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

Remove glasses diffusion model an innovative conditioned of eye glasses removal with image diffusion model

10.11591/ijeecs.v38.i3.pp1503-1516
Yuliza Yuliza , Rachmat Muwardi , Galatia Erica Yehezkiel , Mirna Yunita , Lenni Lenni
The presence of eyeglasses in facial images poses challenges for image processing, particularly in facial recognition. This paper introduces the remove glasses diffusion model (RGDM), a conditioned denoising diffusion probabilistic model (DDPM) designed for precise glasses removal. RGDM employs conditional modeling to focus on the glasses region while seamlessly restoring facial features. An eyes position accuracy mechanism, leveraging facial landmarks, ensures accurate eye restoration post-removal. Comprehensive evaluations on the CelebA dataset demonstrate RGDM’s superior performance, achieving the lowest Fréchet inception distance (FID) of 27.09 and learned perceptual image patch similarity (LPIPS) of 0.299, outperforming state-of-the-art methods such as 3D synthetic, cycleconsistent generative adversarial network (CycleGAN), and eyeglasses removal generative adversarial network (ERGAN). These results highlight the model’s effectiveness in producing natural and high-fidelity facial reconstructions. This work advances glasses removal technology and underscores the significance of conditional models in image processing. The proposed approach has practical implications for facial recognition and image enhancement, paving the way for more accurate and robust real-world applications.
Volume: 38
Issue: 3
Page: 1503-1516
Publish at: 2025-06-01

Camera-based simultaneous localization and mapping: methods, camera types, and deep learning trends

10.11591/ijra.v14i2.pp162-172
Anak Agung Ngurah Bagus Dwimantara , Oskar Natan , Novelio Putra Indarto , Andi Dharmawan
The development of simultaneous localization and mapping (SLAM) technology is crucial for advancing autonomous systems in robotics and navigation. However, camera-based SLAM systems face significant challenges in accuracy, robustness, and computational efficiency, particularly under conditions of environmental variability, dynamic scenes, and hardware limitations. This paper provides a comprehensive review of camera-based SLAM methodologies, focusing on their different approaches for pose estimation, map reconstruction, and camera type. The application of deep learning also will be discussed on how it is expected to improve performance. The objective of this paper is to advance the understanding of camera-based SLAM systems and to provide a foundation for future innovations in robust, efficient, and adaptable SLAM solutions. Additionally, it offers pertinent references and insights for the design and implementation of next-generation SLAM systems across various applications.
Volume: 14
Issue: 2
Page: 162-172
Publish at: 2025-06-01

Trend analysis of machine learning techniques for traffic control based on bibliometrics

10.11591/ijai.v14.i3.pp2402-2411
Hilda Luthfiyah , Eko Syamsuddin Hasrito , Tri Widodo , Sofwan Hidayat , Okghi Adam Qowiy
Machine learning in traffic control for intelligent transportation systems (ML-ITSTC) aims to enhance user coordination and safety within transportation networks, ultimately improving overall traffic system performance. ML-ITSTC is achieved by leveraging data to execute machine learning algorithms in intelligent transportation management and optimizing traffic flow to prevent or reduce congestion. This paper conducts bibliometric analysis to explain the research status, development trajectory, and challenges of ML-ITSTC, drawing insights from literature in the Scopus database literature covering 2013 to November 2023. The bibliometric analysis of ML-ITSTC includes: performance analysis, science mapping analysis, and citation analysis. The evaluation of ML algorithm trends over the 10-year span indicates that traffic prediction (TP), neural networks, and deep learning are frequently used keywords. Further, an examination of keywords used over the entire period and in 2023 (up to November) shows that reinforcement learning (RL) is the latest popular approach for traffic control in transportation. The results provide a comprehensive view of the opportunities and challenges in ML-ITSTC, covering data, models, and applications, offering researchers insights into the current and future directions of ML-ITSTC research.
Volume: 14
Issue: 3
Page: 2402-2411
Publish at: 2025-06-01

Artificial intelligence-blockchain synergy ensures Indonesia’s compliance with European Union’s Deforestation-free regulation

10.11591/ijai.v14.i3.pp1763-1771
Silfi Iriyani , Sarifah Putri Raflesia
This paper introduces a new model that incorporates blockchain and artificial intelligence (AI) in creating traceability on agricultural supply chains to meet European Union's (EU's) regulation on deforestation-free products. This model stands for the system that would be applied for monitoring origins and routes with regard to verifying the status of products being free from deforestation. Particularly, this addressed the European Union's Deforestation-free Regulation products (EUDR)-related issues in Indonesia focused on smallholders and their linkage to traceability tools. The proposed conceptual model demonstrates how blockchain technology combined with AI in agricultural supply chains enhances transparency and reliability in the line of improving environmental sustainability as well as boosting consumers' confidence. Integration of blockchain and AI increases agricultural supply chain transparency, traceability, and reliability whereby smart contracts can execute automatically such as releasing payments once certain conditions are met.
Volume: 14
Issue: 3
Page: 1763-1771
Publish at: 2025-06-01

A new deep steganographic technique for hiding several secret images in one cover

10.11591/ijai.v14.i3.pp2210-2219
Mohamed Htiti , Aziza El Ouaazizi , Ismail Akharraz , Abdelaziz Ahaitouf
Deep learning has been integrated with image steganography to enhance steganographic security by automatically acquiring the ability to hide information. The issue with current models is that if the cover image is accessible, it is possible to expose the hidden information by simply calculating the differences between the cover image and the steganographic image. This paper introduces a novel image steganography model that utilizes convolutional neural network (CNN) to enhance the dissimulation and extraction capabilities. Specifically, we propose a model that hides two images in a single cover image. Before being hidden within the cover image, a random pixel image is generated and combined with the real secret image. Experimental results show that our proposed method is more effective and relevant.
Volume: 14
Issue: 3
Page: 2210-2219
Publish at: 2025-06-01

Understanding golfers’ acceptance behavior toward robotic golf caddies by merging the task technology fits theory and the perceived risk theory

10.11591/ijra.v14i2.pp173-180
Kyuhyeon Joo , Heather Markham Kim , Jinsoo Hwang
The current paper was designed to understand how to form the acceptance behavior of golfers toward robotic golf caddies, which conducted a hypothetico-deductive approach. The study focused on two questions: i) Can the TTF theory explain the acceptance behavior of golfers toward robotic golf caddies? ii) Do perceived risks negatively affect the acceptance behavior of golfers toward robotic golf caddies? Thus, the study postulated the impacts of task/technology characteristics and the five perceived risks (i.e. financial, time, privacy, performance, and psychological) on task technology fit, and the link between task technology fit and behavioral intentions. The data was collected from 387 golfers in South Korea, and the hypotheses tests were conducted by structural equation modeling. The results of the data analysis indicate that both task and technology characteristics increase task technology fit, and the four dimensions of perceived risks, which include time, privacy, performance, and psychology, have a negative influence on task technology fit. In addition, task technology fit also increases behavioral intentions. The study provides theoretical contributions by filling the acknowledged research gaps, and it also presents managerial implications in regard to commercializing robotic caddies in the golf industry.
Volume: 14
Issue: 2
Page: 173-180
Publish at: 2025-06-01

Experimental evaluation of bidirectional encoder representations from transformers models for de-identification of clinical document images

10.11591/ijra.v14i2.pp273-280
Ravichandra Sriram , Siva Sathya Sundaram , S. LourduMarie Sophie
Many health institutes maintain patients’ diagnosis and treatment reports as scanned images. For healthcare analytics and research, large volumes of digitally stored patient information have to be accessed, but the privacy requirements of protected health information (PHI) limit the research opportunities. Particularly in this artificial intelligence (AI) era, deep learning models require large datasets for training purposes, which hospitals cannot share unless the PHI fields are de-identified. Manual de-identification is beyond possible, with millions of patient records generated in hospitals every day. Hence, this work aims to automate the de-identification of clinical document images utilizing AI models, particularly pre-trained bidirectional encoder representations from transformers (BERT) models. For the purpose of experimentation, a synthetic dataset of 550 clinical document images was generated, encompassing data obtained from diverse patients across multiple hospitals. This work presents a two-stage transfer learning approach, initially employing Tesseract character recognition (OCR) to convert clinical document images into text. Subsequently, it extracts PHI fields from the text for de-identification. For the purpose of extraction, BERT models were utilized; in this work, we contrasted six pre-trained versions of such models to examine their effectiveness and achieve the F1 score of 92.45%, thus showing better potential for de-identifying PHI data in clinical documents.
Volume: 14
Issue: 2
Page: 273-280
Publish at: 2025-06-01

Multi-microcontroller system for Mecanum robots with gripper-shooter mechanisms

10.11591/ijra.v14i2.pp181-190
Mohammed Mareai , Juhen Fashikha Wildan , Mochamad Ridho Afkarean
This study presents the design and implementation of a multi-microcontroller digital control system for a Mecanum-wheeled robot with gripper and shooter mechanisms, tailored for agricultural applications. The proposed system integrates an Espressif32 master controller with Arduino Nano slave microcontrollers, enabling precise control of robot movement and functional components. Wireless control is facilitated by a PlayStation 3 controller, while Mecanum wheels ensure omnidirectional mobility in dynamic environments. Experimental results indicate a 66.67% success rate in seedling planting and an 83.33% success rate in ball collection tasks. Despite its notable performance, enhancements in sensor feedback and automation are recommended to improve efficiency. This research underscores the potential of cost-effective, multi-microcontroller systems for advancing real-time control and task execution in agricultural robotics.
Volume: 14
Issue: 2
Page: 181-190
Publish at: 2025-06-01

CP_SDUNet: road extraction using SDUNet and centerline preserving dice loss

10.11591/ijra.v14i2.pp260-272
Bayu Satria Persada , Muhammad Rifqi Priyo Susanto , Laksmita Rahadianti , Aniati Murni Arymurthy
Existing automatic road map extraction approaches on remote sensing images often fail because they cannot understand the spatial context of an image. Mainly because they could not learn the spatial context of the image and only knew the structure or texture of the image. These approaches only focus on regional accuracy instead of connectivity. Therefore, most approaches produce discontinuous outputs caused by buildings, shadows, and similarity to rivers. This study addresses the challenge of automatic road extraction, focusing on enhancing road connectivity and segmentation accuracy by proposing a network-based road extraction that uses a spatial intensifier module (DULR) and densely connected U-Net architecture (SDUNet) with a connectivity-preserving loss function (CP_clDice) called CP_SDUNet. This study analyzes the CP_clDice loss function for the road extraction task compared to the BCE Loss function to train the SDUNet model. The result shows that CP_SDUNet, performs best using an image size of 128×128 pixels and trained with the whole dataset with a combination of 20% clDice and 80% dice loss. The proposed method obtains a clDice score of 0.85 and an Interest over Union (IoU) score of 0.65 for the testing data. These findings demonstrate the potential of CP_SDUNet for reliable road extraction.
Volume: 14
Issue: 2
Page: 260-272
Publish at: 2025-06-01

Enhanced deep auto encoder technique for brain tumor classification and detection

10.11591/ijeecs.v38.i3.pp2031-2040
Syed Jahangir Badashah , Kavita Moholkar , Sunil L. Bangare , Gaurav Gupta , Devi T. , Sammy Francis , Venkatesan Hariram , Batyrkhan Omarov , Kantilal Pitambar Rane , Abhishek Raghuvanshi
A brain tumor can develop due to uncontrolled proliferation of aberrant cells in brain tissue. Malignant tumor can influence the nearby brain tissues, potentially resulting in the person's death. Early diagnosis of a brain tumor is crucial for ensuring the survival of patients. This article introduces an improved method using a deep auto encoder for the classification and detection of brain tumor. Magnetic resonance imaging (MRI) images are obtained from the BraTS data sets. The images undergo preprocessing using an adaptive Wiener filter. Image preprocessing is essential for eliminating noise from the input MRI pictures, hence enhancing the accuracy of MRI image classification. The fuzzy C-means technique is used to accomplish image segmentation. The classification model comprises deep auto encoder, convolution neural network (CNN), and K-nearest neighbor techniques. The classification model is developed and evaluated using MRI image slices from the BraTS dataset. Accuracy of deep auto encoder is 98.81%. Accuracy of CNN is 95.50 and accuracy of K-nearest neighbor (KNN) technique is 91.30%.
Volume: 38
Issue: 3
Page: 2031-2040
Publish at: 2025-06-01

Robust two-stage object detection using YOLOv5 for enhancing tomato leaf disease detection

10.11591/ijai.v14.i3.pp2246-2257
Endang Suryawati , Syifa Auliyah Hasanah , Raden Sandra Yuwana , Jimmy Abdel Kadar , Hilman Ferdinandus Pardede
Deep learning facilitates human activities across various sectors, including agriculture. Early disease detection in plants, such as tomato plant that are susceptible to diseases, is critical because it helps farmers reduce losses and control the disease spread more effectively. However, the ability of machine to recognize diseased leaf objects is also influenced by the quality of data. Data collected directly from the field typically yields lower accuracy due to challenges faced in machine interpretation. To address this challenge, we propose a two-stage detection architecture for identifying infected tomato plant classes, leveraging YOLOv5 to detect objects within the images obtained from the field. We use Inception-V3 for classifying objects into known classes. Additionally, we employ a combination of two dataset: PlantDocs which represent field data, and PlantVillage dataset which serves as a cleaner dataset. Our experimental results indicate that the use of YOLOv5 in handling data under actual field conditions can enhance model performance, although the accuracy value is moderate (62.50 %).
Volume: 14
Issue: 3
Page: 2246-2257
Publish at: 2025-06-01

Identification of potential depression in social media posts

10.11591/ijai.v14.i3.pp2096-2103
Munawar Munawar , Yulhendri Yulhendri
The widespread use of social media to convey emotions (including depression) can be used to identify suspected depression in social media posts by examining the language that they have used on social media. This study aims to develop a system for detecting suspected depression in social media posts using sentiment analysis. This study collected data from X (Twitter) for three months using the keywords depression, mental health, and mental disorders. 1,502 data were generated due to the cleaning process of the 5,000 data collected. The findings of employing the validated by psychologist valence aware dictionary and sentiment reasoner (VADER) and Indonesian sentiment (InSet) lexicons demonstrate that VADER is more accurate (95.1%) than Inset (76.9%). The results of modeling with random forest, naive Bayes, and support vector machine (SVM) showed that random forest had the highest accuracy (83.3%), followed by naive Bayes (80.5%) and SVM (80.4%). Predicting social media data using lexicons and machine learning has limits that can be addressed by validation from clinical psychology. The frequency, timing, and idiom of posts on social media can reveal signs of depression. Depression seems to be best described by words like melancholy, stress, sadness, worthlessness, and depression.
Volume: 14
Issue: 3
Page: 2096-2103
Publish at: 2025-06-01

Lung cancer patients survival prediction using outlier detection and optimized XGBoost

10.11591/ijai.v14.i3.pp2146-2157
Wirot Yotsawat , Peetiphart Suebpeng , Saroch Purisangkaha , Akarapon Poonsawad , Kanyalag Phodong
This research aims to improve the prediction’s model for survival time of lung cancer patients by using outlier detection, hyper-parameter optimization, and machine learning technique. The research compares the performance of several methods including multilayer perceptron (MLP), decision tree (DT), linear regression (LR), Bagging, XGBoost, and random forest (RF). The dataset used for the experiment is obtained from the surveillance, epidemiology, and end result (SEER) cancer database, which contains diagnoses data from 2004 to 2015. The total number of records used is 196,031 with 22 features. 10-fold cross-validation is used for training and testing sets. The evaluation metrics are root mean square error (RMSE), mean squared error (MSE), R-squared (R2), and mean absolute error (MAE). The results show that the lung cancer patient survival prediction model using the optimized XGBoost (O-XGBoost) model performs the best with an RMSE of 13.74 and outperforms the baseline-XGBoost model as well as other models. This research will be useful for developing a clinical decision support system for the care of lung cancer patients. Physicians can use the developed model to assess the patient’s chance of survival in order to plan more effective treatment.
Volume: 14
Issue: 3
Page: 2146-2157
Publish at: 2025-06-01

IoT-based cricket environment system to maximize egg production and reduce mortality rate

10.11591/ijra.v14i2.pp281-289
Dominic Miracle Tjandrata , Suryadiputra Liawatimena
The deployment of Internet of things (IoT) technologies presents an opportunity to improve efficiency in cricket farming. This study investigates the implementation of an IoT-based system utilizing an ESP32 microcontroller, a suite of environmental sensors, and actuators. The system is supported by a ThingsBoard dashboard for data visualization and a Telegram bot for notifications. The setup was tested on a single cricket cage over a 28-day period and compared against a control group. Each cage contained 20 male and 100 female Cliring crickets. Key parameters analyzed included temperature, humidity, soil moisture, egg yield, food conversion ratio (FCR), and mortality rate. Findings show that the IoT-enabled cage consistently maintained optimal environmental conditions—temperature (20 to 32 °C), humidity (65% to 85%), and soil moisture (60% to 80%)—unlike the control, which experienced greater variability. The IoT cage yielded 87.28 grams of eggs, a 33.33% improvement over the control's 65.46 grams. Additionally, FCR improved from 2.53 to 2.01 grams per egg, and mortality rate dropped from 0.816 to 0.708. These results underscore the effectiveness of IoT systems in enhancing environmental stability, productivity, and survival rates in small- to medium-scale cricket farming operations.
Volume: 14
Issue: 2
Page: 281-289
Publish at: 2025-06-01

Supply chain efficiency transformation: analysis of raw material staff selection based on preference selection index

10.11591/ijai.v14.i3.pp2459-2470
Amrullah Amrullah , Akbar Idaman , Al-Khowarizmi Al-Khowarizmi
In the era of intense business globalization, supply chain management is becoming a vital key to improving the efficiency and competitiveness of enterprises. The selection of raw material supply staff is an important aspect of supply chain management, affecting smooth supply, efficiency and cost control. This research focuses on using the preference selection index (PSI) method in the selection of raw material supply staff. PSI is a tool that integrates data from multiple criteria in the selection process. The results show that PSI provides an effective evaluation in staff selection, identifies key variables that affect selection success and analyzes the impact of using PSI on supply chain efficiency and company productivity. This research fills the knowledge gap in the application of PSI in the context of raw material supply staff selection and contributes to the understanding of efficient and sustainable supply chain management. The results provide valuable insights for industries and organizations that depend on reliable raw material supply and demonstrate the potential to improve the overall staff selection process. The outcome of this study found that Muliyono received a PSI score of 0.9643 and was ranked first, while Ramli received a PSI score of 0.9548 and was ranked second.
Volume: 14
Issue: 3
Page: 2459-2470
Publish at: 2025-06-01
1 2 3
...
2097 2098
Show 295 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