smart healthcare systems with intermittent clients using chest X-ray images. Remote hospitals may have imbalanced datasets with intermittent clients communicating with the FL global server. The data augmentation method is used to balance datasets for local model training. In practice, some clients may leave the training process while others join due to technical or connectivity issues. The proposed method is tested with five to eighteen clients and different testing data sizes to evaluate performance in various situations. The experiments show that the proposed FL approach produces competitive results when dealing with two distinct problems, such as intermittent clients and imbalanced data. These findings would encourage medical institutions to collaborate and use rich private data to quickly develop a powerful patient diagnostic model.

AbstractInfectious diseases caused by bacteria, viruses, fungi, and parasites pose a significant societal challenge. In response, scientists are developing advanced technology to enhance the prevention, diagnosis, and treatment of such diseases. One such promising technology is microfluidic systems, which are utilized in organ‐on‐a‐chip systems to replicate the microenvironments of organs. These systems have potential applications in drug screening, disease modeling, and personalized medicine. This review provides an overview of recent advances in organ‐on‐a‐chip platforms and their potential for preventing and diagnosing various infections. After discussing traditional techniques employed in studying infectious diseases, the role of microfluidic platforms in detecting infections is delved in. It is expound on advanced microfluidic‐based strategies for infection diagnosis, such as the polymerase chain reaction‐based microfluidic devices, enzyme linked immunosorbent assay‐based microfluidic devices, hierarchical nanofluidic molecular enrichment systemand µWestern blotting‐based microfluidic devices, and smartphone‐based microfluidic devices. Additionally, future research challenges and perspectives are discussed on microfluidic systems in biomedical and regenerative medicine applications. Consequently, microfluidic platforms have the potential to serve as fundamental frameworks for understanding infectious diseases, thereby leading to personalized regenerative medicine. hierarchical nanofluidic molecular enrichment system

One of the biggest challenges to the sustainable manufacture of liquid ammonia and the prevention of worldwide water contamination is the development of effective electrocatalysts for the electrochemical reduction of nitrate (NO3−) to NH3 with high stability. Herein, a highly active and serviceable electrocatalyst is synthesized by pyrolysis, composed of nanostructure nickel phosphide (Ni2P) embedded in nitrogen phosphide doped carbon (NPC) followed by deposition on carbon paper (CP) to improve the electrocatalytic nitrate reduction. Various characterization techniques investigate the crystallinity, morphology, and chemical components of the Ni2P-NPC/CP nanoparticles. The results support the formation of nanostructure Ni2P and strong synergistic interactions between Ni2P and NPC, which resulted in substantial active sites and high electrical conductivity. Excellent performance of Ni2P-NPC/CP nanoparticles is achieved for electrocatalytic NO3− reduction with an NH4+ yield rate of 2.468 mg h−1 mgcat.−1 and Faradaic efficiency (FE) of 84.6% at −1.2 V vs. RHE. Additionally, Ni2P-NPC/CP nanoparticles exhibit exceptional robustness and endurance. Studies using isotope labeling have been carried out, and the results show that nitrate reduction produces ammonia. Ni2P-based electrocatalysts can effectively treat nitrate wastewater to recover ammonia and facilitate its use in diverse industrial applications.

This paper presents a practical approach to the production of certified reference materials (CRMs) for ferrosilicon magnesium alloys. These new CRMs are predicted to be used in fast X-ray fluorescence spectrometry (XRF) analysis, which does not require sample digestion and does not result in the production of acidic sewage and emissions, contrary to the classical and instrumental techniques currently used in laboratories. Four new certified reference materials (CRMs) were developed to fill the gap in the reference materials market and ensure fast and traceable analyses. The materials were produced with an industrial process and then homogenized and mixed to achieve the required compositions and level of homogeneity. The homogeneity was determined using specially developed analytical methods and confirmed statistically by ANOVA. Additionally, the results of the tests show the short- and long-term stabilities of the new materials. The certified values for specific element contents were determined in interlaboratory tests. All results were assessed statistically for outliers. The results from three or more independent and different analytical methods were used for the calculations. In parallel homogeneity, the stability, and characterization standard uncertainties were calculated and used in the estimation of the final expanded uncertainties of the certified values. Finally, four new CRMs were assisted with the proper certificates according to ISO standards.

This paper presents a thorough examination of methane capture from Polish coal mines, contextualized within the framework of the European Union’s (EU) climate policy objectives. Through a strategic analysis encompassing the interior of coal mines, the surrounding environment, and the macro environment, this study elucidates the complex dynamics involved in methane emissions and capture initiatives. The key findings include a declining trend in absolute methane emissions since 2008, despite fluctuations in coal extraction volumes, and a relatively stable level of methane capture exceeding 300 million m3/year since 2014. The analysis underscores the critical role of government support, both in terms of financial incentives and streamlined regulatory processes, to facilitate the integration of methane capture technologies into coal mining operations. Collaboration through partnerships and stakeholder engagement emerges as essential for overcoming resource competition and ensuring the long-term success of methane capture projects. This paper also highlights the economic and environmental opportunities presented by methane reserves, emphasizing the importance of investment in efficient extraction technologies. Despite these advancements, challenges persist, particularly regarding the low efficiency of current de-methanation technologies. Recommendations for modernization and technological innovation are proposed to enhance methane capture efficiency and utilization.

The paper presents a proposed methodology for designing and planning research on driver behavior at pedestrian crossings using whole-vehicle simulators. It was assumed that dedicated research should be conducted in specific problem contexts. The problems identified were the identification of hazards and the risk of accidents involving vulnerable road users. The purpose of this identification is to determine the determinants of safety at pedestrian crossings, which should constitute guidance when designing new solutions for safety support systems at pedestrian crossings. A number of hazard factors were identified; divided into categories, including types of crossings, location, and surroundings; behavior of vulnerable road users; and attention (focus) distractors, both inside and outside the vehicle. A method for defining and selecting research scenarios and selecting a group of research participants was proposed. Additionally, it was proposed to conduct repeatable test scenarios for different driving speeds and different weather conditions. With respect to the publications on this topic, this work focuses on the process of designing and planning dedicated simulation studies, which may provide a source of guidance and good practices for other researchers. This is an example of how interdisciplinary research involving human factors, traffic organization, and ITS systems can be planned and implemented.

Transforming amines with low boiling points and high volatilities into protic salts is a versatile strategy to utilize low molecular weight compounds as precursors for N-doped carbon structures in a straightforward carbonization procedure. Herein, conventional mineral acids commonly used for the synthesis of protic salts were replaced by bio-derived phytic acid, which, combined with various amines and amino acids, yielded partially or fully bio-derived protic salts. The biomass-based salts showed higher char-forming ability than their mineral acid-based analogs (up to 55.9% at 800°), simultaneously providing carbon materials with significant porosity (up to 1177 m2g−1) and a considerable level of N,P,O-doping. Here, we present the first comprehensive study on the correlation between the structure of the bio-derived protic precursors and the properties of derived carbon materials to guide future designs of biomass-derived precursors for the one-step synthesis of sustainable carbon materials. Additionally, we demonstrate how to improve the textural properties of the protic-salt-derived carbons (which suffer from high brittleness) by simply upgrading them into highly flexible nanocomposites using high-quality single-walled carbon nanotubes. Consequently, self-standing electrodes for the oxygen reduction reaction were created.