Search Results (4,006)

Microplastic (MP) pollution in agricultural ecosystems is an emerging environmental concern, with limited knowledge of its transport and accumulation in rural waterbodies. This study investigates the distribution and sources of MP in drainage ditches influenced by pond connectivity, land use, and soil properties within a small catchment in Nanjing, East China. Sediment was collected from ditches in 18 sites across forest, agricultural, horticultural, and urban areas. Using laser-directed infrared spectroscopy (LDIR), 922 MP particles were identified. Six materials were dominant: fluororubber (FR), polyethylene terephthalate (PET), polyurethane (PU), acrylonitrile (ACR), chlorinated polyethylene (CPE), and polyethylene (PE). MP concentrations varied by land use and pond connectivity, with ditches above ponds exhibiting higher counts (1700 particles/kg) than those below (1050 particles/kg), indicating that ponds act as MP sinks. The analysis revealed site-specific MP sources, with FR linked to road runoff and PET associated with agricultural practices. Correlations between MP shape and soil properties showed that more compact and filled shapes were more commonly associated with coarser soils. PE particle size was negatively correlated with organic matter. This study highlights the need for targeted strategies to reduce MP pollution in rural landscapes, such as reducing plastic use, ditch maintenance, and improved road runoff management. Full article

Due to the excellent quantum conversion and spectral response characteristics of the AlGaN photocathode, it has become the most promising III-V group semiconductor photocathode in solar-blind signal photoconversion devices in the ultraviolet band. Herein, the influence factors of the time-resolved characteristics of the AlGaN photocathode are researched by solving the photoelectron continuity equation and photoelectron flow density equation, such as the AlN/AlGaN interface recombination rate, AlGaN electron diffusion coefficient, and AlGaN activation layer thickness. The results show that the response time of the AlGaN photocathode decreases gradually with the increase in AlGaN photoelectron diffusion coefficient and AlN/AlGaN interface recombination rate, but the response time of the AlGaN photocathode gradually becomes saturated with the further increase in AlN/AlGaN interface recombination rate. When the thickness of the AlGaN photocathode is reduced from 250 nm to 50 nm, the response time of the AlGaN photocathode decreases from 63.28 ps to 9.91 ps, and the response time of AlGaN photocathode greatly improves. This study provides theoretical guidance for the development of a fast response UV detector. Full article

In the quest for computational efficiency, binary neural networks (BNNs) have emerged as a promising paradigm, offering significant reductions in memory footprint and computational latency. In traditional BNN implementation, the first and last layers are typically full-precision, which causes higher logic usage in field-programmable gate array (FPGA) implementation. To solve these issues, we introduce a novel approach named Ponte (Represent Totally Binary Neural Network Toward Efficiency) that extends the binarization process to the first and last layers of BNNs. We challenge the convention by proposing a fully binary layer replacement that mitigates the computational overhead without compromising accuracy. Our method leverages a unique encoding technique, Ponte::encoding, and a channel duplication strategy, Ponte::dispatch, and Ponte::sharing, to address the non-linearity and capacity constraints posed by binary layers. Surprisingly, all of them are back-propagation-supported, which allows our work to be implemented in the last layer through extensive experimentation on benchmark datasets, including CIFAR-10 and ImageNet. We demonstrate that Ponte not only preserves the integrity of input data but also enhances the representational capacity of BNNs. The proposed architecture achieves comparable, if not superior, performance metrics while significantly reducing the computational demands, thereby marking a step forward in the practical deployment of BNNs in resource-constrained environments. Full article

Background: Sweet potato (Ipomoea batatas (L.) Lam.), a key global root crop, faces challenges due to its narrow genetic background. This issue can be addressed by utilizing the diverse genetic resources of sweet potato’s wild relatives, which are invaluable for its genetic improvement. Methods: The morphological differences in leaves, stems, and roots among 13 Ipomoea species were observed and compared. Chromosome numbers were determined by examining metaphase cells from root tips. Fluorescence in situ hybridization (FISH) was used to identify the number of 5S and 18S rDNA sites in these species. PCR amplification was performed for both 5S and 18S rDNA, and phylogenetic relationships among the species were analyzed based on the sequences of 18S rDNA. Results: Three species were found to have enlarged roots among the 13 Ipomoea species. Chromosome analysis revealed that I. batatas had 90 chromosomes, Ipomoea pes-tigridis had 28 chromosomes, while the remaining species possessed 30 chromosomes. Detection of rDNA sites in the 13 species showed two distinct 5S rDNA site patterns and six 18S rDNA site patterns in the 12 diploid species. These rDNA sites occurred in pairs, except for the seven 18S rDNA sites observed in Ipomoea digitata. PCR amplification of 5S rDNA identified four distinct patterns, while 18S rDNA showed only a single pattern across the species. Phylogenetic analysis divided the 13 species into two primary clades, with the closest relationships found between I. batatas and Ipomoea trifida, as well as between Ipomoea platensis and I. digitata. Conclusions: These results enhance our understanding of the diversity among Ipomoea species and provide valuable insights for breeders using these species to generate improved varieties. Full article

The use of foam materials in environments where they come into contact with individuals often results in deterioration, necessitating periodic replacements to maintain safety and hygiene standards. Foam, a lightweight, porous plastic formed by aggregated bubbles, possesses excellent impact-absorbing properties; however, its inherent porosity and susceptibility to wear present challenges. This project aims to develop a technological application for repurposing spent polyurethane (PU) foam from leisure facilities into effective impact absorption systems. By focusing on the reuse of deteriorated foam materials, this initiative seeks to minimize environmental impact while leveraging their beneficial technical characteristics. Addressing issues related to foam degradation, this project endeavors to create sustainable solutions by reintegrating spent foam into new systems. This innovative approach promotes sustainability while enhancing safety through the provision of high-quality, impact-resistant elements. Ultimately, this work aims to contribute to environmental conservation and the advancement of effective impact protection measures in leisure facilities. Full article

The current knowledge of Staphylococcus warneri phages is limited, with few genomes sequenced and characterized. In this study, a prophage, vB_G30_01, isolated from Staphylococcus warneri G30 was characterized and evaluated for its lysogenic host range. The phage was studied using transmission electron microscopy and a host range. The phage genome was sequenced and characterized in depth, including phylogenetic and taxonomic analyses. The linear dsDNA genome of vB_G30_01 contains 67 predicted open reading frames (ORFs), classifying it within Bronfenbrennervirinae. With a total of 10 ORFs involved in DNA replication-related and transcriptional regulator functions, vB_G30_01 may play a role in the genetics and transcription of a host. Additionally, vB_G30_01 possesses a complete set of genes related to host lysogeny and lysis, implying that vB_G30_01 may influence the survival and adaptation of its host. Furthermore, a comparative genomic analysis reveals that vB_G30_01 shares high genomic similarity with other Staphylococcus phages and is relatively closely related to those of Exiguobacterium and Bacillus, which, in combination with the cross-infection assay, suggests possible cross-species infection capabilities. This study enhances the understanding of Staphylococcus warneri prophages, providing insights into phage–host interactions and potential horizontal gene transfer. Full article

Within the area of environmental perception, automatic navigation, object detection, and computer vision are crucial and demanding fields with many applications in modern industries, such as multi-target long-term visual tracking in automated production, defect detection, and driverless robotic vehicles. The performance of computer vision has greatly improved recently thanks to developments in deep learning algorithms and hardware computing capabilities, which have spawned the creation of a large number of related applications. At the same time, with the rapid increase in autonomous systems in the market, energy consumption has become an increasingly critical issue in computer vision and SLAM (Simultaneous Localization and Mapping) algorithms. This paper presents the results of a detailed review of over 100 papers published over the course of two decades (1999–2024), with a primary focus on the technical advancement in computer vision. To elucidate the foundational principles, an examination of typical visual algorithms based on traditional correlation filtering was initially conducted. Subsequently, a comprehensive overview of the state-of-the-art advancements in deep learning-based computer vision techniques was compiled. Furthermore, a comparative analysis of conventional and novel algorithms was undertaken to discuss the future trends and directions of computer vision. Lastly, the feasibility of employing visual SLAM algorithms in the context of autonomous vehicles was explored. Additionally, in the context of intelligent robots for low-carbon, unmanned factories, we discussed model optimization techniques such as pruning and quantization, highlighting their importance in enhancing energy efficiency. We conducted a comprehensive comparison of the performance and energy consumption of various computer vision algorithms, with a detailed exploration of how to balance these factors and a discussion of potential future development trends. Full article

As social media has risen, virtual social media influencers have become a significant tool in modern marketing, utilizing computer-generated images (CGI), machine learning algorithms, and artificial intelligence technologies to connect with consumers via virtual online personas. In this study, the Uses and Gratifications Theory (UGT) is employed as a theoretical framework to explore the effects of educational narrative strategies and evaluative narrative strategies on consumer product preferences, with an analysis of the mediating role of word-of-mouth effectiveness and the moderating role of perceived product usability. It was demonstrated in Experiment 1 that virtual influencers employing educational narrative strategies are more effective than those using evaluative narrative strategies in enhancing consumer product preferences. The boundaries of the study were clarified in Experiment 2, which found that the main effect of educational narrative strategies utilized by social media influencers to increase consumer product preferences is present only in the context of virtual influencers. In Experiment 3, the mediating role of word-of-mouth recommendation effectiveness in the relationship between narrative strategies and consumer product preferences was further verified. The moderating role of perceived product usability was examined in Experiment 4, and it was found that the main effect is more pronounced in contexts where perceived product usability is low. The results of this study provide theoretical and practical guidance on how companies can effectively leverage virtual influencers to promote their products. Full article

Spinal medical image segmentation is critical for diagnosing and treating spinal disorders. However, ambiguity in anatomical boundaries and interfering factors in medical images often cause segmentation errors. Current deep learning models cannot fully capture the intrinsic data properties, leading to unstable feature spaces. To tackle the above problems, we propose Verdiff-Net, a novel diffusion-based segmentation framework designed to improve segmentation accuracy and stability by learning the underlying data distribution. Verdiff-Net integrates a multi-scale fusion module (MSFM) for fine feature extraction and a noise semantic adapter (NSA) to refine segmentation masks. Validated across four multi-modality spinal datasets, Verdiff-Net achieves a high Dice coefficient of 93%, demonstrating its potential for clinical applications in precision spinal surgery. Full article

The cost of hyperspectral image (HSI) classification primarily stems from the annotation of image pixels. In real-world classification scenarios, the measurement and annotation process is both time-consuming and labor-intensive. Therefore, reducing the number of labeled pixels while maintaining classification accuracy is a key research focus in HSI classification. This paper introduces a multi-strategy triple network classifier (MSTNC) to address the issue of limited labeled data in HSI classification by improving learning strategies. First, we use the contrast learning strategy to design a lightweight triple network classifier (TNC) with low sample dependence. Due to the construction of triple sample pairs, the number of labeled samples can be increased, which is beneficial for extracting intra-class and inter-class features of pixels. Second, an active learning strategy is used to label the most valuable pixels, improving the quality of the labeled data. To address the difficulty of sampling effectively under extremely limited labeling budgets, we propose a new feature-mixed active learning (FMAL) method to query valuable samples. Fine-tuning is then used to help the MSTNC learn a more comprehensive feature distribution, reducing the model’s dependence on accuracy when querying samples. Therefore, the sample quality is improved. Finally, we propose an innovative dual-threshold pseudo-active learning (DSPAL) strategy, filtering out pseudo-label samples with both high confidence and uncertainty. Extending the training set without increasing the labeling cost further improves the classification accuracy of the model. Extensive experiments are conducted on three benchmark HSI datasets. Across various labeling ratios, the MSTNC outperforms several state-of-the-art methods. In particular, under extreme small-sample conditions (five samples per class), the overall accuracy reaches 82.97% (IP), 87.94% (PU), and 86.57% (WHU). Full article

The preparation and application of non-isocyanate polyurethane (NIPU) from biomass raw materials as a substitute for traditional polyurethane (PU) has recently become a research hot topic as it avoids the toxicity and moisture sensitivity of isocyanate-based PU. In the work presented here, self-blowing GNIPU non-isocyanate polyurethane (NIPU) rigid foams were prepared at room temperature, based on glucose, with acids as catalysts and glutaraldehyde as a cross-linker. The effects of different acids and glutaraldehyde addition on foam morphology and properties were investigated. The water absorption, compressive resistance, fire resistance, and limiting oxygen index (LOI) were tested to evaluate the relevant properties of the foams, and scanning electron microscopy (SEM) was used to observe the foams’ cell structure. The results show that all these foams have a similar apparent density, while their 24 h water absorption is different. The foam prepared with phosphoric acid as a catalyst presented a better compressive strength compared to the other types prepared with different catalysts when above 65% compression. It also presents the best fire resistance with an LOI value of 24.3% (great than 22%), indicating that it possesses a good level of flame retardancy. Thermogravimetric analysis also showed that phosphoric acid catalysis slightly improved the GNIPU foams’ thermal stability. This is mainly due to the flame-retardant effect of the phosphate ion. In addition, scanning electron microscopy (SEM) results showed that all the GNIPU foams exhibited similar open-cell morphologies with the cell pore sizes mainly distributed in the 200–250 μm range. Full article

In this study, we propose a model called CNN-BiLSTM-AM that utilizes deep learning techniques to forecast severe convective weather based on ERA5 hourly data and observations. The model integrates a CNN with a Bidirectional Long Short-Term Memory (BiLSTM) system and an Attention Mechanism (AM). The CNN is tasked with extracting features from the input data, while the BiLSTM effectively captures temporal dependencies. The AM enhances the results by considering the impact of past feature states on severe weather phenomena. Additionally, we assess the performance of our model in comparison to traditional network architectures, including ConvLSTM, Predrnn++, CNN, FC-LSTM, and LSTM. Our results indicate that the CNN-BiLSTM-AM model exhibits superior accuracy in precipitation forecasting. Especially with the extension of the forecast time, the model performs well across multiple evaluation metrics. Furthermore, an interpretability analysis of the convective weather mechanisms utilizing machine learning highlights the critical role of total precipitable water (PWAT) in short-term heavy precipitation forecasts. It also emphasizes the significant impact of regional variables on convective weather patterns and the role of convective available potential energy (CAPE) in fostering conditions conducive to convection. These findings not only confirm the effectiveness of deep learning in the automatic identification of severe weather features but also validate the suitability of the sample dataset employed. Given its remarkable performance and robustness, we advocate for the adoption of this model to enhance the forecast of severe convective weather across various business applications. Full article

This study investigates the financial behavior of Thai investors in equity crowdfunding (ECF), focusing on the factors that influence their investment intentions. Drawing upon the Information System Success Model (ISSM), the Theory of Diffusion of Innovations, and the Technology Acceptance Model 3 (TAM3), the research examines the platform quality (PQ), platform characteristics (PC), and social influence (SI) as independent variables, with the perceived usefulness (PU) and perceived ease of use (PEOU) acting as mediators. Data were gathered from 275 Thai investors and analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM). The findings reveal that the PU significantly influences investment decisions both directly and indirectly through the PEOU, which also directly affects investment intention. Furthermore, SI, PC, and PQ have indirect effects on investment decisions via the PU and PEOU, with SI being the most influential factor. This study provides valuable insights into optimizing ECF platform design, fostering investor trust, and enhancing regulatory frameworks to facilitate financial inclusion and innovation in the Thai crowdfunding landscape. Full article

The ability to classify and recognize maritime targets based on visual images plays an important role in advancing ship intelligence and digitalization. The current target recognition algorithms for common maritime targets, such as buoys, reefs, other ships, and bridges of different colors, face challenges such as incomplete classification, low recognition accuracy, and a large number of model parameters. To address these issues, this paper proposes a novel maritime target recognition method called DTI-YOL (DualConv Triple Attention InnerEIOU-You Only Look Once). This method is based on a triple attention mechanism designed to enhance the model’s ability to classify and recognize buoys of different colors in the channel while also making the feature extraction network more lightweight. First, the lightweight double convolution kernel feature extraction layer is constructed using group convolution technology to replace the Conv structure of YOLOv9 (You Only Look Once Version 9), effectively reducing the number of parameters in the original model. Second, an improved three-branch structure is designed to capture cross-dimensional interactions of input image features. This structure forms a triple attention mechanism that accounts for the mutual dependencies between input channels and spatial positions, allowing for the calculation of attention weights for targets such as bridges, buoys, and other ships. Finally, InnerEIoU is used to replace CIoU to improve the loss function, thereby optimizing loss regression for targets with large scale differences. To verify the effectiveness of these algorithmic improvements, the DTI-YOLO algorithm was tested on a self-made dataset of 2300 ship navigation images. The experimental results show that the average accuracy of this method in identifying seven types of targets—including buoys, bridges, islands and reefs, container ships, bulk carriers, passenger ships, and other ships—reached 92.1%, with a 12% reduction in the number of parameters. This enhancement improves the model’s ability to recognize and distinguish different targets and buoy colors. Full article