Search Results (447)

The residential separate collection of waste is the first stage in waste recyclability for sustainable development. The paper focuses on designing and implementing a low-cost residential automatic waste sorting bin (RBin) for recycling, alleviating the user’s classification burden. Next, an analysis of two object identification and classification models was conducted to sort materials into the categories of cardboard, glass, plastic, and metal. A major challenge in sorting classification is distinguishing between glass and plastic due to their similar visual characteristics. The research assesses the performance of the Azure Custom Vision Service (ACVS) model, which achieves high accuracy on training data but underperforms in real-time applications, with an accuracy of 95.13%. In contrast, the second model, the Custom Waste Sorting Model (CWSM), demonstrates high accuracy (96.25%) during training and proves to be effective in real-time applications. The CWSM uses a two-tier approach, first identifying the object descriptively using the Google Vision API Service (GVAS) model, followed by classification through the CWSM, a predicate-based custom model. The CWSM employs the LbfgsMaximumEntropyMulti algorithm and a dataset of 1000 records for training, divided equally across the categories. This study proposes an innovative evaluation metric, the Weighted Classification Confidence Score (WCCS). The results show that the CWSM outperforms ACVS in real-world testing, achieving a real accuracy of 99.75% after applying the WCCS. The paper explores the importance of customized models over pre-implemented services when the model uses characteristics and not pixel-by-pixel examination. Full article

Automated head chamber systems (AHCS) are increasingly deployed to measure enteric emissions in vivo. However, guidance for AHCS-derived emissions data analyses pertains to confined settings, such as feedlots, with less instruction for grazing systems. Accordingly, our first objective in this experiment was to determine the utility of two data preprocessing approaches for grazing-based analyses. Using Pearson’s correlation, we compared “simple arithmetic” and “time-bin” averaging to arrive at a single estimate of daily gas flux. For our second objective, we evaluated test period length averaging at 1, 3, 7, and 14 d intervals to determine daily pasture-based emissions estimates under two experimental conditions: herd access to a single AHCS unit vs. two AHCS units. Unlike findings from the confinement-based literature, where slight improvements have been observed, time-bin averaging, compared to simple arithmetic averaging, did not improve gas flux estimation from grazing for CH₄ (p ≥ 0.46) or CO₂ (p ≥ 0.60). Irrespective of experimental condition, i.e., herd access to a single AHCS unit vs. two AHCS units, assessment of variability of diurnal emissions patterns revealed CH₄ flux on pasture had at least half as much variability for the same individuals acclimated in confinement. Using a 7-day test period length interval, aggregating gas flux data at 7 d at a time was adequate for capturing diurnal emissions variation in grazing steers, as no improvement was observed in the percentage of individuals with five of six time bins measured for a 14-day test period length interval. This analysis should provide insights into future research to standardize AHCS data preprocessing across experiments and research groups. Full article

Tetrigidae is one of the largest orthopteran families, but very few studies so far have integrated molecular and morphological data. Unsurprisingly, few species have been DNA barcoded, and the unresolved taxonomy makes Tetrigidae a difficult group to work with. Here, we examined a sample of 90 specimens collected as a part of the Costa Rican DNA barcoding project and identified 20 species assigned to 24 BINs, among which are 8 newly described species: Scaria bimaculata sp. nov., Lophotettix semicristatus sp. nov., Otumba auricarinata sp. nov., Otumba tenuis sp. nov., Otumba ignicula sp. nov., Metrodora mollilobata sp. nov., Metrodora ala sp. nov., and Platythorus inabsolutus sp. nov. We found that coloration and lateral lobe shape are species-specific among the examined species of Batrachideinae and Metrodorinae and that Lophotettiginae and Metrodora might be more closely related than previously assumed. Full article

The service-learning (SL) approach has shown effectiveness in fulfilling both academic and community-oriented objectives. This paper focuses on a specific case study for a Cartography, Remote Sensing, and Geographical Information Systems (GIS) course for Environmental Sciences undergraduate students. The main goals for implementing SL practice were (1) to enhance students’ GIS knowledge and to develop cross-cutting skills by working with real-world problems; (2) to share with society the knowledge acquired by students and ensure that it is valued; and (3) to prompt reflection on urban waste issues among students. The activity consisted of analyzing the waste containers along the 1 km riverbanks in León (Spain) and elaborating a proposal for the location of new rubbish bins to deliver to a City Council’s environmental technician. The results showed an improvement in students’ GIS management skills to solve environmental problems compared to those from the previous 3 years and a satisfactory response from environmental professionals with delivering the results. Together, an increase in students discussing urban waste was observed during the sessions. Projects like this not only teach technical skills but also provide a deeper understanding of data collection and implementation processes in environmental issues, which are closely aligned with professional experiences, and awareness of the practical application of the knowledge acquired. Full article

The phenomenal growth of digital learning platforms has brought new learner engagement and retention challenges to higher education. This study proposes a framework that integrates game mechanics—leveling systems, badges, and timely feedback—with artificial intelligence (AI)-driven personalization to meet the challenges of enhanced adaptability, motivation, and learning outcomes in online environments. Key design elements were identified through literature reviews and consultations with instructional design experts, leading to the development an adaptive learning platform prototype. The prototype underwent an eight-week pilot study with 250 Prince Sattam Bin Abdulaziz University (PSAU) students randomly assigned to a control group (non-adaptive system) or an experimental group (adaptive system). Data sources included pre- and post-tests, platform engagement analytics, and learner perception surveys. The results showed that the adaptive group outperformed the control group in the post-test scores (M = 85.2, SD = 6.4 vs. M = 78.5, SD = 7.2) and motivation levels (M = 4.2, SD = 0.7 vs. M = 3.6, SD = 0.8). Additionally, 82% of the adaptive group achieved mastery-level performance compared to 64% in the control group. These findings demonstrate the potential of integrating game mechanics and AI-driven personalization to transform digital learning, offering a roadmap for scalable, data-driven adaptive platforms. Future research will address long-term retention and diverse subject applications. Full article

Recent advances in autonomous driving have led to an increased use of LiDAR (Light Detection and Ranging) sensors for high-frequency 3D perceptions, resulting in massive data volumes that challenge in-vehicle networks, storage systems, and cloud-edge communications. To address this issue, we propose a bounded-error LiDAR compression framework that enforces a user-defined maximum coordinate deviation (e.g., 2 cm) in the real-world space. Our method combines multiple compression strategies in both axis-wise metric Axis or Euclidean metric L2 (namely, Error-Bounded Huffman Coding (EB-HC), Error-Bounded 3D Compression (EB-3D), and the extended Error-Bounded Huffman Coding with 3D Integration (EB-HC-3D)) with a lossless Huffman coding baseline. By quantizing and grouping point coordinates based on a strict threshold (either axis-wise or Euclidean), our method significantly reduces data size while preserving the geometric fidelity. Experiments on the KITTI dataset demonstrate that, under a 2 cm bounded-error, our single-bin compression reduces the data to 25–35% of their original size, while multi-bin processing can further compress the data to 15–25% of their original volume. An analysis of compression ratios, error metrics, and encoding/decoding speeds shows that our method achieves a substantial data reduction while keeping reconstruction errors within the specified limit. Moreover, runtime profiling indicates that our method is well-suited for deployment on in-vehicle edge devices, thereby enabling scalable cloud-edge cooperation. Full article

As global waste production continues to rise, improper handling of household waste significantly contributes to environmental pollution and resource depletion. Inefficient sorting at the household level leads to the contamination of recyclables, reducing recycling efficiency and increasing landfill waste. Effective waste sorting is essential for conserving manual labor, protecting the environment, and ensuring sustainable development for human progress. Recently, advancements in deep learning and computer vision have offered a promising pathway to improve the sorting process, though significant developmental steps are still required. Enhancing the efficiency of automated waste detection and classification through computer vision could bring substantial societal and environmental benefits. However, classifying and identifying waste materials presents challenges due to the complex and diverse nature of waste, coupled with the limited availability of data on waste management. This paper presents a real-time waste detection and classification system based on the YOLOv8 deep learning model, designed to enhance waste sorting processes at the household level. The proposed system detects and classifies a diverse range of household waste items. Experiments were conducted on a custom waste dataset comprising 3775 images across 17 types of common household waste. The one-stage YOLOv8 model demonstrated superior performance, outperforming traditional two-stage detectors. To improve the accuracy and robustness of the original YOLOv8, five data augmentation techniques and two attention mechanisms were incorporated. Notably, the enhanced YOLOv8-CBAM model achieved a mean average precision (mAP) of 89.5%, a significant improvement with a 4.2% increase over the baseline model. The methodology and improvements applied provide a more efficient and effective AI framework for real-time applications in smart bins, robotic waste pickers, and large-scale recycling systems. Full article

Background/Objectives: Liquid chromatography coupled with mass spectrometry (LC-MS) is a commonly used platform for many metabolomics studies. However, metabolite annotation has been a major bottleneck in these studies in part due to the limited publicly available spectral libraries, which consist of tandem mass spectrometry (MS/MS) data acquired from just a fraction of known compounds. Application of deep learning methods is increasingly reported as an alternative to spectral matching due to their ability to map complex relationships between molecular fingerprints and mass spectrometric measurements. The objectives of this study are to investigate deep learning methods for molecular fingerprint based on MS/MS spectra and to rank putative metabolite IDs according to similarity of their known and predicted molecular fingerprints. Methods: We trained three types of deep learning methods to model the relationships between molecular fingerprints and MS/MS spectra. Prior to training, various data processing steps, including scaling, binning, and filtering, were performed on MS/MS spectra obtained from National Institute of Standards and Technology (NIST), MassBank of North America (MoNA), and Human Metabolome Database (HMDB). Furthermore, selection of the most relevant m/z bins and molecular fingerprints was conducted. The trained deep learning models were evaluated on ranking putative metabolite IDs obtained from a compound database for the challenges in Critical Assessment of Small Molecule Identification (CASMI) 2016, CASMI 2017, and CASMI 2022 benchmark datasets. Results: Feature selection methods effectively reduced redundant molecular and spectral features prior to model training. Deep learning methods trained with the truncated features have shown comparable performances against CSI:FingerID on ranking putative metabolite IDs. Conclusion: The results demonstrate a promising potential of deep learning methods for metabolite annotation. Full article

An inverse-square probability mass function (PMF) is at the Newcomb–Benford law (NBL)’s root and ultimately at the origin of positional notation and conformality. $Pr\left(Z\right)={\left(2Z\right)}^{-2}$, where $Z\in {\mathbb{Z}}^{+}$. Under its tail, we find information as harmonic likelihood $\mathcal{L}\left(\left[s,t\right)\right)=H_{t-1}-H_{s-1}$, where $H_n$ is the nth harmonic number. The global $\mathbb{Q}$-NBL is $Pr\left(b,q\right)={\displaystyle \raisebox{1ex}{$\mathcal{L}\left(\left[q,q+1\right)\right)$}\!\left/ \!\raisebox{-1ex}{$\mathcal{L}\left(\left[1,b\right)\right)$}\right.}={\left(q{H}_{b-1}\right)}^{-1}$, where b is the base and q is a quantum ($1\le q<b$). Under its tail, we find information as logarithmic likelihood $\ell \left(\left[i,j\right)\right)=ln{\displaystyle \raisebox{1ex}{$j$}\!\left/ \!\raisebox{-1ex}{$i$}\right.}$. The fiducial $\mathbb{R}$-NBL is $Pr\left(r,d\right)={\displaystyle \raisebox{1ex}{$\ell \left(\left[d,d+1\right)\right)$}\!\left/ \!\raisebox{-1ex}{$\ell \left(\left[1,r\right)\right)$}\right.}={log}_r\left(1+{\displaystyle \raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$d$}\right.}\right)$, where $r\ll b$ is the radix of a local complex system. The global Bayesian rule multiplies the correlation between two numbers, s and t, by a likelihood ratio that is the NBL probability of bucket $\left[s,t\right)$ relative to b’s support. To encode the odds of quantum j against i locally, we multiply the prior odds ${\displaystyle \raisebox{1ex}{$Pr\left(b,j\right)$}\!\left/ \!\raisebox{-1ex}{$Pr\left(b,i\right)$}\right.}$ by a likelihood ratio, which is the NBL probability of bin $\left[i,j\right)$ relative to r’s support; the local Bayesian coding rule is $\tilde{o}\left(j:i|r\right)={\displaystyle \raisebox{1ex}{$i$}\!\left/ \!\raisebox{-1ex}{$j$}\right.}{log}_r{\displaystyle \raisebox{1ex}{$j$}\!\left/ \!\raisebox{-1ex}{$i$}\right.}$. The Bayesian rule to recode local data is $\tilde{o}\left(j:i|r^{\prime }\right)=\tilde{o}\left(j:i|r\right){\displaystyle \raisebox{1ex}{$lnr$}\!\left/ \!\raisebox{-1ex}{$ln{r}^{\prime }$}\right.}$. Global and local Bayesian data are elements of the algebraic field of “gap ratios”, ${\displaystyle \frac{A-B}{C-D}}$. The cross-ratio, the central tool in conformal geometry, is a subclass of gap ratio. A one-dimensional coding source reflects the global Bayesian data of the harmonic external world, the annulus $\left\{x\in \mathbb{Q}|1\le \left|x\right|<b\right\}$, into the local Bayesian data of its logarithmic coding space, the ball $\left\{x\in \mathbb{Q}|\left|x\right|<1-{\displaystyle \raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$b$}\right.}\right\}$. The source’s conformal encoding function is $y={log}_r\left(2x-1\right)$, where x is the observed Euclidean distance to an object’s position. The conformal decoding function is $x={\displaystyle \frac{1}{2}}\left(1+r^y\right)$. Both functions, unique under basic requirements, enable information- and granularity-invariant recursion to model the multiscale reality. Full article

The exponential growth of Internet of Things (IoT) devices has led to massive volumes of data, challenging traditional centralized processing paradigms. The cloud–edge continuum computing model has emerged as a promising solution to address this challenge, offering a distributed approach to data processing and management and improved performances in terms of the overhead and latency of the communication network. In this paper, we present a novel resource allocation optimization solution in cloud–edge continuum architectures designed to support multiple heterogeneous mobile clients that run a set of applications in a 5G-enabled environment. Our approach is structured across three layers, mist, edge, and cloud, and introduces a set of innovative resource allocation models that addresses the limitations of the traditional bin-packing optimization problem in IoT systems. The proposed solution integrates task offloading and resource allocation strategies designed to optimize energy consumption while ensuring compliance with Service Level Agreements (SLAs) by minimizing resource consumption. The evaluation of our proposed solution shows a longer period of active time for edge servers because of the lower energy consumption. These results indicate that the proposed solution is viable and a sustainability model that prioritizes energy efficiency in alignment with current climate concerns. Full article

The color of the rind is one of the most crucial agronomic characteristics of watermelon (Citrullus lanatus L.). Its genetic analysis was conducted to provide the identification of genes regulating rind color and improving the quality of watermelon appearance. In this study, a mapping population of 505 F₂ plants, derived from a cross between green (CG058) and light-green (CG265) rinds, along with a high-density genetic linkage (average 0.9 cM distance between bin markers), was used to map and identify possible candidate genes. The green rind trait was determined to be regulated by a single Mendelian locus and was precisely located within a 110 kb genomic site on chromosome nine (Chr 9). In the respective region, two potential genes, Cla97C09G175170 and Cla97C09G175180, were substantially downregulated in the light-green rind in comparison to the green rind. Previous studies revealed that Cla97C09G175170, encoding a two-component response regulator-like protein (APRR2), is possibly involved in the green rind trait in watermelon. Virus-induced gene silencing (VIGS) assay confirmed that ClAPRR2 is a key gene responsible for green rind color. Moreover, qRT-PCR analysis revealed that the transcription levels of multiple key genes in the chlorophyll (Chl) biosynthesis pathway were downregulated in the light-green rind relative to the green rind. The current findings have the potential to clarify the regulatory mechanisms that underlie the color of the watermelon rind. These data would provide valuable insights for the targeted molecular design and development of watermelon rinds. Full article

Background/Objectives: Noninvasive brain stimulation (NIBS) can boost motor recovery after a stroke. Certain movement phases are more responsive to NIBS, so a system that auto-detects these phases would optimize stimulation timing. This study assessed the effectiveness of various machine learning models in identifying movement phases in hemiparetic individuals undergoing simultaneous NIBS and EEG recordings. We hypothesized that transcranial direct current stimulation (tDCS), a form of NIBS, would enhance EEG signals related to movement phases and improve classification accuracy compared to sham stimulation. Methods: EEG data from 10 chronic stroke patients and 11 healthy controls were recorded before, during, and after tDCS. Eight machine learning algorithms and five ensemble methods were used to classify two movement phases (hold posture and reaching) during each of these periods. Data preprocessing included z-score normalization and frequency band power binning. Results: In chronic stroke participants who received active tDCS, the classification accuracy for hold vs. reach phases increased from pre-stimulation to the late intra-stimulation period (72.2% to 75.2%, p < 0.0001). Late active tDCS surpassed late sham tDCS classification (75.2% vs. 71.5%, p < 0.0001). Linear discriminant analysis was the most accurate (74.6%) algorithm with the shortest training time (0.9 s). Among ensemble methods, low gamma frequency (30–50 Hz) achieved the highest accuracy (74.5%), although this result did not achieve statistical significance for actively stimulated chronic stroke participants. Conclusions: Machine learning algorithms showed enhanced movement phase classification during active tDCS in chronic stroke participants. These results suggest their feasibility for real-time movement detection in neurorehabilitation, including brain–computer interfaces for stroke recovery. Full article

Background: Soil microbes play a vital role in the ecosystem as they are able to carry out a number of vital tasks. Additionally, metagenomic studies offer valuable insights into the composition and functional potential of soil microbial communities. Furthermore, analyzing the obtained data can improve agricultural restoration practices and aid in developing more effective environmental management strategies. Methodology: In November 2023, sandy soil samples were collected from ten sites of different geographical areas surrounding natural lakes and artificial water points in the Tubaiq conservation area of King Salman Bin Abdulaziz Royal Natural Reserve (KSRNR), Saudi Arabia. In addition, genomic DNA was extracted from the collected soil samples, and 16S rRNA sequencing was conducted using high-throughput Illumina technology. Several computational analysis tools were used for gene prediction and taxonomic classification of the microbial groups. Results: In this study, sandy soil samples from the surroundings of natural and artificial water resources of two distinct natures were used. Based on 16S rRNA sequencing, a total of 24,563 OTUs were detected. The metagenomic information was then categorized into 446 orders, 1036 families, 4102 genera, 213 classes, and 181 phyla. Moreover, the phylum Pseudomonadota was the most dominant microbial community across all samples, representing an average relative abundance of 34%. In addition, Actinomycetes was the most abundant class (26%). The analysis of clustered proteins assigned to COG categories provides a detailed understanding of the functional capabilities and adaptation of microbial communities in soil samples. Amino acid metabolism and transport were the most abundant categories in the soil environment. Conclusions: Metagenome analysis of sandy soils surrounding natural lakes and artificial water points in the Tubaiq conservation area of KSRNR (Saudi Arabia) has unveils rich microbial activity, highlighting the complex interactions and ecological roles of microbial communities in these environments. Full article

The conventional methods used to produce N-acetyl-β-D-glucosamine (GlcNAc) from seafood waste require pretreatment steps that use acids or bases to achieve the extraction and decrystallization of chitin prior to enzymatic conversion. The development of an enzymatic conversion method that does not require the pretreatment of seafood waste is essential for the efficient and clean production of GlcNAc. In this study, the annotated metagenomic assembly data of domesticated microbiota (XHQ10) were analyzed to identify carbohydrate-active enzymes (CAZymes), and an in-depth analysis of the high-quality genome FS13.1, which was obtained from metagenomic binning, was performed; this enabled us to elucidate the catabolic mechanism of XHQ10 by using shrimp shell chitin as a carbon and nitrogen source. The only β-N-acetylglucosaminidase (named XmGlcNAcase) was cloned from FS13.1 and biochemically characterized. The direct production of GlcNAc from shrimp shell powder (SSP) via the use of a chitin enzyme cocktail was evaluated. Under the action of a chitin enzyme cocktail containing 5% recombinant XmGlcNAcase and a crude XHQ10 enzyme solution, the yield and purity of the final conversion of SSP to GlcNAc were 2.57 g/L and 82%, respectively. This is the first time that metagene-derived GlcNAcase has been utilized to achieve the enzymatic conversion of untreated seafood waste, laying the foundation for the low-cost and sustainable production of GlcNAc. Full article

Background/Objectives: COVID-19 has significantly impacted lives, and data show that receiving a booster vaccination has been demonstrated to lower the spread of COVID-19 and reduce the severity of the risk of infection. The Saudi government has actively promoted booster dose vaccines among university students who can spread the virus to older populations, especially in high-density environments, where the risk of virus transmission and spread is elevated. This study focuses on the acceptance of COVID-19 booster shots among students at Imam Abdulrahman bin Faisal University. The study assessed students’ willingness to receive a COVID-19 booster dose and the factors influencing their decision. Methods: A descriptive, cross-sectional study design using an online self-administered survey was conducted among medical and non-medical students at Imam Abdulrahman bin Faisal University. A convenience sampling technique was used to recruit participants via email and social media platforms (WhatsApp version 2.3). Quantitative analysis was performed using IBM SPSS version 28.0. using descriptive statistics. Logistic regression analysis was used to predict factors affecting COVID-19 booster dose acceptance and hesitancy. Results: Among 315 respondents, 171 (54.3%) were males, and 144 (45.7%) were females. All the respondents fell in the 18–25 years age group. About 173 (54.9%) respondents were from health-related colleges. Overall, 24.44% (77/315) agreed to get a COVID-19 vaccine booster dose. However, 77.14% (243/315) were confident of getting the vaccine whenever they wanted. About 48.88% (154/315) of respondents considered COVID-19 a serious severe infection, while 14.06% (46/315) of respondents were concerned about the probability of receiving COVID-19 immunization (World Health Organization, 2021). Conclusions: The study revealed that students were not accepting COVID-19 booster doses, highlighting the need for awareness campaigns to dispel myths and improve vaccination rates. Full article