• The Journal of the Acoustical Society of Korea
• /
• v.40 no.2
• /
• pp.121-129
• /
• 2021

This paper deals with a super-resolution that improves the resolution of side scan sonar images using learning-based compressive sensing. Learning-based compressive sensing combined with deep learning and compressive sensing takes a structure of a feed-forward network and parameters are set automatically through learning. In particular, we propose a method that can effectively extract additional information required in the super-resolution process through various initialization methods. Representative experimental results show that the proposed method provides improved performance in terms of Peak Signal-to-Noise Ratio (PSNR) and Structure Similarity Index Measure (SSIM) than conventional methods.

• Journal of the Korean Society of Safety
• /
• v.36 no.5
• /
• pp.52-60
• /
• 2021

As the prevention of fatal accidents is considered an essential part of social responsibilities, both government and individual have devoted efforts to mitigate the unsafe conditions and behaviors that facilitate accidents. Several studies have analyzed the factors that cause fatal accidents and compared them to those of non-fatal accidents. However, studies on mathematical and systematic analysis techniques for identifying the features of fatal accidents are rare. Recently, various industrial fields have employed machine learning algorithms. This study aimed to apply machine learning algorithms for the classification of fatal and non-fatal accidents based on the features of each accident. These features were obtained by text mining literature on accidents. The classification was performed using four machine learning algorithms, which are widely used in industrial fields, including logistic regression, decision tree, neural network, and support vector machine algorithms. The results revealed that the machine learning algorithms exhibited a high accuracy for the classification of accidents into the two categories. In addition, the importance of comparing similar cases between fatal and non-fatal accidents was discussed. This study presented a method for classifying accidents using machine learning algorithms based on the reports on previous studies on accidents.

• Journal of the Economic Geographical Society of Korea
• /
• v.25 no.4
• /
• pp.499-513
• /
• 2022

This study targets the rechargeable battery industry, which has been rapidly growing recently. The rechargeable battery industry is closely related to the electric vehicle industry. However, other factors also influence it. Currently, rechargeable battery companies show a pattern of restructuring production space by various means. To determine the causes of these production spaces, the factors affecting regional and national scales were thoroughly examined. As a result, the location factors for rechargeable battery-related companies are determined by cooperative relationships with assembled car companies, government policy regulations, and the stability of supply of key materials. And a spatial strategy was implemented to make the most of these circumstances.

• Journal of the Korean Society of Industry Convergence
• /
• v.26 no.5
• /
• pp.867-874
• /
• 2023

In this paper, a wireless communication-based sensor data monitoring device with an explosion-proof (Exd IIC) case was implemented to enable installation at explosion-risk industrial sites such as plants. In existing industrial plant sites, most of the temperature sensors and vibration and impact sensors are wired up to several kilometers, which takes a lot of time and money to bury long pipes and cables. In addition, there are not many cases where some wireless devices have been applied to actual plant industry sites due to communication quality problems. Therefore, in order to solve this problem, zigbee mesh wireless communication was applied to provide high reliability wireless communication quality to industrial plant sites, and the time and cost incurred in new or additional installation of sensors could be greatly reduced. In particular, in the event of loss or error of some wireless communication devices, the communication network is automatically bypassed or recovered to enable real-time data monitoring.

• ETRI Journal
• /
• v.45 no.6
• /
• pp.1079-1089
• /
• 2023

Stroke is one of the leading causes of long-term disability worldwide, placing huge burdens on individuals and society. Further, automatic human activity recognition is a challenging task that is vital to the future of healthcare and physical therapy. Using a baseline long short-term memory recurrent neural network, this study provides a novel dataset of stretching, upward stretching, flinging motions, hand-to-mouth movements, swiping gestures, and pouring motions for improved model training and testing of stroke-affected patients. A MATLAB application is used to output textual and audible prediction results. A wearable sensor with a triaxial accelerometer is used to collect preprocessed real-time data. The model is trained with features extracted from the actual patient to recognize new actions, and the recognition accuracy provided by multiple datasets is compared based on the same baseline model. When training and testing using the new dataset, the baseline model shows recognition accuracy that is 11% higher than the Activity Daily Living dataset, 22% higher than the Activity Recognition Single Chest-Mounted Accelerometer dataset, and 10% higher than another real-world dataset.

• IMMUNE NETWORK
• /
• v.22 no.3
• /
• pp.27.1-27.13
• /
• 2022

Little is known of the lung cellular immunophenotypes in patients with non-tuberculous mycobacterial lung disease (NTM-LD). Flow-cytometric analyses for the major myeloid and lymphoid cell subsets were performed in less- and more-diseased areas of surgically resected lungs from six patients with NTM-LD and two with Pseudomonas aeruginosa lung disease (PsA-LD). Lymphocytes, comprised mainly of NK cells, CD4⁺ and CD8⁺ T cells, and B cells, accounted for ~60% of all leukocytes, with greater prevalence of T and B cells in more-diseased areas. In contrast, fewer neutrophils were found with decreased number in more-diseased areas. Compared to NTM-LD, lung tissues from patients with PsA-LD demonstrated relatively lower numbers of T and B lymphocytes but similar numbers of NK cells. While this study demonstrated a large influx of lymphocytes into the lungs of patients with chronic NTM-LD, further analyses of their phenotypes are necessary to determine the significance of these findings.

• Korean Journal of Health Education and Promotion
• /
• v.27 no.5
• /
• pp.15-25
• /
• 2010

Objectives: The objectives of this study was to evaluate Wonju Healthy City project and identify its problems, and seeking a way for its improvement based on the Healthy City project philosophy and strategies. Methods: We used the SPIRIT Checklist that was a process evaluation tool and developed by Alliance for Healthy Cities for the study. We analyzed 39 related materials and gathered opinions on the evaluation result with Healthy City Team staffs, related department staffs and the advisory committee. Finally, a joint meeting with AFHC SPIRIT evaluation expert verified the result of the analysis. Results: The evaluation of Wonju Healthy City project confirmed that Wonju city is equipped with the resources, such as mid-term plan, infrastructure, cooperative organizations, and the Healthy City network to enable the consistent implementation of the Healthy City project based on strong political commitment. However, the necessity of additional complementary processes as well as the application of further improvements to assist health promotion strategies was evident. Conclusion: It is required to improve Wonju Healthy City project that activation of health promotion programs based on the political support and cooperation with public health center and Healthy City project departments in city hall.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.30 no.11B
• /
• pp.750-757
• /
• 2005

This paper reports on the study of P2P traffic behaviors in a high-speed campus network under a simple firewall which drops packets with default port numbers for the well-blown P2P applications. Among several ways of detecting P2P traffic, the easiest method is to filter out packets with the default port number of each P2P application. After deploying the port-based firewall against P2P-traffic, it is expected that the amount of P2P traffic will be decreased. However, during the eight-month measurement period, three new commercial P2P applications have been identified and their traffic usages have reached up to $30/5.6\%$ of the total outbound/inbound traffic volumes at the end of the measurement period. In addition, the most famous P2P application, eDonkey, has adapted and has escaped detection through port hopping. The measurement result shows that the amount of eDonkey traffic is around $6.7/4.0\%$ of the total outbound/inbound traffic volume. From the measurement results, it is observed that the port-based firewall is not effective to limit the usage of P2P applications and that the P2P traffic is steadily growing due to not only the evolution of existing P2P applications such as port hopping but also appearances of new P2P applications.

• Journal of IKEEE
• /
• v.13 no.2
• /
• pp.140-149
• /
• 2009

Traffic anomaly detection is one of important technology that should be considered in network security and administration. In this paper, we propose an abnormal traffic detection mechanism that includes traffic monitoring and traffic analysis. We develop analytical passive monitoring system called WISE-Mon which can inspect traffic behavior. We establish a criterion by analyzing the characteristics of a traffic training set. To detect abnormal traffic, we derive a hyperplane by using Fisher linear discriminant and chi-square distribution as well as the analyzed characteristics of traffic. Our mechanism can support reliable results for traffic anomaly detection and is compatible to real-time detection. In addition, since the trend of traffic can be changed as time passes, the hyperplane has to be updated periodically to reflect the changes. Accordingly, we consider the self-learning algorithm which reflects the trend of the traffic and so enables to increase the pliability of detection probability. Numerical results are presented to validate the accuracy of proposed mechanism. It shows that the proposed mechanism is reliable and relevant for traffic anomaly detection.

• The Journal of the Acoustical Society of Korea
• /
• v.39 no.6
• /
• pp.615-621
• /
• 2020

In this paper, we propose a multi-site based earthquake event classification method using graph convolution networks. In the traditional earthquake event classification methods using deep learning, they used single-site observation to estimate seismic event class. However, to achieve robust and accurate earthquake event classification on the seismic observation network, the method using the information from the multi-site observations is needed, instead of using only single-site data. Firstly, our proposed model employs convolution neural networks to extract informative embedding features from the single-site observation. Secondly, graph convolution networks are used to integrate the features from several stations. To evaluate our model, we explore the model structure and the number of stations for ablation study. Finally, our multi-site based model outperforms up to 10 % accuracy and event recall rate compared to single-site based model.