• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.34 no.12C
• /
• pp.1177-1183
• /
• 2009

In this paper, support vector machines (SVM) are applied to multi-user detector (MUD) for direct sequence (DS)-CDMA system. This work shows an analytical performance of SVM based multi-user detector with some of kernel functions, such as linear, sigmoid, and Gaussian. The basic idea in SVM based training is to select the proper number of support vectors by maximizing the margin between two different classes. In simulation studies, the performance of SVM based MUD with different kernel functions is compared in terms of the number of selected support vectors, their corresponding decision boundary, and finally the bit error rate. It was found that controlling parameter, in SVM training have an effect, in some degree, to SVM based MUD with both sigmoid and Gaussian kernel. It is shown that SVM based MUD with Gaussian kernels outperforms those with other kernels.

• Proceedings of the KIEE Conference
• /
• 2008.04a
• /
• pp.219-220
• /
• 2008

The paper is proposed maximum torque control of SynRM drive using adaptive fuzzy neuro inference system(AFNIS) and artificial neural network(ANN). The control method is applicable over the entire speed range and considered the limits of the inverter's current and voltage rated value. For each control mode, a condition that determines the optimal d-axis current $i_d$ for maximum torque operation is derived. The proposed control algorithm is applied to SynRM drive system controlled AFNIS and ANN controller, the operating characteristics controlled by maximum torque control are examined in detail. Also, this paper is proposed the analysis results to verify the effectiveness of the AFNIS and ANN controller.

• Journal of Information Technology Applications and Management
• /
• v.19 no.1
• /
• pp.85-100
• /
• 2012

The development of technology causes a lot of change. Many researchers have insisted that the proliferation of cyber space changes the physical space. Their insistences have been accumulated into three aspects. Firstly, the proliferation of cyber space brings out the concentration in the physical space, secondly the decentralization and lastly both at the same time. In Korea, the concentration of plastic surgery clinics has taken place in Kangnam-gu area at similar period of the Internet proliferation. In this research, we execute empirical study of whether the concentration of plastic surgery in specific areas correlates with the proliferation of cyber space or not. In order to do this, we verified homogeneity of plastic surgery websites between Kangnam-gu and Non-Kangnam-gu areas. Also, we used three statistical and data-mining techniques which are Multi-discriminant analysis, Decision tree analysis and artificial neural network analysis. As a result, there was homogeneity between two different area plastic surgery clinics websites, but there was not big heterogeneity as well. Therefore, in this case of concentration of plastic surgery in Korea, the proliferation of cyber space restrictively correlates with the concentration of physical space.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.10 no.1
• /
• pp.75-80
• /
• 2010

Noise handing is very important in voice recognition of vehicle environment. that has been studying about to hardware and software approach. hardware method that is noise filter circuit design, basically using Low-pass filter. it was shown a good result. and the side of software that has been developing about to algorithm for Noise canceler, NN(neural network), etc. in this paper we have analysis about to classified parameter pattern matting level for voice recognition on car noise environment that use of DTW(Dynamic Time Warping) which is applicable time series pattern recognition algorithm.

• Physical Therapy Korea
• /
• v.12 no.3
• /
• pp.67-73
• /
• 2005

Prospective memory (PM) is related to remember to carry out a previously intented behaviour. The purpose of this study was to develop a paradigm for measuring PM function to diagnosis in mild cognitive impairment 1 or brain injury in patients 2. among brain injured patients Thirty-eight normal healthy subjects participated in current study. The paradigm was composed of four conditions: a baseline and three intention conditions (expectation, execution 1 and 2). In the expectation condition, subjects were asked to make a new response to intented stimuli during ongoing task, but the intented stimuli never occurred. In the execution 1 (one type of expected stimulus) and 2 (two types of expected stimuli), the intended stimuli did occur in 20% of trials. The reaction time and error rate were calculated in each condition. Repeated measures using ANOVA of subject's mean reaction times (RTs) and mean error rates (ERs) showed main effects of conditions during ongoing task. The comparison of PM tasks in executive condition 1 and 2 also showed significance in RTs and ERs. This paradigm reflects sufficiently the performance of prospective memory function during ongoing task in normal individuals. Thus, we suggest that the paradigm will be helpful to study neural network of PM function using brain imaging techniques and diagnosis of PM dysfunction.

• Journal of Advanced Navigation Technology
• /
• v.5 no.1
• /
• pp.85-96
• /
• 2001

Recent development of a smart structures module and its successful integration with a multidisciplinary design optimization software $ASTROS^*$ and an Aeroservoelasticity module is presented. A modeled F-16 wing using piezoelectric actuators is used as an example to demonstrate the integrated software capability to design a flutter suppression system. For an active control design, neural network based controller is used for this study. A smart structures module is developed by modifying the existing thermal loads module in $ASTROS^*$ in order to include the effects of the induced strain due to piezoelectric actuation. The control surface/piezoelectric equivalence model principle is developed, which ensures the interchangeability between the control surface force input and the piezoelectric force input to the Aeroservoelasticity modules in $ASTROS^*$. The results show that the developed controller can increase the flutter speed.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.20 no.3
• /
• pp.861-871
• /
• 1996

In this study, a fuzzy control algorithm was developed for the randomly irregular shape of cold-rolled strip. Currently developed fuzzy control algorithm consists of two parts: the first part calculates the changes of work and intermediate roll bender forces based on the symmetric part of the irregular strip shape, and the second part calculates the weighting factors based on the asymmetric part and modifies the pre-determined roll bender forces according to the weighting factors. As a result of this, bender froces applied at the both sides of the cold-rolled strip were different. In order to simulate the continuous shape control. fuzzy controller developed was linked with emulator which was developed based on neural network. The fuzzy controller and emulator developed simulated the cold rolling process until irregular shape converged to a tolerable range in producing uniform cross-sectional strip shape. The results obtained from the simulation were reasonable for various irregular strip shapes.

• 한국HCI학회:학술대회논문집
• /
• 2006.02a
• /
• pp.988-992
• /
• 2006

본 논문은 PC카메라 환경에서 명암왜곡에 강인한 얼굴검출을 위한 컴포넌트 기반 얼굴검출 기법을 제시한다. 영상 내의 얼굴검출을 위해 에지(edge) 분석, 색상 분석, 형판정합(template matching), 신경망(Neural Network), PCA(Principal Component Analysis), LDA(Linear Discriminant Analysis) 등의 기법들이 사용되고 있고, 영상의 왜곡을 보정하기 위해 히스토그램 분석(평활화, 명세화), gamma correction, log transform 등의 영상 보정 방법이 사용되고 있다. 그러나 기존의 얼굴검출 방법과 영상보정 방법은 검출대상 객체의 부분적인 잡음 및 조명의 왜곡에 대처하기가 어려운 단점이 있다. 특히 PC카메라 환경에서 획득된 이미지와 같이 전면과 후면, 상하좌우에서 비추어지는 조명에 의해 검출 대상 객체의 일부분이 왜곡되는 상황이 발생될 경우 기존의 방법으로는 높은 얼굴 검출 성능을 기대할 수 없는 상황이 발생된다. 본 논문에서는 기울어진 얼굴 및 부분적으로 명암 왜곡된 얼굴을 효율적으로 검출할 수 있도록 얼굴의 좌우 대칭성을 고려한 가로방향의 대칭평균화로 얼굴검출을 위한 모델을 생성하여 얼굴검출에 사용한다. 이 방법은 부분적으로 명암왜곡된 얼굴이미지를 기존의 영상 보정기법을 적용한 것 보다 잘 표현하며, 얼굴이 아닌 후보는 비얼굴 이미지의 형상을 가지게 하는 특성이 있다.

• 한국HCI학회:학술대회논문집
• /
• 2006.02a
• /
• pp.547-554
• /
• 2006

우리 춤은 우리 고유의 정서를 담고 있는 종합예술이므로 우리 춤을 분석하고 이해하는 것은 큰 의미가 있다. 본 논문에서는 기존의 춤 동작의 정량적인 분석을 통한 감정인식 기술을 이용하여 우리 춤에 내포된 감정 패턴의 변화를 살펴본다. 먼저 한국 전통춤으로부터 무용전문가들의 정성적 분석에 기반하여 추출된 우리 춤사위를 정해진 각 감정별로 재구성하여 창작하고 창작된 우리 춤을 무용전문가가 시연한다. 이를 카메라를 이용하여 획득하고, 영상처리를 통해서 시연자의 실루엣을 뽑아낸 후, 정량적 특징량들을 추출한다. 이어 신경회로망을 이용하여 각 감정별 춤사위를 학습 시킨 후, 임의의 춤사위에 내포된 감정을 인식 한다. 본 논문에서는 정면, 좌, 우 세 시점에서 획득된 다시점 영상을 이용하여 학습시킴으로써 보다 안정적으로 동작하는 인식 시스템을 제안한다. 그리고, 시스템에 의해 인식된 감정 패턴과 변화의 정성적 의미를 이해하기 위해 무용전문가들에 의해 정립된 정성적 분석 결과와 비교, 분석한다. 이는 정성적인 분석에만 국한되던 우리 춤의 특성에 대한 이해를 객관적이고 정량화된 분석을 통한 이해의 차원으로 확장시키는 것으로, 우리 춤의 특성을 새롭게 정의하는 계기를 마련할 수 있다. 다양한 장르의 한국 전통춤 가운데 우리 춤을 대표할 수 있는 춤사위를 선정하고, 정성적/정량적으로 분석함으로써 우리 춤의 특성을 이해하기 위한 체계적인 틀을 제공하고자 한다.

• Journal of Institute of Control, Robotics and Systems
• /
• v.12 no.4
• /
• pp.389-394
• /
• 2006

Inaccurate localization exposes a robot to many dangerous conditions. It could make a robot be moved to wrong direction or damaged by collision with surrounding obstacles. There are numerous approaches to self-localization, and there are different modalities as well (vision, laser range finders, ultrasonic sonars). Since sensor information is generally uncertain and contains noise, there are many researches to reduce the noise. But, the correctness is limited because most researches are based on statistical approach. The goal of our research is to measure more exact robot location by matching between built VRML 3D model and real vision image. To determine the position of mobile robot, landmark-localization technique has been applied. Landmarks are any detectable structure in the physical environment. Some use vertical lines, others use specially designed markers, In this paper, specially designed markers are used as landmarks. Given known focal length and a single image of three landmarks it is possible to compute the angular separation between the lines of sight of the landmarks. The image-processing and neural network pattern matching techniques are employed to recognize landmarks placed in a robot working environment. After self-localization, the 2D scene of the vision is overlaid with the VRML scene.