• Proceedings of the KSRS Conference
• /
• 2007.10a
• /
• pp.270-273
• /
• 2007

In this paper, a SAR-despeck1ing approach of adaptive iteration based a Bayesian model using the lognormal distribution for image intensity and a Gibbs random field (GRF) for image texture is proposed for noise removal of the images that are corrupted by multiplicative speckle noise. When the image intensity is logarithmically transformed, the speckle noise is approximately Gaussian additive noise, and it tends to a normal probability much faster than the intensity distribution. The MRF is incorporated into digital image analysis by viewing pixel types as states of molecules in a lattice-like physical system. The iterative approach based on MRF is very effective for the inner areas of regions in the observed scene, but may result in yielding false reconstruction around the boundaries due to using wrong information of adjacent regions with different characteristics. The proposed method suggests an adaptive approach using variable parameters depending on the location of reconstructed area, that is, how near to the boundary. The proximity of boundary is estimated by the statistics based on edge value, standard deviation, entropy, and the 4th moment of intensity distribution.

• 한국가시화정보학회:학술대회논문집
• /
• 2004.04a
• /
• pp.1-7
• /
• 2004

When a vortex diffracts upon encountering a vortex, many strong and weak waves are produced in the course of interaction. They are the cause of shock wave attenuation and noise production. This phenomenon is fundamental to understanding the more complex supersonic turbulent Jet noise. In this paper we have reviewed the research on shock-vortex interaction we have carried on last seven years. We have computationally investigated the parameter effect. When a shock is strong, shock diffraction pattern becomes complex since the slip lines from the triple points on Mach stem curl into the vortex, causing an entropy layer. When the vortex is unstable, vortexlets are brought about each of which make shock diffraction of a reduced intensity. Strong vortex produces quadrupole noise as it impinges into a vortex. Elementary interaction models such as shock splitting, shock reflection, and shock penetration are presented based on shock tube experiment. These models are also verified by computational approach. They easily explain production and propagation of the aforementioned quadrupole noise, Diverging acoustics are explained in terms of shock-vortexlet interactions for which a computational model Is constructed.

• Annual Conference on Human and Language Technology
• /
• 2008.10a
• /
• pp.183-188
• /
• 2008

본 연구에서는 문서 안에 있는 문장들 중 비교 문장을 추출해낸다. 비교 문장이란 두 개 이상의 객체, 혹은 한 객체의 시간차, 공간차 등에 따른 변화를 비교하는 내용을 포함하는 문장을 말한다. 비교 문장을 구별해내는 작업은 많은 분야에서 응용될 수 있는데, 특히 객체(사람, 상품 등)에 대한 평가 면에서 매우 직접적이고 확실한 자료로 활용될 수 있다. 비교문장 추출을 위해 본 연구에서는 비교어휘를 이용한 추출 및 MEM(Maximum Entropy Model)을 적용하였으며, 뉴스기사(news article), 상품에 대한 고객리뷰(customer review) 등의 문서를 대상으로 실험하여 재현율 88.40%, 정확률 88.68%의 결과를 산출하였다.

• International Journal of Air-Conditioning and Refrigeration
• /
• v.17 no.2
• /
• pp.74-79
• /
• 2009

The performance of cryogenic refrigerator greatly depends on the effectiveness of heat exchanger, which generates major entropy at low temperature. There are numerous types of heat exchanger available, but it is not easy to apply most of them to cryogenic application because the cryogenic heat exchanger must have high effectiveness value as well as small conduction loss in the environment of considerable temperature difference. In this paper, two kinds of heat exchanger are noticeably introduced for high-effectiveness miniature cryogenic recuperator(recuperative heat ex-changer). Also, the flow mal-distribution problem, which is a critical issue of performance deterioration in a high-effectiveness recuperator, is addressed with simplified model, and its alleviation method is discussed.

• Annual Conference on Human and Language Technology
• /
• 2016.10a
• /
• pp.249-251
• /
• 2016

다중 화자 대화 시스템에서, 시스템의 입장에서 어느 시점에 참여해야하는지를 아는 것은 중요하다. 이러한 참여 모델을 구축함에 있어서 본 연구에서는 다수의 화자가 대화에 참여하는 영화 대본으로 구축된 MovieDic 말뭉치를 사용하였다. 구축에 필요한 자질로써 의문사, 호칭, 명사, 어휘 등을 사용하였고, 훈련 알고리즘으로는 Maximum Entropy Classifier를 사용하였다. 실험 결과 53.34%의 정확도를 기록하였으며, 맥락 자질의 추가로 정확도 개선을 기대할 수 있다.

• Journal of the Korean Society of Propulsion Engineers
• /
• v.5 no.1
• /
• pp.1-9
• /
• 2001

BGK schemes are developed by improving the standard BGK numerical method. Shceme 1 uses the Osher's Godunov type solution and scheme 2 are developed to overcome the problems of scheme 1.The improved schemes show many unique properties such as entropy condition, positivity condition, higher order gas evolution model, which lead to an high degree of robustness and accuracy. The scheme 2 especially overcomes the shortcomings of the scheme 1 and posseses many superior properties that cannot be found ohter numerical schemes, and is expected to apply various problems with high accuracy and robustness.

• Annual Conference on Human and Language Technology
• /
• 2006.10e
• /
• pp.213-219
• /
• 2006

본 논문에서는 스팸메일에 나타나는 스팸성 자질과 URL 자질을 이용한 최대엔트로피모델 기반 스팸 필터 시스템을 제안한다. 스팸성 자질은 스패머들이 스팸메일에 인위적으로 넣는 강조 패턴이나 필터 시스템을 통과하기 위해 비정상적으로 변형시킨 단어들을 말한다. 스팸성 자질 외에 반복적으로 나타나는 URL과 비정상적인 Ink도 자질로 사용하였다. 메일 수신자에게 추가적인 정보 제공을 목적으로 하이퍼링크로 연결시키거나 메일에 직접 타이핑한 URL 중 필터 시스템을 피하기 위해 유효하지 알은 비정상적인 URL들이 스팸 메일을 걸러내는데 도움을 줄 수 있기 때문이다. 또한 스팸성 자질과 URL을 각각 적용한 두 분류기를 통합하였다. 분류기의 통합은 각 분류기에 이용된 자질을 독립적으로 사용할 수 있다는 장점을 가지고 있다. 실험 결과를 통해 스팸성 자질과 URL을 이용함으로써 스팸 필터 시스템의 성능을 향상시킬 수 있음을 확인할 수 있었다.

• Journal of information and communication convergence engineering
• /
• v.13 no.2
• /
• pp.113-122
• /
• 2015

Feature subset selection is as a pre-processing step in learning algorithms. In this paper, we propose an efficient algorithm, ModifiedFAST, for feature subset selection. This algorithm is suitable for text datasets, and uses the concept of information gain to remove irrelevant and redundant features. A new optimal value of the threshold for symmetric uncertainty, used to identify relevant features, is found. The thresholds used by previous feature selection algorithms such as FAST, Relief, and CFS were not optimal. It has been proven that the threshold value greatly affects the percentage of selected features and the classification accuracy. A new performance unified metric that combines accuracy and the number of features selected has been proposed and applied in the proposed algorithm. It was experimentally shown that the percentage of selected features obtained by the proposed algorithm was lower than that obtained using existing algorithms in most of the datasets. The effectiveness of our algorithm on the optimal threshold was statistically validated with other algorithms.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2019.10a
• /
• pp.1101-1104
• /
• 2019

키넥트는 RGBD 카메라로 인체의 뼈대와 관절을 3D 공간에서 스켈레톤 데이터수집을 가능하게 해주었다. 스켈레톤 데이터를 활용한 행동 분류는 RNN, CNN 등 다양한 인공 신경망으로 접근하고 있다. 본 연구는 키넥트를 이용해서 Skeleton Joints를 수집하고, DNN 기반 스켈레톤 모델링 학습으로 행동을 분류한다. Skeleton Joints Processing 과정은 키넥트의 Depth Map 기반의 Skeleton Tracker로 25가지 Skeleton Joints 좌표를 얻고, 학습을 위한 전처리 과정으로 각 좌표를 상대좌표로 변경하고 데이터 수를 제한하며, Joint가 트래킹 되지 않은 부분에 대한 예외 처리를 수행한다. 스켈레톤 모델링 학습 과정에선 3계층의 DNN 신경망을 구축하고, softmax_cross_entropy 함수로 Skeleton Joints를 집는 모션, 내려놓는 모션, 팔짱 낀 모션, 얼굴을 가까이 가져가는 모션 해서 4가지 행동으로 분류한다.

• Bulletin of the Korean Chemical Society
• /
• v.4 no.3
• /
• pp.133-139
• /
• 1983

To understand the importance of Y-base adjacent to the anticodon stabilizing codon-anticodon interaction, a study has been undertaken for the model compound involving the interaction between Y-base and anticodon as well as the participation of water molecule by calculating the conformational free energy using an empirical potential function. We restrict our analysis to sites directly associated with Y-base by varying only the backbone torsion angles of Y-base. The hydration and $Mg^{+2}$ binding effects on the conformational stability of Y-base are calculated and discussed. The free Y-base is proved to be less stable than the hydrated one. The free energy change due to the hydration of Y-base amounts to -119.5 kcal/mole, in which the conformational energy change is -142.4 kcal/mole and the configurational entropy change is -76.9 e. u. It is found that the water molecules bound to Y-base and $Mg^{+2}$ contribute to the conformation of Y-base dominantly.