• 디지털융복합연구
• /
• 제13권1호
• /
• pp.227-235
• /
• 2015

빅데이터분석은 조직의 문제해결을 위한 융합적 수단이다. 효과적인 문제해결을 위해서는 문제의 형태, 데이터의 유형 및 존재여부, 데이터 분석역량, 분석을 위한 기반정보기술의 수준 등 다양한 요인을 융합적으로 고려하여 문제해결의 접근법이 결정되어야 한다. 본 연구에서는 기획 접근법으로 논리적인 하향식 접근법, 데이터기반의 상향식 접근법, 그리고 문제해결 환경의 불확실성을 극복하기 위한 프로토타이핑 접근법 등 세 가지 유형을 제안한다. 특히, 이 유형 중에서 창의적 문제해결과 상향식 접근법이 어떤 연관성을 갖는지 살펴본다. 또한 데이터 거버넌스와 데이터 분석역량을 융합적으로 고려하여 조직의 빅데이터분석의 소싱과 관련한 주요 전략적 이슈를 도출한다.

• Steel and Composite Structures
• /
• 제20권3호
• /
• pp.623-649
• /
• 2016

Most of the early studies on plates vibration are focused on two-dimensional theories, these theories reduce the dimensions of problems from three to two by introducing some assumptions in mathematical modeling leading to simpler expressions and derivation of solutions. However, these simplifications inherently bring errors and therefore may lead to unreliable results for relatively thick plates. The main objective of this research paper is to present 3-D elasticity solution for free vibration analysis of continuously graded carbon nanotube-reinforced (CGCNTR) rectangular plates resting on two-parameter elastic foundations. The volume fractions of oriented, straight single-walled carbon nanotubes (SWCNTs) are assumed to be graded in the thickness direction. In this study, an equivalent continuum model based on the Eshelby-Mori-Tanaka approach is employed to estimate the effective constitutive law of the elastic isotropic medium (matrix) with oriented, straight carbon nanotubes (CNTs). The proposed rectangular plates have two opposite edges simply supported, while all possible combinations of free, simply supported and clamped boundary conditions are applied to the other two edges. The formulations are based on the three-dimensional elasticity theory. A semi-analytical approach composed of differential quadrature method (DQM) and series solution is adopted to solve the equations of motion. The fast rate of convergence of the method is demonstrated and comparison studies are carried out to establish its very high accuracy and versatility. The 2-D differential quadrature method as an efficient and accurate numerical tool is used to discretize the governing equations and to implement the boundary conditions. The convergence of the method is demonstrated and to validate the results, comparisons are made between the present results and results reported by well-known references for special cases treated before, have confirmed accuracy and efficiency of the present approach. The novelty of the present work is to exploit Eshelby-Mori-Tanaka approach in order to reveal the impacts of the volume fractions of oriented CNTs, different CNTs distributions, various coefficients of foundation and different combinations of free, simply supported and clamped boundary conditions on the vibrational characteristics of CGCNTR rectangular plates. The new results can be used as benchmark solutions for future researches.

• 융합정보논문지
• /
• 제7권6호
• /
• pp.113-121
• /
• 2017

IT 기술의 발달로 유해 멀티미디어가 무분별하게 유포되고 있다. 또한 선정적, 폭력적 유해 콘텐츠는 청소년에게 약 영향을 끼친다. 따라서 본 논문에서는 선정성, 폭력성이 드러나는 영상 콘텐츠 차단을 위한 다중 기법을 제안한다. 다중 기법 내에는 선정성, 폭력성을 검출하는 두 가지 모듈이 있다. 선정성 검출 모듈 내에는 성인 점수와 외설점수를 기반으로 선정성을 검출하는 모델이 있다. 폭력성 검출을 위한 모듈 내에는 RGB 영역을 이용한 피 검출 모델과 폭력적인 움직임은 방향과 크기 변화가 크다는 것에 착안한 움직임 추출 모델 두 가지가 있다. 이와 같은 총 세가지 모델의 검출 결과에 따라 해당 콘텐츠의 유해 여부를 판단한다. 본 논문의 유해 콘텐츠 차단 다중 기법은 무분별하게 유포되는 선정적, 폭력적 유해 콘텐츠를 차단한다.

• 융합신호처리학회논문지
• /
• 제2권1호
• /
• pp.1-11
• /
• 2001

자동차 번호판은 조명과 카메라에 따라 영상에서 다양한 형태로 나타나고 영상내의 잡음으로 인해 알고리즘 방식으로 자동차 번호판을 인식하기가 쉽지 않다. 이러한 문제에 적합한 해결 방법으로 본 논문에서는 학습 기반의 자동차 번호판 인식 시스템을 제안한다. 제안한 시스템은 자동차 검출 모듈, 번호판 추출 모듈, 번호판 문자 인식 모듈로 구성된다 본 논문에서는 자동차 번호판 추출을 위해서 입력 영상의 잡음에 상대적인 영향이 적은 시간-지연 신경망(Time-Delay Neural Networks : TDNN)과 번호판 인식을 위해서 일반적인 신경망보다 일반화 성능이 뛰어난 서포트 벡터 머신(Support Vector Machines : SVMs)을 시스템에 적용한다. 주차장과 톨게이트에서 여러 시간대의 움직이는 자동차 영상들을 실험한 결과, 번호판 추출율은 97.5%, 번호판 문자 인식률은 97.2%의 성능을 내었고, 전체 시스템 성능은 947%이며 처리 시간은 약 1조 미만이다. 따라서 본 논문에서 제안한 시스템은 실세계에서 유용하게 적용될 수 있다.

• ETRI Journal
• /
• 제33권2호
• /
• pp.251-258
• /
• 2011

Applying face alignment after face detection exerts a heavy influence on face recognition. Many researchers have recently investigated face alignment using databases collected from images taken at close distances and with low magnification. However, in the cases of home-service robots, captured images generally are of low resolution and low quality. Therefore, previous face alignment research, such as eye detection, is not appropriate for robot environments. The main purpose of this paper is to provide a new and effective approach in the alignment of small and blurred faces. We propose a face alignment method using the confidence value of Real-AdaBoost with a modified census transform feature. We also evaluate the face recognition system to compare the proposed face alignment module with those of other systems. Experimental results show that the proposed method has a high recognition rate, higher than face alignment methods using a manually-marked eye position.

• Nuclear Engineering and Technology
• /
• 제54권2호
• /
• pp.601-607
• /
• 2022

In recent years, monitoring of spent nuclear fuel inside dry cask storage has become an important area of national security. Muon tomography is a useful method for monitoring spent nuclear fuel because it uses high energy muons that penetrate deep into the target material and provides a 3-D structure of the inner materials. We designed a muon tomography system consisting of four 2-D position sensitive detector and characterized and optimized the system parameters. Each detector, measuring 200 × 200 cm², consists of a plastic scintillator, wavelength shifting (WLS) fibers and, SiPMs. The reconstructed image is obtained by extracting the intersection of the incoming and outgoing muon tracks using a Point-of-Closest-Approach (PoCA) algorithm. The Geant4 simulation was used to evaluate the performance of the muon tomography system and to optimize the design parameters including the pixel size of the muon detector, the field of view (FOV), and the distance between detectors. Based on the optimized design parameters, the spent fuel assemblies were modeled and the line profile was analyzed to conduct a feasibility study. Line profile analysis confirmed that muon tomography system can monitor nuclear spent fuel in dry storage container.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2021년도 추계학술발표대회
• /
• pp.805-807
• /
• 2021

In-hospital cardiac arrest is a significant problem for medical systems. Although the traditional early warning systems have been widely applied, they still contain many drawbacks, such as the high false warning rate and low sensitivity. This paper proposed a strategy that involves a deep learning approach based on a novel interpretable deep tabular data learning architecture, named TabNet, for the Rapid Response System. This study has been processed and validated on a dataset collected from two hospitals of Chonnam National University, Korea, in over 10 years. The learning metrics used for the experiment are the area under the receiver operating characteristic curve score (AUROC) and the area under the precision-recall curve score (AUPRC). The experiment on a large real-time dataset shows that our method improves compared to other machine learning-based approaches.

• International journal of advanced smart convergence
• /
• 제12권4호
• /
• pp.142-146
• /
• 2023

We present a method for generating 3D structures and rendering objects by combining VAE (Variational Autoencoder) and GAN (Generative Adversarial Network). This approach focuses on generating and rendering 3D models with improved quality using residual learning as the learning method for the encoder. We deep stack the encoder layers to accurately reflect the features of the image and apply residual blocks to solve the problems of deep layers to improve the encoder performance. This solves the problems of gradient vanishing and exploding, which are problems when constructing a deep neural network, and creates a 3D model of improved quality. To accurately extract image features, we construct deep layers of the encoder model and apply the residual function to learning to model with more detailed information. The generated model has more detailed voxels for more accurate representation, is rendered by adding materials and lighting, and is finally converted into a mesh model. 3D models have excellent visual quality and accuracy, making them useful in various fields such as virtual reality, game development, and metaverse.

• Journal of Microbiology and Biotechnology
• /
• 제29권2호
• /
• pp.304-310
• /
• 2019

Interleukin-21 is a common ${\gamma}$-chain cytokine that controls the immune responses of B cells, T cells, and natural killer cells. Targeting IL-21 to strengthen the immune system is promising for the development of vaccines as well as anti-infection and anti-tumor therapies. However, the practical application of IL-21 is limited by the high production cost. In this study, we improved IL-21 production by codon optimization and selection of appropriate signal peptide in CHO-K1 cells. Codon-optimized or non-optimized human IL-21 was stably transfected into CHO-K1 cells. IL-21 expression was 10-fold higher for codon-optimized than non-optimized IL-21. We fused five different signal peptides to codon-optimized mature IL-21 and evaluated their effect on IL-21 production. The best result (a 3-fold increase) was obtained using a signal peptide derived from human azurocidin. Furthermore, codon-optimized IL-21 containing the azurocidin signal peptide promoted $IFN-{\gamma}$ secretion and STAT3 phosphorylation in NK-92 cells similar to codon-optimized IL-21 containing original signal peptide. Collectively, these results indicate that codon optimization and azurocidin signal peptides provide an efficient approach for the high-level production of IL-21 as a biopharmaceutical.

• Korean Journal of Radiology
• /
• 제23권11호
• /
• pp.1089-1101
• /
• 2022

Immunotherapy has revolutionized and opened a new paradigm for cancer treatment. In the era of immunotherapy and molecular targeted therapy, precision medicine has gained emphasis, and an early response assessment is a key element of this approach. Treatment response assessment for immunotherapy is challenging for radiologists because of the rapid development of immunotherapeutic agents, from immune checkpoint inhibitors to chimeric antigen receptor-T cells, with which many radiologists may not be familiar, and the atypical responses to therapy, such as pseudoprogression and hyperprogression. Therefore, new response assessment methods such as immune response assessment, functional/molecular imaging biomarkers, and artificial intelligence (including radiomics and machine learning approaches) have been developed and investigated. Radiologists should be aware of recent trends in immunotherapy development and new response assessment methods.