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

31,377 Article Results

A novel approach for real-time traffic sign recognition framework

10.11591/csit.v7i2.p224-230
Kshatrapal Singh
Traffic sign recognition plays a critical role in enhancing road safety and enabling autonomous driving systems. This paper presents a comprehensive approach to real-time traffic sign recognition using advanced computer vision techniques and machine learning models. The proposed system employs convolutional neural networks (CNNs) for accurate detection and classification of traffic signs under diverse environmental conditions, including varying lighting, weather, and occlusions. Real-time processing is achieved through the integration of optimized algorithms and hardware acceleration techniques, ensuring minimal latency and high throughput. Experimental results demonstrate that the system achieves state-of-the-art performance on benchmark datasets, with an accuracy of over 95% and a recognition speed suitable for real-world applications. The findings underscore the potential of the system to improve driver assistance systems and pave the way for safer autonomous vehicles.
Volume: 7
Issue: 2
Page: 224-230
Publish at: 2026-07-01

Improving Botnet host prediction with encryption and GRU for enhanced network security

10.11591/csit.v7i2.p141-158
Omega Joel Patria Moata , Irwansyah Saputra
This paper examines the challenges of reliably and securely predicting Botnet hosts, a crucial aspect of network security. Existing Botnet detection systems often fail to address data privacy concerns and struggle with evolving attack methods. This study proposes an innovative approach to improve the security and accuracy of Botnet host prediction by integrating deep learning with encryption. The proposed method employs encryption techniques such as data encryption standard (DES) and blum-blum-shub (BBS) to protect sensitive data in a text data set of 2,100 IP addresses, consisting of Botnet hosts and benign hosts. Several pre-processing techniques, including moving average and missing value handling, are implemented to optimize the model performance. The effectiveness of the system is evaluated using performance metrics such as F1-score, recall, accuracy, and precision. Experimental results show that the proposed approach significantly outperforms existing methods in accuracy, which have not achieved the maximum accuracy per IP Host within a given time frame, while providing enhanced security through encryption on text data. The study concludes that combining deep learning with encryption on text data offers a promising solution for reliable and secure Botnet host prediction data. Future research will focus on testing larger and more diverse data sets, as well as analyzing the impact of different encryption techniques on the overall accuracy and security of the system.
Volume: 7
Issue: 2
Page: 141-158
Publish at: 2026-07-01

Matter protocol-enabled device onboarding for cross-platform internet of things systems

10.11591/ijres.v15.i2.pp406-415
Geetishree Mishra , Hemavathi Hemavathi , Harish V Mekali
The Matter protocol, created by the connectivity standards alliance (CSA), comes in with a single standard to make sure these devices can connect and be controlled across platforms like Google Home, Apple HomeKit, Amazon Alexa, and Samsung SmartThings. The rapid expansion of the internet of things (IoT) is driving the urgent need for secure and efficient onboarding processes for a wide range of connected devices. It necessitates a robust framework to seamlessly integrate new additions into existing systems while upholding security standards. This initiative focuses on implementing the Matter protocol on ESP32 devices, employing a Raspberry Pi hub as the central communication point to facilitate smooth device-to-hub interactions. This work presents the onboarding devices for interconnected IoT systems using the Matter protocol. The Matter device is configured and tested within the Amazon ecosystem using an Alexa Echo Dot, as well as with the smart home assistant ecosystem along with a smartphone application. By configuring the Raspberry Pi hub as a designated Matter hub and exploring interactions within the home assistant ecosystem supporting diverse platforms like Apple HomeKit and Google Home, the work enhanced interoperability and broadened the utility of IoT devices within an interconnected network. This initiative forges a foundation for an adaptable and cohesive IoT environment.
Volume: 15
Issue: 2
Page: 406-415
Publish at: 2026-07-01

Performance evaluation of the deep learning system for weed recognization

10.11591/csit.v7i2.p167-178
Abd Abrahim Mosslah , Reyadh Hazim Mahdi , Hassan Kassim Albahadily
Numerous approaches based on machine learning have emerged in recent years to enhance crop protection efficiency. One example is the utilization of deep neural networks (DNNs) to differentiate between various weed types in actual events scenarios. Nevertheless, these methods often need substantial input from experts who work iteratively to design the robust deep learning system. To simplify such process and conserve resources, researchers have explored a fresh method known as automated deep learning our technology’s recognization of weeds through the use of machine learning was evaluated using plant seedlings and weed collections from plants dataset to address a issue of weed recognization. The study compared various configurations, including plant segmentation, using a collection of classifiers in place of Softmax, and training with datasets that contain noise. The findings indicated ensuring performance, with F1-scores of 93.1% and 90.2% based on the dataset utilised. These results align together with automated machine learning (AutoML-linked) studies, while fall short of manually fine-tuned deep-learning-based systems created through human specialists. To conclude, exploring the potential of combining manual expert work and automated deep learning could be a promising direction for enhancing efficiency in plant defence.
Volume: 7
Issue: 2
Page: 167-178
Publish at: 2026-07-01

A real-time multi-modal deep learning framework for student attentiveness assessment in online learning environments

10.11591/ijres.v15.i2.pp450-460
Rajasekaran Mariswamy , P.V. Praveen Sundar
The rapid growth of online learning platforms has increased the need for intelligent systems capable of monitoring student attentiveness in real time to improve learning effectiveness and adaptive instruction. This paper proposes a multi-modal deep learning framework for attentiveness assessment by integrating visual, behavioral, and temporal information extracted from online classroom interactions. The proposed system consists of four major components, namely data acquisition, preprocessing and normalization, deep feature extraction with temporal learning, and attentiveness evaluation with analytics generation. Visual and spatial characteristics are learned using a convolutional neural network (CNN), while temporal behavioral patterns are captured through a long short-term memory (LSTM) network to model sequential engagement dynamics. The framework is designed to operate in both real-time and offline modes, enabling live monitoring during virtual classes as well as post-session analysis of recorded lectures. The computational pipeline is optimized through fixed-point processing, parallel convolution execution, and latency-aware temporal modeling, making it suitable for field programmable gate array (FPGA)-based and embedded implementations under constrained computational resources. Experimental evaluation conducted on an in-house dataset demonstrates that the proposed framework achieves 92.9% classification accuracy and a 91.9% F1-score, while maintaining strong generalization capability on cross-dataset benchmarks. Furthermore, latency analysis shows an average processing time of 31.6 ms per frame, enabling near real-time inference at approximately 30 frames per second.
Volume: 15
Issue: 2
Page: 450-460
Publish at: 2026-07-01

Binary hybrid pathfinder algorithm for efficient feature selection in resource-constrained embedded systems

10.11591/ijres.v15.i2.pp504-513
Rahul Mirajkar , Premanand Ghadekar , Vijay Dasharath Chougule , Renuka Bhandari , Hridaynath Khandagale , Mahavir A. Devmane , Mangesh Hajare , Kuldeep B. Vayadande
Feature selection is critical for embedded machine learning systems where computational resources and memory are severely constrained. This paper presents the binary quadratically interpolated hybrid pathfinder algorithm (BQIHPFA), a novel metaheuristic optimization method designed for efficient feature subset selection in resource-limited classification tasks. BQIHPFA adapts the continuous QIHPFA to binary search spaces through sigmoid transfer functions and employs a hybrid two-group enhancement strategy combining pathfinder dynamics with salp swarm algorithm-inspired exploration. We evaluate BQIHPFA against three established binary optimization algorithms (binary particle swarm optimization (BPSO), binary grey wolf optimizer (BGWO), and binary whale optimization (BWO)) on three benchmark datasets with varying dimensionalities: Língua Brasileira de Sinais (Brazilian Sign Language) movement (90 features), Parkinson's disease detection (22 features), and Sonar Rock vs. Mine (60 features). Experimental results demonstrate that BQIHPFA achieves competitive classification accuracy (average 83.57%) with substantial feature reduction (average 64.1%) while executing 5.2 times faster than complex baselines and consuming minimal memory (peak: 45-58 MB). Ablation experiments demonstrate that every algorithmic part makes a 8-24% contribution to the total performance. BQIHPFA offers an easy-to-use, non-specific feature selection method to automated resource-constrained embedded classification systems, applicable to be deployed to low-power computing environments, and internet of things (IoT) edge systems.
Volume: 15
Issue: 2
Page: 504-513
Publish at: 2026-07-01

Fuzzy logic–based consensus protocol for educational blockchain networks

10.11591/csit.v7i2.p131-140
Igor Ivanov , Svetlana Zhdanova
This paper addresses the growing challenge of ensuring trust, authenticity, and transparency in the management and verification of educational credentials within modern, digitally oriented learning ecosystems. Rapid expansion of e-learning, lifelong learning, and global mobility has intensified document fraud, revealing the limitations of traditional verification mechanisms. To respond to these systemic risks, the study proposes a socially oriented block-validation protocol integrated into a distributed blockchain environment designed specifically for educational data security. The protocol forms the core of the EduBLOCK system, developed by the authors, and introduces an innovative consensus mechanism that incorporates human-centered reputation assessments rather than computational or financial power. The approach employs fuzzy-set theory to evaluate user activity, institutional credibility, and delegate reputation, enabling a more nuanced and context-sensitive model of trust. Delegates responsible for validating blocks are selected through a dynamic, reputation-driven procedure that excludes financial contributions and subjective parameter tuning. The proposed algorithm combines cryptographic guarantees, peer-to-peer (P2P) communication, and soft-computing methods to ensure fairness, prevent manipulation, and maintain stable system functioning. Block validity is determined through open voting, requiring approval by more than two-thirds of elected delegates.
Volume: 7
Issue: 2
Page: 131-140
Publish at: 2026-07-01

Analysis of a compact wideband DGS-inspired octagonal patch antenna for sub-6 GHz 5G and IoT wireless systems

10.11591/ijres.v15.i2.pp479-489
Komalavalli Subramanian , Divya Subramani , Sornalatha Ravindran , Muthu Manickam Anbarasu , Mohan Chinnasamy , Anita Daniel
The proposed compact wideband antenna is developed to meet the increasing demand for efficient and miniaturized radiators in sub-6 GHz fifth generation (5G) and internet of things (IoT) wireless systems. The design features an octagonal radiating patch integrated with modified H-shaped slots to enhance the current path and impedance matching, while a graded defected ground structure (DGS) is introduced to improve bandwidth (BW) and suppress unwanted surface wave effects. Fabricated on an FR4 substrate and energised by a simple stripline feed, the antenna maintains a compact size of 18×15 mm² without compromising performance. It achieves a wide fractional BW of 42.81% spanning 3.1–5.8 GHz, with a resonance centered at 4.6 GHz and obtained reflection coefficient of −36 dB, indicating excellent impedance matching. Additionally, the suggested antenna provides the maximum gain of 3.14 dB and an overall radiation efficiency of 80.5%, demonstrating stable radiation characteristics while making it ideal for small, low-profile 5G and IoT communication devices.
Volume: 15
Issue: 2
Page: 479-489
Publish at: 2026-07-01

Tracking a person and determining the location by using convolutional neural network technology

10.11591/csit.v7i2.p203-213
Zinah Shiker Makki , Ahmet Zengin
Tracking individuals in real-world environments requires robust, non-intrusive methods that overcome the limitations of device-based systems. This study proposes a convolutional neural network (CNN)-driven person-tracking framework that identifies targeted individuals directly from camera feeds, eliminating the need for wearable or global positioning system (GPS) devices and addressing a major drawback of traditional tracking technologies. The system utilizes a TensorFlow-trained CNN model that can detect, recognize, and locate persons of interest in real-time, even under varying illumination conditions. Unlike conventional approaches, our method integrates facial feature extraction with encrypted identity management, enabling secure multi-person detection and rapid location reporting. Experimental results demonstrate a 92% accuracy in low-light settings and 100% accuracy under normal lighting, confirming the system’s effectiveness for security-oriented applications. The findings highlight the novelty of combining lightweight CNN architecture, real-time facial recognition, and hash-based identity protection within a unified tracking pipeline.
Volume: 7
Issue: 2
Page: 203-213
Publish at: 2026-07-01

Revolutionizing night-time object detection in autonomous vehicles with SCL-YOLOv11 and ROA optimization

10.11591/ijres.v15.i2.pp534-552
Kondapalli Sri Vijaya , Gokula Krishnan Vasudevan , Pinagadi Venkateswara Rao , Therasa Michael , Balasubramanian Lalithambigai , Boddula Prathusha Laxmi
Accurate object detection under low-light conditions is a critical requirement for reliable perception in autonomous driving systems. However, night-time environments often suffer from poor illumination, noise, and reduced feature visibility, which significantly degrade the performance of conventional object detection models. To address this challenge, this paper proposes spatial contrast learning (SCL)-you only look once version 11 (YOLOv11), an enhanced object detection framework designed for night-time scenarios. The proposed approach integrates SCL to improve feature discrimination in dark regions and employs the revolution optimization algorithm (ROA) for effective model parameter optimization. The framework is evaluated on three benchmark night-time datasets, ExDark, LLVIP, and BDD100K, to assess its detection performance. Experimental results demonstrate that the proposed model achieves a mAP@50 of 72.9%, improving the baseline YOLOv11 by 9.5% while also reducing inference latency by 18.3%. Comparative evaluations with existing detectors further confirm that the proposed method provides improved accuracy and efficiency for night-time object detection. These results indicate that the proposed framework can enhance perception reliability for autonomous driving applications operating in low-light environments.
Volume: 15
Issue: 2
Page: 534-552
Publish at: 2026-07-01

Adaptive fractional-order PID-controlled DVR optimized by zebra algorithm for harmonic suppression

10.11591/ijeecs.v42.i3.pp786-797
Milind Paraye , Rajendra G. Sutar
Dynamic voltage restorers (DVRs) are widely employed to mitigate power quality disturbances in modern power grids. Existing DVR control strategies frequently struggle to adequately suppress harmonic distortions and voltage sags due to nonlinear grid behaviour, rapidly varying disturbances, and limited tuning flexibility. We suggest a grid-connected DVR with an adaptive fractional order proportional integral derivative (FOPID) controller whose parameters are improved using an improved zebra algorithm (IZA) in order to close this gap. The IZA algorithm is used to improve the FOPID controller parameters, ensuring rapid convergence and superior accuracy. The effectiveness of the proposed system is assessed under two different operating conditions. In case 1, the harmonic compensation is analyzed, in which the DVR reduces systemic harmonic disturbances. The results reveal that the proposed controller reduces the total harmonic distortion (THD) from 1.36% to 0.01% while maintaining a constant voltage amplitude of around 0.9986 V, demonstrating strong harmonic suppression capability. Voltage sag mitigation is assessed in Case 2. The load voltage is effectively restored from 0.722 V to 0.9986 V by the DVR, which also reduces THD from 32.97% to 1.6% by injecting the required compensatory current. Overall, the results confirm that the adaptive FOPID–IZA controlled DVR significantly improves power quality and voltage stability in grid-connected systems by effectively mitigating both harmonic distortion and voltage sags.
Volume: 42
Issue: 3
Page: 786-797
Publish at: 2026-06-10

Evaluation of aerodynamic and structural design to enhance solar energy absorption for e-Cars

10.11591/ijeecs.v42.i3.pp649-665
Mohamed Abubakr Mahgoub Hassan , Belal Ahmed Hamida , El Sayed Soliman , Muhammed Zaharadeen Ahmed
The integration of photovoltaic (PV) systems into electric vehicles (EVs) offers a promising solution for extending driving range and reducing dependence on grid-based charging. However, vehicle-integrated PV systems are limited by aerodynamic drag, structural integration challenges, thermal losses, and inefficient energy management. This study presents a multidisciplinary simulation framework to evaluate aerodynamic and structural optimization strategies for enhancing solar energy absorption in EVs. Computational fluid dynamics (CFD) was used to analyze airflow and reduce aerodynamic drag, while finite element analysis (FEA) assessed structural integrity and weight optimization after PV integration. PV energy flow and thermal models were also developed to evaluate power generation, battery charging behavior, and temperature-dependent efficiency losses. The optimized design reduced the drag coefficient from 0.310 to 0.236, a 23.7% improvement, while maintaining a structural safety factor above 1.75 through lightweight composite materials. The optimized PV configuration increased solar conversion efficiency from 17.6% to 22.3% and daily energy generation from 2.91 kWh/day to 3.75 kWh/day, corresponding to a 28.9% increase in harvested energy. Thermal management strategies lowered average PV operating temperature by about 12 °C, improving efficiency by an additional 5%–7%. Unlike existing studies that examine aerodynamic, structural, or PV performance separately, this work provides a unified framework that evaluates their combined impact on solar energy harvesting in EVs. The proposed integrated design approach demonstrates that coordinated aerodynamic, structural, thermal, and energy-management optimization can substantially improve the practicality and energy contribution of solar-assisted EVs in high-irradiance environments.
Volume: 42
Issue: 3
Page: 649-665
Publish at: 2026-06-10

Development of an IoT-based waste monitoring and notification system for smart environmental management

10.11591/ijeecs.v42.i3.pp729-741
Enggar Utari , Ika Rifqiawati , Wahyuni Martiningsih , Izzal Ihasani , Aditya Rahman , Bagus Dwicahyono
Rapid urban population growth has intensified solid waste generation, while many existing waste management systems still rely on manual inspection and single-parameter monitoring, resulting in delayed responses and inefficient handling. Previous studies have primarily focused on isolated sensing or offline monitoring, highlighting the need for integrated, real-time, and user-oriented waste monitoring solutions. This study used RnD method, proposes a smart garbage level and information hub (SIGALIH), an IoT based waste monitoring and notification system designed to address these limitations. SIGALIH combines multi-parameter sensing, including waste level, temperature–humidity, gas concentration, and ambient light, with an ESP32 microcontroller, a cloud-based data platform, and a real-time notification service using a messaging bot. System evaluation involved sensor accuracy testing, communication latency analysis, and functional verification. Experimental results indicate an average sensor accuracy of 96.8%, with an average data transmission latency of 1.84 seconds and a notification delay of 2.14 seconds, indicating reliable real-time performance under varying network conditions. Functional testing confirmed stable operation of all system modules. The system was also integrated into an Environmental Education learning module to support environmental literacy and awareness through contextual learning on sustainable waste management. SIGALIH is designed for small- to medium-scale urban and community-based applications. However, performance depends on wireless network availability, which may reduce reliability in low-connectivity areas. Overall, SIGALIH provides a low-cost, scalable, integrated solution supporting smart environmental management and sustainable urban waste initiatives.
Volume: 42
Issue: 3
Page: 729-741
Publish at: 2026-06-10

Voice portraits: building faces through voice analysis

10.11591/ijeecs.v42.i3.pp902-912
Anandhu T. G. , John K. Joseph , Navneeth Krishnan J. , Richu Shibu , Elizabeth Isaac
Generation of a person’s appearance from their voice alone is an intriguing challenge. The proposed framework centers on recreating a person’s facial image based solely on a short audio recording of that person speaking. Using a deep neural network trained on millions of YouTube recordings where faces and voices appear together, the system learns voice-face relationships, enabling it to generate images that capture physical traits such as age, gender, and ethnicity. Operating in a self-supervised manner, this method takes advantage of the pairing of faces and voices in online videos, eliminating the need for explicit property modeling. The model achieved a classification accuracy of (95%) for gender, (83%) for age, and (65%) for race prediction from voice inputs, demonstrating an exceptional performance in demographic trait identification. The generated images are evaluated against real photographs of the speakers, assessing how closely these reconstructions resemble actual appearance. This framework has practical applications in forensic analysis, security systems, and privacy-conscious biometric identification, offering a non-invasive alternative to traditional facial recognition methods.
Volume: 42
Issue: 3
Page: 902-912
Publish at: 2026-06-10

Velocity hemodynamic patterns in aortic valve stenosis: a study of inlet velocity during systole phase

10.11591/ijeecs.v42.i3.pp884-891
Nur’Afifah Yousri , Nabilah Ibrahim , Ishkrizat Taib
This work highlighted a close-up version of the aortic valve that provide detail parameter and clearer graphics compared to the 3D version. Therefore, four simplified models are designed and simulated by using computational fluid dynamics (CFD) which are one healthy valve model (100%) and three stenotic models with varying valve opening (70%, 50%, and 30%). The model dimensions and setup parameters are determined by comparing the healthy aortic valve with the previous data. The analysis focused on two different views, which are the view on a targeting line velocity along the x-axis, and the view at the y-axis around the aortic valve. Results on the evaluation graph at the x-axis and y-axis show significant differences in flow patterns between healthy and aortic valve stenosis. The healthy model of 100% valve opening depicted a lower velocity (m/s) at 1.5m/s compared to the stenotic model of 70%, 50%, and 30% valve opening that showed higher velocities of 3.24 m/s, 6.09 m/s, and 14.57 m/s, respectively, due to the narrowing of the valve opening. Thus, the smallest orifice of the valve produced a higher velocity. This finding highlights the importance of hemodynamic assessment in aortic valve stenosis by providing valuable insight for clinicians in pre-surgical evaluation.
Volume: 42
Issue: 3
Page: 884-891
Publish at: 2026-06-10
1 2 3
...
2091 2092
Show 58 of 2092

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