• Journal of the Korea Society of Computer and Information
• /
• v.25 no.12
• /
• pp.63-71
• /
• 2020

In this paper, we propose a fire detection technology using YOLOv3 and EfficientDet, the most reliable artificial intelligence detection algorithm recently, an alert service that simultaneously transmits four kinds of notifications: text, web, app and e-mail, and an AWS system that links fire detection and notification service. There are two types of our highly accurate fire detection algorithms; the fire detection model based on YOLOv3, which operates locally, used more than 2000 fire data and learned through data augmentation, and the EfficientDet, which operates in the cloud, has conducted transfer learning on the pretrained model. Four types of notification services were established using AWS service and FCM service; in the case of the web, app, and mail, notifications were received immediately after notification transmission, and in the case of the text messaging system through the base station, the delay time was fast enough within one second. We proved the accuracy of our fire detection technology through fire detection experiments using the fire video, and we also measured the time of fire detection and notification service to check detecting time and notification time. Our AI fire detection and notification service system in this paper is expected to be more accurate and faster than past fire detection systems, which will greatly help secure golden time in the event of fire accidents.

• Journal of Korea Multimedia Society
• /
• v.25 no.6
• /
• pp.769-782
• /
• 2022

A study on agricultural automation became more important. In Korea, sweet persimmon farmers spend a lot of time and effort on classifying profitable persimmons. In this paper, we propose and implement an efficient grading algorithm for persimmons before shipment. We gathered more than 1,750 images of persimmons, and the images were graded and labeled for classifications purpose. Our main algorithm is based on EfficientDet object detection model but we implemented more exquisite method for better classification performance. In order to improve the precision of classification, we adopted a machine learning algorithm, which was proposed by PyCaret machine learning workflow generation library. Finally we acquired an improved classification model with the accuracy score of 81%.

• Journal of Electrochemical Science and Technology
• /
• v.8 no.2
• /
• pp.87-95
• /
• 2017

Enzymatic fuel cells are promising low cost, compact and flexible energy resources. The basis of enzymatic fuel cells is transfer of electron from enzyme to the electrode surface and vice versa. Electron transfer is done either by direct or mediated electron transfer (DET/MET), each one having its own advantages and disadvantages. In this study, the DET and MET of laccase-based biocathodes are compared with each other. The DET of laccase enzyme has been studied using two methods; assemble of needle-like carbon nanotubes (CNTs) on the electrode, and CNTs/Nafion polymer. MET of laccase enzyme also is done by use of ceramic electrode containing, ABTS (2,2'-azino-bis [3-ethylbenzthiazoline-6-sulphonic acid]) /sol-gel. Cyclic voltammetric results of DET showed a pair of well-defined redox peaks at $200{\mu}A$ and $170{\mu}A$ in a solution containing 5and $10{\mu}M$ o-dianisidine as a substrate for needle-like assembled CNTs and CNTs-Nafion composite respectively. In MET method using sol-gel/ABTS, the maximum redox peak was $14{\mu}A$ in the presence of 15 M solution o-dianisidine as substrate. The cyclic voltammetric results showed that laccase immobilization on needle-like assembled CNTs or CNTs-Nafion is more efficient than the sol-gel/ABTS electrode. Therefore, the expressed methods can be used to fabricate biocathode of biofuel cells or laccase based biosensors.

• Korea Journal of Hospital Management
• /
• v.23 no.3
• /
• pp.87-103
• /
• 2018

Purpose: This study investigates the best appointment pattern which can enhance customer's satisfaction and hospital's efficient management reflecting plastic surgery clinic's service characteristics. Methodology: The data of this study is obtained from Plastic surgery Clinic which is located in the civic center. By collecting and analyzing the data, we build the simulation model using ARENA. Based on 5 appointment patterns that was suggested in formal appointment scheduling studies, we experiment 3 simulation models; 'Basic Appointment Pattern' that has no restriction, 'Restriction on Second Customer' that restricts the number of second customer's in each slot, 'Restriction on Process Time' that restricts the number of second customer who has long process time in each slot. We can check robustness of the appointment patterns by experimenting on off-peak day and peak day, during peak season. Findings: This study confirms that these 2 restrictions can give a better result than 'Basic Appointment Pattern' that just simply distributes customers by number. Especially, the performance of Triangle-like pattern which is the best appointment pattern in the formal study has been improved by adding restrictions. Based on 'DET', 'Restriction on Second Customer' shows a better result. Meanwhile, based on 'E(WT)', 'Restriction on Process Time' shows a better result. Overall, based on 'DET+E(WT)', 'Restriction on Second Customer' shows a better result. Practical Implications: The purpose of each hospital may alter as demand for plastic surgery grows increasingly. Thus, each hospital should be always prepared to introduce appointment pattern for changed purpose. In order to respond flexibly to these changes, it is necessary for medical personnel to improve the awareness or for hospital to create an environment by constructing appointment program so that medical personnel does not need to put more labor on work.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2024.01a
• /
• pp.73-75
• /
• 2024

본 논문에서는 밀 수확량을 증가시키기 위한 일반화된 검출 모델을 제안한다. 일반화 성능을 높이기 위해 CutMix 알고리즘으로 데이터를 증식시켰고, 라벨링 되지 않은 데이터를 최대한 활용하기 위해 Fast R-CNN 기반 Pseudo labeling을 사용하였다. 학습의 정확성과 효율성을 높이기 위해 사전에 훈련된 EfficientDet 모델로 학습하였으며, OOF를 이용하여 검증하였다. 최신 객체 검출 모델과 IoU(Intersection over Union)를 이용한 성능 평가 결과, 제안된 모델이 가장 높은 성능을 보이는 것을 확인하였다.

• Proceedings of the Korea Society for Energy Engineering kosee Conference
• /
• 2004.11a
• /
• pp.133-138
• /
• 2004

Proton exchange membrane fuel cell (PEMFC) is considered as a clean and efficient energy conversion det ice for mobile and stationary applications. Anions all the components of the PEMFC, the interface between the electrolyte ,and electrode catalyst plays an important role in determining tile cell performance since the electrochemical reactions take place at the interface in contact with tile reactant gases. Therefore, to increase the interface area and obtain a high-performance PEMFC, surface of the electrolyte membrane was roughened by Ar$^{+}$ beam bombardment. The results imply that by modifying surface of the electrolyte membrane, platinum loading can be reduced significantly without performance loss. To optimize the surface treatment condition, effects of ion dose density on characteristics of the membrane/electrode interface were examined by measuring the cell performance, impedance spectroscopy, and cyclic voltammograms. Surface of the modified membranes were characterized using scanning electron microscopy and FT-IR.R.

• Geomechanics and Engineering
• /
• v.38 no.1
• /
• pp.29-44
• /
• 2024

This research investigates the effectiveness of data augmentation techniques in the automated analysis of B-scan images from ground-penetrating radar (GPR) using deep learning. In spite of the growing interest in automating GPR data analysis and advancements in deep learning for image classification and object detection, many deep learning-based GPR data analysis studies have been limited by the availability of large, diverse GPR datasets. Data augmentation techniques are widely used in deep learning to improve model performance. In this study, we applied four data augmentation techniques (geometric transformation, color-space transformation, noise injection, and applying kernel filter) to the GPR datasets obtained from a testbed. A deep learning model for GPR data analysis was developed using three models (Faster R-CNN ResNet, SSD ResNet, and EfficientDet) based on transfer learning. It was found that data augmentation significantly enhances model performance across all cases, with the mAP and AR for the Faster R-CNN ResNet model increasing by approximately 4%, achieving a maximum mAP (Intersection over Union = 0.5:1.0) of 87.5% and maximum AR of 90.5%. These results highlight the importance of data augmentation in improving the robustness and accuracy of deep learning models for GPR B-scan analysis. The enhanced detection capabilities achieved through these techniques contribute to more reliable subsurface investigations in geotechnical engineering.

• Journal of Broadcast Engineering
• /
• v.27 no.2
• /
• pp.185-197
• /
• 2022

Recently, deep learning networks with high performance for object recognition are emerging. In the case of object recognition using deep learning, it is important to build a training data set to improve performance. To build a data set, we need to collect and label the images. This process requires a lot of time and manpower. For this reason, open data sets are used. However, there are objects that do not have large open data sets. One of them is data required for license plate detection and recognition. Therefore, in this paper, we propose an artificial license plate generator system that can create large data sets by minimizing images. In addition, the detection rate according to the artificial license plate arrangement structure was analyzed. As a result of the analysis, the best layout structure was FVC_III and B, and the most suitable network was D2Det. Although the artificial data set performance was 2-3% lower than that of the actual data set, the time to build the artificial data was about 11 times faster than the time to build the actual data set, proving that it is a time-efficient data set building system.