Articles

Access the latest knowledge in applied science, electrical engineering, computer science and information technology, education, and health.

Advanced Search
Filter Icon

Filters article

Years

FAQ Arrow
Single Year
Range Year
0
0

Source Title

FAQ Arrow

Authors

FAQ Arrow

31,128 Article Results

Attention-enhanced VGG-16 architecture for precision weed detection in wheat fields

10.11591/ijict.v15i3.pp1303-1312
Akanksha Bodhale , Seema Verma , Aishwary Bodhale
This study delineates five advanced convolutional neural network designs that employ deep learning for the classification of wheat weeds. The dataset consists of greyscale photos taken in agricultural fields, enhanced with RGB lighting and cropped to 256×256×3 pixels. Normalization, batch-wise augmentation, contrast stretching, or histogram equalization were all part of the preprocessing that improved picture quality and model learning efficiency. These enhancements developed feature extraction by increasing picture contrast and homogeneity. A VGG16-based model with further Conv2D layers and spatial attention is one of the five models recommended. An additional design, inspired by ResNet50 that applies residual blocks and worldwide average pooling. A hybrid RNN integrated ICNA-CNN or LSTM for spatial-temporal content learning is another. Lastly, InceptionResNetV2 is enhanced with CNN layers and a query-key attention mechanism. We used the Adam optimizer and categorical cross-entropy to estimate the loss throughout the training for the round. At the end, all the models had ReLU activation, batch normalization, MaxPooling2D, and a dropout charge of 0.5 to keep them from overfitting. Upon evaluating their performance based on accuracy, recall, reliability, and training loss, VGG16 emerged as the standout model, achieving an impressive 99.07% precision and a remarkably low training loss of just 0.0079. The study determined that VGG16 is the optimal choice for precise wheat–weed categorization because to its superior generalizability and accuracy.
Volume: 15
Issue: 3
Page: 1303-1312
Publish at: 2026-09-01

A pilot review of the Google cluster workload trace 2019, methodology and its alternatives: analysis of workload in large scale data centres

10.11591/ijict.v15i3.pp1167-1178
Akash Patel , Amit Nayak , Khushi Patel , Anand Patel
Cloud data centres require shared services that are highly available, elastic capacity, managed operations, and robust recovery capabilities to facilitate the next generation of efficient, reliable, and diverse connected computing environments. Nevertheless, large-scale cloud infrastructures continue to fail regularly despite their availability, scalability, and cost efficiency, primarily due to low resource utilisation and inadequate early-stage failure management. The key to effective resource management and minimizing failures in such settings is understanding the nature of the workload and its failure modes. The current review considers the Google cluster workload trace 2019 to investigate workload and failure patterns and to generalise the results of 24 articles. The analysis is also compared with other major datasets, such as Microsoft Azure Trace, Tencent Trace, and Alibaba Trace. The paper establishes the relevance of Google cluster traces, describes the key contents of the 2019 dataset, and contrasts prior literature with respect to research objectives, trace datasets, significant results, and limitations. Moreover, it briefly describes methods for analyzing and modeling cluster traces and identifies gaps in the research that should be addressed to advance the study of cluster traces.
Volume: 15
Issue: 3
Page: 1167-1178
Publish at: 2026-09-01

An AI-powered knowledge graph-based question answering system for Charak Samhita: integrating sanskrit NLP and graph data science

10.11591/ijict.v15i3.pp1197-1207
Sharayu Mirasdar , Mangesh Bedekar
Originating in India, Ayurveda is an ancient medical system focused on holistic healing that considers the mind, body, and spirit. This study utilizes knowledge graph (KG) technology to develop a KG model for an Ayurveda question-and-answer system. The system includes modules for knowledge extraction from चरकसंहिता, कायहचहकत्सा, भैषज्यरत्नावली and द्रव्यगुण संग्रि, construction of KG from this extracted knowledge and construction of AI supported Question answer system. In the methodology, domain-specific KG is constructed in Neo4j. Entities such as diseases (Vyadhi व्याधी), symptoms (Lakshana लक्षण), doshas (दोष), herbs, and treatments are incorporated. Advanced Sanskrit natural language processing (NLP) pipelines using ByT5-Sanskrit, SanskritBERT, and fine-tuned BioBERT facilitate named entity recognition (NER) and relation extraction. Graph-based reasoning models such as graph attention networks (GAT) and graph reasoning enhanced language models (GREASELM) enhance multi-hop reasoning across Ayurvedic concepts. Evaluation was conducted using a gold-standard annotated dataset of Charak Samhita verses mapped to disease–symptom–treatment relationships. Performance metrics included precision, recall, F1-score, mean reciprocal rank (MRR), and overlap coefficient. Superior accuracy can be seen in the proposed model as compared to baseline BERT-QA and subgraph QA approaches. This research has integrated Sanskrit computational linguistics and KG science. The approach mentioned in this paper has mentioned a framework that is scalable, interpretable and culturally significant. With the focus on Ayurveda, the methodology also mentions the potential for developing cross-cultural medical questions–answering systems, thereby bridging ancient wisdom with modern technological approaches.
Volume: 15
Issue: 3
Page: 1197-1207
Publish at: 2026-09-01

Wavelet-based spectrum sensing with improved thresholding for enhanced detection in cognitive radio networks

10.11591/ijict.v15i3.pp1123-1134
Nur Hanis Abdul Rani , Mas Haslinda Mohamad , Nurusolihah Zamri , Nor Khairiah Ibrahim
Cognitive radio (CR) technology is an adaptive, intelligent radio and network technology that can automatically detect available channels in a wireless spectrum. Spectrum sensing is the most important component in CR due to its ability to sense and recognize parameters related to the radio channel characteristics. However, there are some spectrums that are not used known as spectrum holes. It is challenging to accurately identify these spectrum holes, especially when employing traditional energy detection techniques, which suffer from incorrect threshold selection at low signal-to noise ratio (SNR) levels. This work suggests a wavelet-based spectrum sensing technique in conjunction with an enhanced thresholding method to improve detection accuracy and decrease noise to overcome this constraint. MATLAB simulations are used for evaluating three threshold functions: hard, soft, and improved. The results indicate that the improved threshold achieves superior denoising performance and a higher detection probability compared to the traditional energy detection method. In this study, the energy detection technique was also implemented for comparison with the wavelet-based approach. The findings reveal that wavelet-based sensing consistently provides a higher detection probability (𝑃𝐷𝐸𝑇), demonstrating its effectiveness and reliability for cognitive radio (CR) application.
Volume: 15
Issue: 3
Page: 1123-1134
Publish at: 2026-09-01

Classification of encryption attacks and strategies for mitigation

10.11591/ijict.v15i3.pp1238-1253
Anas Maaifi , Khalid Zine-Dine
Cryptography is essential for securing digital communications, yet it remains vulnerable to various malicious attacks. These attacks can be classified based on the type of cryptography they target symmetric or asymmetric. This paper presents a comprehensive classification of encryption attacks, examining the specific vulnerabilities associated with each cryptographic approach. By analyzing these attack vectors, the study shows the importance of understanding weaknesses in cryptographic systems. Furthermore, it proposes several mitigation strategies to strengthen defenses and enhance protection of sensitive information in the digital domain.
Volume: 15
Issue: 3
Page: 1238-1253
Publish at: 2026-09-01

Metaverse based immersive learning prototype for satellite communication using silvercoms model

10.11591/ijict.v15i3.pp1408-1418
Komputerio Akbar , Meyliana Meyliana , Harco Leslie Hendric Spits Warnars , Ilvico Sonata
Advancements in immersive technologies and the metaverse are transforming engineering education. Satellite communication, as a complex and abstract domain, requires innovative approaches to enhance conceptual understanding and learner engagement beyond traditional methods. This study proposes the Silvercoms model, an immersive learning framework tailored for satellite communication systems education, integrating metaverse technology with pedagogical design. The model combines 3D interactive simulations, collaborative virtual environments, and systems-level content delivery, structured using the 6E instructional model and supported by the motivated strategies for learning questionnaire (MSLQ) to address both cognitive and motivational aspects of learning. The system is developed using a systems engineering approach and implemented with Unity3D in a virtual reality (VR) based metaverse environment. The prototype includes modules such as satellite orbit simulation, satellite history, and interactive satellite systems, enabling experiential and concept-driven learning. This study contributes by integrating immersive technology with structured pedagogical frameworks, offering a novel approach to improving learning effectiveness in satellite communication systems education.
Volume: 15
Issue: 3
Page: 1408-1418
Publish at: 2026-09-01

Exploring multi-answer visual question answering with object detection: a systematic review

10.11591/ijict.v15i3.pp1097-1114
Nidaul Hasanati , Taufik Djatna , Imas Sukaesih Sitanggang , Arif Imam Suroso
Visual question answering (VQA) is a challenging research area that enables machines to answer natural language questions based on visual content by jointly understanding images and text. Conventional VQA systems typically produce a single answer for each image–question pair. However, many real world visual questions are ambiguous or complex, allowing multiple valid answers to exist. This systematic literature review (SLR) focuses on multi answer VQA systems and the use of object detection, following the PRISMA 2020 guidelines. We analyzed 58 peer-reviewed journal articles retrieved from the Scopus database published between 2020 and 2025. Ten of these studies clearly stated that generating multiple answers was their main goal. Forty-eight others indirectly supported answer variability by using object-based or multi-instance reasoning. Through this review, we examine the current methodologies for supporting multi-answer generation, including model architecture, datasets, and evaluation metrics. Most multi answer generation approaches utilize attention mechanisms, graph neural networks, and transformer-based models. Additionally, we propose a taxonomy of multi-answer VQA organized along four dimensions. Limitations are identified in datasets and evaluation metrics (i.e., answer ambiguity/subjectivity). Future research should focus on improving model interpretability and designing an evaluation framework that incorporates subjective and context-sensitive responses.
Volume: 15
Issue: 3
Page: 1097-1114
Publish at: 2026-09-01

Spatial-channel reconstruction for efficient multiscale attention in robotic object detection

10.11591/ijra.v15i3.pp544-552
Mohammed Maiza , Chahira Cherif , Samira Chouraqui , Abdelmalik Taleb-Ahmed
Real-time object detection is a core capability for autonomous robots, unmanned aerial vehicles (UAVs), and self-driving systems operating in resource-constrained environments. This paper presents spatial-channel enhanced multiscale attention (SCEMA), a novel lightweight attention module designed to enhance robotic perception while minimizing computational overhead for embedded deployment. SCEMA employs a parallel dual-branch architecture that synergistically combines spatial-channel reconstruction with multiscale attention mechanisms. When integrated into the YOLOv8n framework (3.01M baseline parameters), the proposed YOLO-SCEMA model achieves significant performance gains across multiple challenging benchmarks relevant to robotics automation. Experiments on an NVIDIA RTX 4080 GPU demonstrate that on the ExDark dataset, YOLO-SCEMA improves mAP@50 by 7.37% over the baseline (69.07% to 76.44%) while reducing parameters by 36.88% (3.01M to 1.90M) and computational cost by 8.64% (8.1 to 7.4 GFLOPs). Consistent improvements are also observed on VisDrone2019 (+3.24% mAP@50) and FYP (+1.50% mAP@50) datasets. Comparative analysis demonstrates that YOLOSCEMA achieves superior accuracy-efficiency trade-offs, making it particularly suitable for deployment in low-light conditions, dense scenes, and complex structural environments for autonomous navigation, robotic surveillance, and industrial automation applications.
Volume: 15
Issue: 3
Page: 544-552
Publish at: 2026-09-01

Design of a portable IoT robot with azure machine learning for monitoring mine workers’ health

10.11591/ijra.v15i3.pp577-588
Shanthi Natarajan , Vijayaraja Loganathan , Dhanasekar Ravikumar , Diwakar Venkat Nalini , Harish Elangovan , Balaji Arikrishnan
The mining environment exposes workers to physical, environmental, and health risks. The lack of effective real-time health monitoring systems leads to delayed medical responses. Hence, this paper discusses developing a portable Internet of Things (IoT) robot with advanced machine learning and cloud computing to monitor mine workers’ health and send out alerts. The system is equipped with IoT sensors that monitor parameters continuously, such as heart rate, body temperature, blood pressure, and environmental factors (gas concentrations, air quality). Data collected in real-time is transmitted to a cloud-based platform for analysis using advanced machine learning algorithms. MQ-135 detects harmful gases, and DHT11 measures humidity and transmits data to the Arduino UNO. The HC-SR04 sensor measures object distances by emitting ultrasonic waves and detecting their echoes, aiding in obstacle detection. The NEO-6M GPS with GSM SIM900 modules transmit location data and emergency alerts via the GSM network, enabling responses to potential dangers. Simulation via Proteus validates the robot’s transceiver connectivity, mobility, and sensing functions. To enhance monitoring precision, the system adopts XGBoost, which classifies mine conditions, and the training model achieves 96.77% accuracy with high precision and recall. The system with Azure Machine Learning improves detection accuracy, raising temperature, CO, NH₄, and NO₂ precision by 7.25%, 15%, 17%, and 18%, respectively. Thus, the system features an intelligent alert mechanism to notify users of emergencies, enhancing worker safety and minimizing health-related risks in mining operations.
Volume: 15
Issue: 3
Page: 577-588
Publish at: 2026-09-01

Comparative study of MPPT algorithm on photovoltaic string under partial shading

10.11591/ijape.v15.i3.pp1422-1438
Dikpride Despa , Gigih Forda Nama , Zulmiftahul Huda , Stefanus Debiarto Marudut Sagala
Partial shading significantly degrades the photovoltaic (PV) performance value by introducing multiple peaks in power voltage (P-V) curve, complicating maximum power point tracking (MPPT). This research aims to presents a systematic comparative study of 4 MPPT algorithms, that are i) perturb and observe (P&O), ii) incremental conductance (InC), iii) particle swarm optimization (PSO), and iv) flower pollination algorithm (FPA), under 6 systematically testbeds modeled irradiance scenarios. The evaluation focused on tracking accuracy, convergence speed, and also robustness against local maxima entrapment. The findings indicated that slope-based algorithms (P&O and InC algorithm) achieved rapid convergence (
Volume: 15
Issue: 3
Page: 1422-1438
Publish at: 2026-09-01

Conversational AI in museums: a systematic literature review using the people–process–technology framework

10.11591/ijict.v15i3.pp955-966
Diana Utomo , Siti Elda Hiererra
The adoption of conversational artificial intelligence (AI) in museums has opened a new opportunity to create a richer visiting experience to tell stories for the preservation of culture. This paper contributes a systematic literature review (SLR) of 33 peer-reviewed papers covering the period from 2020 to mid-2025, by applying the people–process–technology (PPT) model to examine the social technological aspects of AI implementation. This combination is novel in a museum context, as previous research has largely treated these separately. Findings indicate a pronounced shift from text based chatbots (21% or 7 of 33 papers) to more immersive and interactive platforms (30% or 10 of 33 papers), reflecting the transition from the pandemic (2020 – 2023) to the post-pandemic period (2024 – 2025). Besides the evolution of these technologies, the technology component highlights the importance of data governance, digital preparedness, and value alignment at and across different levels. The people component includes the relevance of hedonic and utilitarian values. Meanwhile, the process component emphasizes both strategic and technical aspects of AI design and implementation, such as knowledge structuring, media selection, and narrative representation. In line with this agenda, this study addresses the trends in literature and provides a path for sustainable appropriation by museums.
Volume: 15
Issue: 3
Page: 955-966
Publish at: 2026-09-01

Prototype of real-time Mexican sign language classifier

10.11591/ijict.v15i3.pp986-994
Alan Ramírez-Noriega , Yobani Martínez-Ramírez , Samantha Jiménez , Marcos Murillo-Corrales
Mexican sign language (MSL) is the language used by the deaf community in Mexico. Like Spanish, it has its own distinct grammar, syntax, and vocabulary. However, instead of relying on sounds, MSL conveys meaning through gestures, facial expressions, and body movement. This research proposes the creation of an image dataset of the MSL alphabet for real-time sign detection. A neural network model was developed to recognize these signs, achieving an accuracy of approximately 60%. Although this result is modest, the study establishes a foundation for future work that could facilitate communication for MSL users or lead to the development of educational applications for language learning.
Volume: 15
Issue: 3
Page: 986-994
Publish at: 2026-09-01

Prognosis of vector borne dengue disease outbreak in urban areas using multivariate analysis

10.11591/ijict.v15i3.pp1066-1077
Pratik S. Machchar , Purvi N. Ramanuj , Rajan Patel , Jitendra Bhatia , Kuntesh Jani
Vector borne disease like dengue continues to pose a significant climate-sensitive public health challenge in tropical regions such as Brazil, Peru, and India. This study examines the feasibility of predicting dengue outbreaks using weekly multivariate time-series data from San Juan (SJ), Puerto Rico and Iquitos (IQ), Peru. Dengue incidence was analyzed alongside meteorological, environmental, and vegetation-based variables to capture key climatic influences. Several machine learning and deep learning approaches were evaluated, including LightGBM. Model performance was assessed using root mean square error (RMSE) and mean absolute error (MAE). The results show that LightGBM achieved the low est RMSE/MAE, indicating strong short-term predictive accuracy and excellent interpretability. Feature importance analysis and principal component analysis (PCA) identified precipitation, dew point temperature, and humidity as the most influential predictors of dengue incidence. The study demonstrates that advanced machine learning models can serve as reliable early warning systems for vector-borne diseases. While this research focuses on dengue, the methodology is adaptable to other vector-bone datasets and diseases, offering a flexible tool for public health authorities to predict and mitigate outbreaks in diverse urban contexts.
Volume: 15
Issue: 3
Page: 1066-1077
Publish at: 2026-09-01

A recent hybrid of IoT with adaptive extended Kalman filter fuzzy logic for children’s health dietary

10.11591/ijict.v15i3.pp1217-1225
Noorrezam Yusop , Massila Kamalrudin , Mohd Nazrien Zaraini , Siti Fairuz Nurr Sardikan
The increasing prevalence of childhood obesity highlights the critical need for intelligent dietary monitoring systems that are tailored to individual nutritional requirements. This work describes the creation and testing of an internet of things (IoT) hybrid using an adaptive extended Kalman filter and fuzzy logic (AEKFFL-IoT) model aimed at providing personalised food calorie prediction for children. The system uses IoT devices to collect real-time sensor data, such as height, weight, and BMI, and then employs extended Kalman filter (EKF) algorithms to denoise signals and anticipate trends. Fuzzy logic inference is then utilised to adaptively calculate caloric requirements based on biometric data. Experimental results reveal that the proposed AEKFFL model has a training root mean square error (RMSE) of 21.43 kcal and a testing RMSE of 22.35 kcal, exceeding existing rule-based, wearable, and ANN-driven models in terms of accuracy and generalisation. Furthermore, the system achieves high classification accuracy (94.5%) for BMI categorisation and fuzzy rule application. Comparative examination confirms the model’s adaptability, real-time integration, and mobile deployment capability. This study presents a scalable and intelligent approach for child-centered dietary monitoring, paving the path for personalised digital health interventions.
Volume: 15
Issue: 3
Page: 1217-1225
Publish at: 2026-09-01

Intelligent engineering framework for managing hospital cardiac arrest resources

10.11591/ijict.v15i3.pp1290-1302
Chams Eddine Fathoun , Mohamed Ridda Laouar , Safa Abid , Sean B. Eom
In-hospital cardiac arrest in intensive care remains frequent (often cited incidence roughly 0.5%-7.8% of admissions), while causes differ in what staff and equipment must be ready. We ask whether vital-sign trajectories from a standard EHR can classify which of three cardiac-related mechanisms is most salient arrhythmia, acute myocardial infarction (AMI), or respiratory failure or hypoxia so ICU resources can be aligned with risk. Using MIMIC-IV, we extracted diagnoses and charted vitals in the 12 hours before the index event, applied cleaning, aggregation, label encoding, sequence padding, and class balancing (3,000 cases per class), then trained and compared eXtreme gradient boosting (XGBoost), random forest (RF), support vector machine (SVM), and logistic regression (LR) with 5-fold cross-validation on an 80/20 split. XGBoost performed best (about 93% accuracy; sensitivity 89.15%; specificity 90.43%; AUC-ROC 0.94). Feature importance highlighted heart rate, oxygen saturation, and blood pressure patterns consistent with bedside monitoring practice. The study supports mechanism-oriented triage labels derived from widely recorded vitals, as a complement to generic early warning scores, for prioritizing telemetry, respiratory support, and cardiology pathways. External validation and prospective evaluation are needed before deployment.
Volume: 15
Issue: 3
Page: 1290-1302
Publish at: 2026-09-01
1 2 3
...
2075 2076
Show 16 of 2076

Discover Our Library

Embark on a journey through our expansive collection of articles and let curiosity lead your path to innovation.

Explore Now
Library 3D Ilustration