• Journal of Internet of Things and Convergence
• /
• v.10 no.5
• /
• pp.173-178
• /
• 2024

This paper proposes the design of a personalized learning support system for students with learning disabilities, utilizing biometric signals. The system leverages EEG (electroencephalography) and eye-tracking data to monitor the learner's state in real-time, identifying signs of decreased concentration, boredom, or diminished interest. By providing customized feedback and an adaptive learning environment, the system aims to enhance the learning experience and effectiveness. Key components of the system include data collection using Emotiv Epoc X and eye-tracking devices, data preprocessing, and the application of AI models such as Convolutional Neural Networks (CNNs) and Long Short-Term Memory (LSTM) networks. Additionally, Random Forest and Gradient Boosting techniques are employed to predict learner characteristics and optimize feedback, while Decision Trees are used to analyze learning outcomes and deliver individualized recommendations. The proposed system aims to provide an optimal learning environment for students with learning disabilities, with the ultimate goal of improving educational performance and motivation.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.19 no.6
• /
• pp.1271-1280
• /
• 2024

JPEG AI refers to an international standard for learning-based image coding, leveraging deep learning techniques that have made groundbreaking advancements in compression performance. It addresses the rapid increase in the generation and utilization of image data, and is one of the latest standardization efforts in this field. JPEG AI aims to meet the requirements of a wide range of applications, including cloud systems, video surveillance, autonomous vehicles, image data monitoring, and media distribution. To achieve this, it reduces the bandwidth and storage space requirements by up to 50% for the same visual quality and provides a framework that allows the compressed bitstream to be directly used for computer vision and image processing tasks. This paper discusses the JPEG AI, explaining its goals, the selection of training datasets, the standardization process, a performance comparison of key proposals, and the future standardization schedule, in order to understand the characteristics of this new international standard.

• Journal of the Semiconductor & Display Technology
• /
• v.23 no.4
• /
• pp.89-94
• /
• 2024

Optical Emission Spectroscopy (OES) is widely used for real-time diagnostics in semiconductor processes. Recently, virtual metrology (VM) techniques utilizing OES data have gained attention to reduce wafer measurements and improve productivity. However, the selection of key wavelengths that influence process outcomes is still based on expert experience or key wavelengths used in end point detection (EPD), limiting automation and objectivity. In this study, CF₄/O₂ plasma with added N₂ was used to etch SiO₂ and Si₃N₄ simultaneously, and OES data were analyzed using machine learning algorithms to identify key wavelengths closely related to the etch rate (ER) and selectivity of SiO₂ and Si₃N₄. Unlike traditional experience-based methods, this data-driven approach enabled the identification of previously overlooked wavelengths, demonstrating their significant impact on process outcomes in real-time monitoring.

• Journal of the Korean Geotechnical Society
• /
• v.40 no.6
• /
• pp.161-169
• /
• 2024

In this study, a low-cost, Arduino-based inclination measurement system was developed to reduce the acquisition and maintenance costs associated with conventional wall inclination monitoring devices. The developed system utilizes an embedded inertial measurement unit (IMU) sensor for data acquisition and employs various filtering techniques, including the moving average, exponentially weighted moving average, complementary, and Kalman filters, to minimize noise and error. Achieving the required precision with a single filter proved challenging; however, combining multiple filtering methods considerably enhanced the accuracy. The validation results demonstrated that the selected Arduino board, embedded IMU sensor, and integrated filtering system can accurately measure wall inclination or horizontal displacement. This study demonstrates the feasibility of implementing low-cost, open-source hardware-based measurement systems in practical field applications.

• Journal of The Korean Astronomical Society
• /
• v.52 no.1
• /
• pp.11-21
• /
• 2019

Intensive Monitoring Survey of Nearby Galaxies (IMSNG) is a high cadence observation program monitoring nearby galaxies with high probabilities of hosting supernovae (SNe). IMSNG aims to constrain the SN explosion mechanism by inferring sizes of SN progenitor systems through the detection of the shock-heated emission that lasts less than a few days after the SN explosion. To catch the signal, IMSNG utilizes a network of 0.5-m to 1-m class telescopes around the world and monitors the images of 60 nearby galaxies at distances D < 50 Mpc to a cadence as short as a few hours. The target galaxies are bright in near-ultraviolet (NUV) with $M_{NUV}$ < -18.4 AB mag and have high probabilities of hosting SNe ($0.06SN\;yr^{-1}$ per galaxy). With this strategy, we expect to detect the early light curves of 3.4 SNe per year to a depth of R ~ 19.5 mag, enabling us to detect the shock-heated emission from a progenitor star with a radius as small as $0.1R_{\odot}$. The accumulated data will be also useful for studying faint features around the target galaxies and other science projects. So far, 18 SNe have occurred in our target fields (16 in IMSNG galaxies) over 5 years, confirming our SN rate estimate of $0.06SN\;yr^{-1}$ per galaxy.

• The Sea:JOURNAL OF THE KOREAN SOCIETY OF OCEANOGRAPHY
• /
• v.16 no.3
• /
• pp.147-154
• /
• 2011

Decision making in Environmental Impact Assessment (EIA) and Consultation on the Coastal Area Utilization (CCAU) is footing on the survey reports, thus requires concrete and accurate information on the natural habitats. In spite of the importance of reporting the ecological quality and status of habitats, the accumulated knowledge and recent techniques in ecology such as the use of investigated cases and indicators/indices have not been utilized in evaluation processes. Even the EIA report does not contain sufficient information required in a decision making process for conservation and development. In addition, for CCAU, sampling efforts were so limited that only two or a few stations were set in most study cases. This hampers transferring key ecological information to both specialist review and decision making processes. Hence, setting the effective number of sampling stations can be said as a prior step for better assessment. We introduced a few statistical techniques to determine the number of sampling stations in macrobenthos surveys. However, the application of the techniques requires a preliminary study that cannot be performed under the current assessment frame. An analysis of the spatial configuration of sampling stations from 19 previous studies was carried out as an alternative approach, based on the assumption that those configurations reported in scientific journal contribute to successful understanding of the ecological phenomena. The distance between stations and number of sampling stations in a $4{\times}4$ km unit area were calculated, and the medians of each parameter were 2.3 km, and 3, respectively. For each study, approximated survey area (ASA, $km^2$) was obtained by using the number of sampling stations in a unit area (NSSU) and total number of sampling stations (TNSS). To predict either appropriate ASA or NSSU/TNSS, we found and suggested statistically significant functional relationship among ASA, survey purpose and NSSU. This empirical approach will contribute to increasing sampling effort in a field survey and communicating with reasonable data and information in EIA and CCAU.

• Archives of Hand and Microsurgery
• /
• v.23 no.4
• /
• pp.230-238
• /
• 2018

Purpose: Metacarpal fractures are common injuries of the hand. They are treated using closed reduction (CR) or open reduction (OR) techniques. The management strategy depends on fracture site characteristic and fixation methods. In this study, we evaluated pre- and postoperative fracture angulation, when metacarpal fractures bad been treated using two different techniques: CR with parallel transverse pinning and OR with plate fixation. Methods: Forty-six patients undergoing anatomic reduction to treat extra-articular metacarpal fractures were recruited. They were included in one of two therapeutic groups: Group 1, CR with parallel transverse pinning (n=21); Group 2, OR with plate fixation (n=25). Fracture angulation values have been measured on pre- and postoperative radiologic images. Values were compared between pre- and postoperative states, and between corresponding measurements of each group. Results: All extra-articular metacarpal fractures were successfully treated without wound related complications or the limit of joint motion. Both groups demonstrated adequate reduction at immediate postoperative period (postoperative angulation of group 1, $20^{\circ}{\pm}7^{\circ}$; group 2, $19^{\circ}{\pm}5^{\circ}$). During the observation at follow-up period, Group 1 exhibited slight recurrence (follow-up angulation of group 1, $24^{\circ}{\pm}10^{\circ}$). Nonetheless, Group 2 showed adequate reduction state in both immediate postoperative and long-term follow-up periods (follow-up angulation of group 2, $18^{\circ}{\pm}6^{\circ}$). Conclusion: Extra-articular metacarpal fractures were successfully restored without functional complications. CR with parallel transverse pinning method exhibited recurrence after pin removal, which necessitates cautious postoperative exercise and monitoring.

• Journal of Intelligence and Information Systems
• /
• v.27 no.3
• /
• pp.57-73
• /
• 2021

Maintenance and prevention of failure through anomaly detection of ICT infrastructure is becoming important. System monitoring data is multidimensional time series data. When we deal with multidimensional time series data, we have difficulty in considering both characteristics of multidimensional data and characteristics of time series data. When dealing with multidimensional data, correlation between variables should be considered. Existing methods such as probability and linear base, distance base, etc. are degraded due to limitations called the curse of dimensions. In addition, time series data is preprocessed by applying sliding window technique and time series decomposition for self-correlation analysis. These techniques are the cause of increasing the dimension of data, so it is necessary to supplement them. The anomaly detection field is an old research field, and statistical methods and regression analysis were used in the early days. Currently, there are active studies to apply machine learning and artificial neural network technology to this field. Statistically based methods are difficult to apply when data is non-homogeneous, and do not detect local outliers well. The regression analysis method compares the predictive value and the actual value after learning the regression formula based on the parametric statistics and it detects abnormality. Anomaly detection using regression analysis has the disadvantage that the performance is lowered when the model is not solid and the noise or outliers of the data are included. There is a restriction that learning data with noise or outliers should be used. The autoencoder using artificial neural networks is learned to output as similar as possible to input data. It has many advantages compared to existing probability and linear model, cluster analysis, and map learning. It can be applied to data that does not satisfy probability distribution or linear assumption. In addition, it is possible to learn non-mapping without label data for teaching. However, there is a limitation of local outlier identification of multidimensional data in anomaly detection, and there is a problem that the dimension of data is greatly increased due to the characteristics of time series data. In this study, we propose a CMAE (Conditional Multimodal Autoencoder) that enhances the performance of anomaly detection by considering local outliers and time series characteristics. First, we applied Multimodal Autoencoder (MAE) to improve the limitations of local outlier identification of multidimensional data. Multimodals are commonly used to learn different types of inputs, such as voice and image. The different modal shares the bottleneck effect of Autoencoder and it learns correlation. In addition, CAE (Conditional Autoencoder) was used to learn the characteristics of time series data effectively without increasing the dimension of data. In general, conditional input mainly uses category variables, but in this study, time was used as a condition to learn periodicity. The CMAE model proposed in this paper was verified by comparing with the Unimodal Autoencoder (UAE) and Multi-modal Autoencoder (MAE). The restoration performance of Autoencoder for 41 variables was confirmed in the proposed model and the comparison model. The restoration performance is different by variables, and the restoration is normally well operated because the loss value is small for Memory, Disk, and Network modals in all three Autoencoder models. The process modal did not show a significant difference in all three models, and the CPU modal showed excellent performance in CMAE. ROC curve was prepared for the evaluation of anomaly detection performance in the proposed model and the comparison model, and AUC, accuracy, precision, recall, and F1-score were compared. In all indicators, the performance was shown in the order of CMAE, MAE, and AE. Especially, the reproduction rate was 0.9828 for CMAE, which can be confirmed to detect almost most of the abnormalities. The accuracy of the model was also improved and 87.12%, and the F1-score was 0.8883, which is considered to be suitable for anomaly detection. In practical aspect, the proposed model has an additional advantage in addition to performance improvement. The use of techniques such as time series decomposition and sliding windows has the disadvantage of managing unnecessary procedures; and their dimensional increase can cause a decrease in the computational speed in inference.The proposed model has characteristics that are easy to apply to practical tasks such as inference speed and model management.

• Journal of Trauma and Injury
• /
• v.18 no.2
• /
• pp.161-171
• /
• 2005

Background: The management of hepatic injuries has changed dramatically during the past two decade after the technologic breakthroughs in radiologic imaging techniques. Recently, the non-operative management of blunt hepatic trauma has become the standard of care in hemodynamically stable patients. We reviewed our experience of the non-operative management of blunt hepatic trauma. And the purpose of this study was to examine the prognostic factors and indicators affecting the decision for treatment modality of emergent hepatic trauma. Methods: The medical records of 84 patients who were treated for blunt hepatic injury at Masan Samsung Hospital from January 2002 to December 2003. The patients were divided two groups, non-operative(Non-OP) and operative(OP), according to the treatment modality. The two groups were compares for age, sex, mechanism of injury, grade of liver injury scale, combined injury, systolic blood pressure, pulse rate, hemoglobin, hematocrit, WBC count, S-GOT, S-GPT, ALP, transfusion amount during initial 24 hours, amount of infused crystalloid fluid, length of ICU stay, length of ward care, morbidity and mortality. The grade of the liver injury were determined by using the organ injury scale(OSI). Results: Among the 84 patients, 46 cases(54.8%) were managed non-surgically, and 3 cases of Non-OP group were treated by transarterial embolization. Between the two groups, there were significant difference in age, injury grade, combined injury, hemoglobin, hematocrit, initial systolic blood pressure, amount of infused crystalloid fluid, amount of transfusion during the first 24 hours, and length of ICU care, morbidity and mortality.(p<0.05) The overall mortality rate was 8.3%, but 2.2% mortality in the non-operative group. Conclusion: Non-operative management may be considered as a first choice in hemodynamic stable patients with blunt liver trauma. The reliable indicators affecting the treatment modality of blunt hepatic trauma were systolic BP, Hb, Hct, amount of infused crystalloid fluid, amount of transfusion during the first 24 hours, liver injury grade and combined injury. Strict selection of treatment madality and aggresive monitoring with intensive care unit were more important.

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.10 no.4
• /
• pp.383-395
• /
• 2008

The grouting is one of the improved techniques which is aim to decrease the permeability and to strengthen the soft ground. But The grouting method has many problems about a suitability of grouting procedure and an effectiveness of grouting after grouting work because of a technical characteristic operated inside the soil. The grouting $p{\sim}q{\sim}t$ chart of a typical index about grouting rate and time alteration of grouting pressure is one method to estimate the suitability of grouting factor with monitoring during grouting procedure. This study is automatic grouting system (AGS) which can control the testing and grouting procedures. It can make the detailed $p{\sim}q{\sim}t$ chart and analyze the grouting characters of the ground by comparing the detailed pattern of $p{\sim}q{\sim}t$ chart with standard pattern. If using the $p{\sim}q{\sim}t$ chart derived from AGS in the grouting work, it is an objective standard estimating the suitability of grouting factor with grouting materials, grouting method, grouting rate and grouting pressure, as results it expects successfully to improve reliability of the grouting work.