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

31,456 Article Results

Driver inattention detection system using multi-task cascaded convolutional networks

10.11591/ijai.v13.i4.pp4249-4262
Abdelfettah Soultana , Faouzia Benabbou , Nawal Sael , Soukaina Bouhsissin
Driver inattention has emerged as a critical concern impacting road safety, resulting in an alarming surge in accidents and fatalities. This research introduces a novel system for detecting inattention, structured across six levels: perception, facial feature extraction, tracking driver face, and driver secondary task using pre-trained deep learning models, inattention detection, risk estimation, and alert. The system is based on image processing captured from two strategically positioned cameras that simultaneously capture the driver’s activities while driving and their facial expressions. The second contribution concerns the driver facial features extraction using multi-task cascaded convolutional networks (MTCNN), and it is comparison with the histogram of gradient (HOG)-based frontal face detector, and haar feature based cascade classifier. The algorithms were compared based on their runtime efficiency, robustness in handling varying lighting conditions, and various head movements. The MTCNN achieves high performance, reaching accuracy levels ranging from 96.4% to 99.5% on two datasets including realistic driving scenarios: the DrivFace dataset and, the driver drowsiness dataset. The comparative analysis sheds light on the strengths and weaknesses of each algorithm, providing valuable insights for selecting the most suitable face detection algorithm to use in our system.
Volume: 13
Issue: 4
Page: 4249-4262
Publish at: 2024-12-01

Internet of things important roles in hybrid photovoltaic and energy storage system: a review

10.11591/ijece.v14i6.pp6182-6194
Habiba Ahmed , Eva-Denisa Barbulescu , Mohamad Nassereddine , Obada Al Khatib
Renewable energy systems have become integral components of the electrical grid, offering environmental benefits and cost-effective power generation. Technological advancements have introduced internet of things (IoT) devices with robust data collection and execution capabilities. Solar photovoltaic systems, reliant on unpredictable solar radiation, require hybrid systems incorporating various renewable energy sources and energy storage to ensure system stability. Successful operation necessitates data gathering through IoT devices, which have played a crucial role in enhancing system sustainability. This paper provides a comprehensive review of the role of IoT in photovoltaic systems and energy storage, highlighting its significant contributions to system efficiency, fault detection, output prediction, system stability, and load management. The study underscores the critical dependence of advancements in the renewable energy sector on IoT systems.
Volume: 14
Issue: 6
Page: 6182-6194
Publish at: 2024-12-01

Stacking-based ensemble learning for identifying artist signatures on paintings

10.11591/ijeecs.v36.i3.pp1683-1693
Shintami Chusnul Hidayati , Agustinus Aldi Irawan Rahardja , Nanik Suciati
Identifying artist signatures on paintings is essential for authenticating artworks and advancing digital humanities. An artist’s signature is a consistent element included in each painting that the artist creates, providing a unique identifier for their work. Traditional methods that rely on expert analysis and manual comparison are time-consuming and are prone to human error. Although convolutional neural networks (CNNs) have shown promise in automating this process, existing single-model approaches struggle with the diversity and complexity of artistic styles, leading to limitations in their performance and generalizability. Therefore, this study proposes an ensemble learning approach that integrates the predictive power of multiple CNN-based models. The proposed framework leverages the strengths of three state-of-the-art CNNs: EfficientNetB4, ResNet-50, and Xception. These models were independently trained, and the predictions were combined using a meta-learning strategy. To address class imbalance, data augmentation techniques and weighted loss functions were employed. The experimental results obtained on a dataset of more than 8,000 paintings from 50 artists demonstrate significant improvements over individual CNN architectures and other ensemble methods, thereby effectively capturing complex features and improving generalizability.
Volume: 36
Issue: 3
Page: 1683-1693
Publish at: 2024-12-01

Enhancing text classification through novel deep learning sequential attention fusion architecture

10.11591/ijai.v13.i4.pp4642-4653
Shilpa Shilpa , Shridevi Soma
Text classification is a pivotal task within natural language processing (NLP), aimed at assigning semantic labels to text sequences. Traditional methods of text representation often fall short in capturing intricacies in contextual information, relying heavily on manual feature extraction. To overcome these limitations, this research work presents the sequential attention fusion architecture (SAFA) to enhance the features extraction. SAFA combines deep long sort-term memory (LSTM) and multi-head attention mechanism (MHAM). This model efficiently preserves data, even for longer phrases, while enhancing local attribute understanding. Additionally, we introduce a unique attention mechanism that optimizes data preservation, a crucial element in text classification. The paper also outlines a comprehensive framework, incorporating convolutional layers and pooling techniques, designed to improve feature representation and enhance classification accuracy. The model's effectiveness is demonstrated through 2-dimensional convolution processes and advanced pooling, significantly improving prediction accuracy. This research not only contributes to the development of more accurate text classification models but also underscores the growing importance of NLP techniques.
Volume: 13
Issue: 4
Page: 4642-4653
Publish at: 2024-12-01

Risk factors and trends analysis of unwanted pregnancy in Indonesia

10.11591/ijphs.v13i4.23702
Supriyadi supriyadi , Tris Eryando , Purnawan Junadi
Unwanted pregnancy has serious and adverse health, social, and economic impacts, as well as the risk of maternal and child mortality. In Indonesia, unwanted pregnancy is one of the most troubling public health problems and a major reproductive health problem. The sample consisted of 2002 (12,612), 2007 (14,471), 2012 (14,212), and 2017 (8,838), so the total sample was 50,233 couples of childbearing age by random sample method. This study was analyzed using logistic regression analysis. Unintended pregnancies were associated with maternal age, age at first marriage, number of children, maternal education, contraceptive use, contraceptive failure, residence, and the interval between pregnancies. Unwanted pregnancies at the mother's vulnerable age have increased over the past 10 years and decreased after entering the last 5 years, tend to be stable and constant at the age of first marriage. About one and a half women in Indonesia encounter unwanted pregnancies. Unwanted pregnancies fluctuate over 20 years. Women, especially the most vulnerable should be empowered to avoid unwanted pregnancies for their awareness and independence. Programs to expand access and quality of information and counseling centers, access and quality of women's health services, and specific reproductive health services as needed.
Volume: 13
Issue: 4
Page: 1864-1871
Publish at: 2024-12-01

Image analysis for classifying coffee bean quality using a multi-feature and machine learning approach

10.11591/ijai.v13.i4.pp4241-4248
Anindita Septiarini , Hamdani Hamdani , Aji Ery Burhandeny , Damar Nurcahyono , Surya Eka Priyatna
Price and customer satisfaction depend on coffee bean quality. The coffee industry must analyze coffee bean quality. Global demand for robusta coffee is high. Coffee industry professionals mostly understand coffee bean quality. Thus, an image analysis using a computer vision-based approach for classifying robusta coffee bean quality is required. Image acquisition, region of interest (ROI) detection, pre-processing, segmentation, feature extraction, feature selection, and classification are covered in this study. A multi-feature derived based on color, shape, and texture features was employed in feature extraction, followed by feature selection using principal component analysis (PCA). Several machine-learning methods classified the coffee beans. The method performance was assessed using precision, recall, and accuracy. The selected features using the backpropagation neural network (BPNN) classifier outperformed others with 98.54% accuracy.
Volume: 13
Issue: 4
Page: 4241-4248
Publish at: 2024-12-01

Android malware detection using the random forest algorithm

10.11591/ijeecs.v36.i3.pp1876-1883
Anas El Attaoui , Norelislam El Hami , Younes Koulou
The rapid growth in Android device usage has resulted in a significant increase in malware targeting this platform, posing serious threats to user security and privacy. This research tackles the challenge of Android malware detection by leveraging advanced machine learning techniques, with a particular emphasis on the random forest (RF) algorithm. Our primary objective is to accurately identify and classify malicious applications to enhance the security of Android devices. In this study, we employed the RF algorithm to analyze a comprehensive dataset of Android applications, where the classification of each application as either malware or benign is known. The method was rigorously tested, yielding impressive results: an average accuracy of 98.47%, a sensitivity of 98.60%, and an F-score of 98.60%. These metrics underscore the effectiveness of our approach. Moreover, we conducted a comparative analysis of the RF algorithm against other malware detection methods. The results demonstrate that the RF algorithm outperforms these alternative methods, offering superior detection capabilities and contributing to more robust Android security measures.
Volume: 36
Issue: 3
Page: 1876-1883
Publish at: 2024-12-01

Optical coherence tomography angiography image classification and analysis of diabetic retinopathy, using Wasserstein generative adversarial network augmentation

10.11591/ijece.v14i6.pp7046-7056
Pranali Pradeep Hatode , Maniroja Edinburgh
Deep learning algorithms effectively work with, a significant amount of data. trained on small datasets tend to have poor generalization. Data augmentation techniques can be used to make better use of existing training data, improving the applicability of deep learning methods. However, traditional data augmentation methods often produce limited additional credible data. The deep learning approach's performance can be enhanced by generating new data by employing generative adversarial networks (GANs). Although GANs have been extensively used to improve the performance of convolutional neural networks (CNNs), there has been relatively less research on data augmentation methods specifically for GAN training. This study focuses on using a Wasserstein GAN (WGAN) architecture for generating synthetic optical coherence tomography angiography (OCTA) images of diabetic retinopathy to aid in the detection of different types of diabetic retinopathy diseases, including proliferative diabetic retinopathy (PDR), Severe non-PDR (NPDR), Moderate NPDR, and Mild NPDR. WGAN, provides the generator with a more informative learning signal, making training more stable, particularly in high-dimensional spaces. The trained WGAN model is saved in .h5 file format (HDF), converted to portable network graphics (PNG) image format, and then classified into different categories of diabetic retinopathy using a ResNet50 model with various fine-tuning methods. The proposed model has demonstrated better results than the previous study. 99.95% accuracy is exhibited.
Volume: 14
Issue: 6
Page: 7046-7056
Publish at: 2024-12-01

Application of K-means clustering and B-value algorithms for analysis of earthquake-dangerous zones in Java Island

10.11591/ijaas.v13.i4.pp907-915
Bentang Anggarajati , Y Yatini , Wiji Raharjo
Java Island is an island with a high earthquake vulnerability. Therefore, earthquake mitigation measures are needed to reduce the impact of earthquakes. Earthquake mitigation is done by knowing the zones with a high risk of earthquakes and high levels of rock stress. The methods used to map earthquake-prone zones are K-means clustering and B-value. The K-means clustering method can provide earthquake clusters based on their characteristics and the B-value can produce rock stress conditions in the area. The results of this study are that the K-means clustering method produces 7 earthquake clusters with 5 classifications of very low, low, medium, high, and very high. In contrast, the B-value process has a high B-value with a value of 1.2-1.5 in West Java and a low B-value with a value of 0.9-1.2 in the central to the eastern part of Java.
Volume: 13
Issue: 4
Page: 907-915
Publish at: 2024-12-01

Enhancing road image clarity with residual neural network dehazing model

10.11591/ijai.v13.i4.pp4147-4157
Aerun Martin , Mohd Nazeri Kamaruddin , Zamani Md Sani , Hadhrami Abdul Ghani
Lane markers, or road markers, are the painted lines on a roadway that separate different lanes of traffic. Lane markers guide drivers and ensure orderly vehicle flow. They are essential for advanced driver assistance systems (ADAS), providing reference points for vehicle positioning on the road. These markers enable ADAS to give warnings, assistance, and automation features that enhance driver safety and convenience. However, unpredictable illumination, such as a foggy environment, can suppress marker visibility, impacting ADAS's performance. Deep learning-based methods are well-known for their superiority in handling various haze patterns. This paper presents a residual network (ResNet)-based deep learning model to improve road image clarity impacted by fog. The residual neural network dehaze model (RNN-D) utilises a joint loss function to produce haze-free images with improved lighting conditions and enhanced details. The model was trained, validated, and fine-tuned using hazy and corresponding non-hazy datasets to ensure that the model is quantitatively superior in the peak signal-to-noise ratio (PSNR) and the structural similarity index measure (SSIM). RNN-D achieved an average PSNR of 27.98 and SSIM of 0.8 on multiple open sourced datasets. The proposed algorithm's superior performance and visually appealing results make it a powerful tool for real-world image dehazing applications.
Volume: 13
Issue: 4
Page: 4147-4157
Publish at: 2024-12-01

A comparative analysis of constant impedance and constant power loads in a distribution network

10.11591/ijece.v14i6.pp6111-6121
Ramya Ramya , Rex Joseph
Most conventional power systems adopt radial distribution network wherein multiple loads are connected across the distribution transformer. As the number of loads increases, it results in poor voltage profile at the distant receiving end reducing power delivery. This issue worsens with the large-scale influx of electric vehicles and power converter-fed loads, which draw constant power irrespective of supply voltage. Such loads exhibit negative incremental resistance behavior and also have a dynamic response which affects the network in a manner different from constant impedance loads. This paper compares the effects of constant power and constant impedance loads by modeling adjustable converter dynamics for constant power loads. It analyzes line currents, load voltages and power transmitted in a four-load radial test system with optional distributed sources. Results show poorer voltage profile and the effect of power converter dynamics in constant power loads compared to conventional loads. Adding distributed sources improves voltage profile considerably, and transmission losses are reduced. Steady state analysis is then extended to an IEEE 31-bus 23 kV distribution test system with similar results. Transmission losses are computed along different branches, and the influence of loads and sources are analyzed. The outcomes of the analysis can be used in arrival of loss allocation in a system where peer to peer energy sharing is envisaged.
Volume: 14
Issue: 6
Page: 6111-6121
Publish at: 2024-12-01

Structural equation modeling: exploring relationships of body mass index, waist circumference, and hypertension in elderly

10.11591/ijphs.v13i4.24415
Wahyu Tri Sudaryanto , Isnaini Herawati , Wahyuni Wahyuni , Ika Yuli Ayuningrum , Bhisma Murti , Rizki Setiawan , Nuristiqomah Dwi Putri
Waist circumference (WC) measurements in the elderly are uncommon and erratic. Furthermore, its correlation to hypertension has received less attention in Indonesia. This study attempted to investigate direct and indirect risk factors for high blood pressure in the elderly using structural equation modeling (SEM). A cross-sectional study was conducted in Surakarta, Central Java, Indonesia. The 297 elderly aged ≥60 is eager to take part. The path model of factors associated with blood pressure was analyzed using Stata 13. Male (b=-0.43; 95% CI=-5.67 to 4.81; p=0.872), age (b=0.52; 95% CI=0.12 to 0.93; p=0.012), and WC (b=0.35; 95% CI=0.17 to 0.53; p<0.001) directly affected systolic blood pressure (SBP). Male (b=1.83; 95% CI=-1.17 to 4.82; p=0.231), age (b=-0.03; 95% CI=-0.26 to 0.20; p=0.790), BMI (b=0.40; 95% CI=-0.03 to 0.83; p=0.067), and WC (b=0.04; 95% CI=-0.09 to 0.16; p=0.571) directly affected diastolic blood pressure (DBP). Increasing age, body mass index (BMI), and waist circumference (WC) are the important variables that influence blood pressure in older people. This study supports the evidence that body composition and weight control are necessary to prevent and control blood pressure in the elderly.
Volume: 13
Issue: 4
Page: 1599-1608
Publish at: 2024-12-01

Electrical circuit approaches to modeling brain chaos: insights into neural dynamics

10.11591/ijeecs.v36.i3.pp1503-1511
Kaouther Selmi , Kods Bachta , Kais Bouallegue
This paper investigates the simulation of brain chaos dynamics using a combination of the Chua circuit and diode tunnel mechanisms, aiming to examine chaotic behavior in brain networks. Leveraging the inherent chaotic properties of the Chua circuit, Fitzhugh-Nagumo (FHN) function and the nonlinear characteristics of diode tunneling, our model offers a platform to mimic the intricate synaptic interactions observed in the brain. By subjecting the model to various stimuli and perturbations, we analyze the emergence and evolution of chaotic patterns, shedding light on the underlying mechanisms of cerebral chaos. Through numerical simulations and experimental validation, we demonstrate the effectiveness of our approach in replicating key features of brain chaos and highlight its potential implications for understanding neurological disorders and cognitive processes. This research contributes to the broader effort of leveraging computational models to explore the complex dynamics of the brain and their implications for neuroscience and microengineering.
Volume: 36
Issue: 3
Page: 1503-1511
Publish at: 2024-12-01

Comparison of algorithms for the detection of marine vessels with machine vision

10.11591/ijece.v14i6.pp6332-6338
José Rodríguez-Gonzales , Junior Niquin-Jaimes , Ernesto Paiva-Peredo
The detection of marine vessels for revenue control has many tracking deficiencies, which has resulted in losses of logistical resources, time, and money. However, digital cameras are not fully exploited since they capture images to recognize the vessels and give immediate notice to the control center. The analyzed images go through an incredibly detailed process, which, thanks to neural training, allows us to recognize vessels without false positives. To do this, we must understand the behavior of object detection; we must know critical issues such as neural training, image digitization, types of filters, and machine learning, among others. We present results by comparing two development environments with their corresponding algorithms, making the recognition of ships immediately under neural training. In conclusion, it is analyzed based on 100 images to measure the boat detection capability between both algorithms, the response time, and the effectiveness of an image obtained by a digital camera. The result obtained by YOLOv7 was 100% effective under the application of processing techniques based on neural networks in convolutional neural network (CNN) regions compared to MATLAB, which applies processing metrics based on morphological images, obtaining low results.
Volume: 14
Issue: 6
Page: 6332-6338
Publish at: 2024-12-01

Hybrid optimization algorithm for resource-efficient and data-driven performance in agricultural IoT

10.12928/telkomnika.v23i1.26361
Depa; Universiti Teknikal Malaysia Melaka (UTeM) Ramachandraiah Kumar Raja , Zuraida; Universiti Teknikal Malaysia Melaka (UTeM) Abal Abas , Goshtu; Dayananda Sagar University Hemanth Kumar , Chakana; Mohan Babu University Ravindra Murthy , Venappagari; REVA University Eswari
The agricultural sector is undergoing a significant transformation with the adoption of the agricultural internet of things (IoT), yet it faces persistent challenges in optimizing resource efficiency and data-driven performance due to limitations in current optimization algorithms. This research assesses the effectiveness of four prominent algorithms such as ant colony optimization (ACO), genetic algorithms (GA), particle swarm optimization (PSO), and artificial bee colony (ABC) in addressing these challenges within agricultural IoT (AIoT). Introducing a novel hybrid optimization algorithm (HOA), we aim to overcome these limitations by prioritizing both resource efficiency and data-driven performance. Through a thorough evaluation, HOA demonstrates its superiority in enhancing both aspects, thereby establishing itself as a compelling solution for AIoT applications. The introduction of HOA sets the stage for sustainable, cost-effective, and data-driven precision agriculture, significantly enhancing resource efficiency and data accuracy within the IoT network.
Volume: 23
Issue: 1
Page: 201-210
Publish at: 2024-11-26
1 2 3
...
2097 2098
Show 424 of 2098

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