• Journal of information and communication convergence engineering
• /
• 제2권1호
• /
• pp.40-45
• /
• 2004

This paper describes a modeling of service composition manager based on TINA (Telecommunication Information Networking Architecture). The Service composition function is mainly motivated by the desire to easily generate new service using existing services from retailers or $3^{rd}$-party service providers. The TINA-C specification for the service composition does not include the detailed composition procedure and its object models. In this paper, we propose a model of components for the service composition, which adapts a static composition feature in a single provider domain. To validate the proposed modeling, we implemented prototype service composition function, which combines two multimedia services; a VOD service and a VCS service. As a result, we obtain the specification of the detailed composition architecture between a retailer domain and a $3^{rd}$-party service provider domain.

• 대한전자공학회논문지SP
• /
• 제43권4호
• /
• pp.46-54
• /
• 2006

최근 각종 온라인 상거래 및 개인 신분카드 이용이 늘어나면서 개인 인증의 중요성이 부각되고 있다. RFID(Radio Frequency Identification) tag가 내장된 개인 신분 카드가 점차 증가하고 있지만, 본인의 인증을 할 수 있는 방법이 미비하기 때문에, 자동화 할 수 있는 대책이 시급하다. RFID tag는 현재 메모리 용량이 매우 작기 때문에, 개인의 생체정보를 저장하기 위해서는 효율적인 특징추출 방법이 필요하며, 저장된 특징들을 비교하기 위해서는 새로운 인식방법이 필요하다. 본 논문에서는 인간의 인지학적인 두뇌 원리인 해마 신경망을 공학적으로 모델링하여 얼굴 영상의 특징 벡터들을 고속 학습하고, 각 영상의 최적의 특정을 구성할 수 있는 해마 신경망 모델링 알고리즘을 이용한 개인생체 인증 시스템에 관한 연구를 수행하였다. 시스템은 크게 NMF(Non-negative Matrix Factorization)와 LDA(Linear Discriminants Analysis) 혼합 알고리즘을 이용한 특징 추출 부분과 해마신경망을 모델링하고 인식 성능을 실험하는 것으로 구성 되어 있다. 제안한 시스템의 성능을 평가하기 위하여 실험은 표정변화와 포즈변화가 포함된 이미지를 각각 구분하여 인식률을 확인하였다. 실험 결과, 본 논문에서 제안하는 특정 추출 방법과 학습 방법을 다른 방법들과 비교하였을 때, 학습시간비용과 인식률에서 우수함을 확인하였다.

• ETRI Journal
• /
• 제29권4호
• /
• pp.527-529
• /
• 2007

In this paper, a decision-tree-based Markov model for phrase break prediction is proposed. The model takes advantage of the non-homogeneous-features-based classification ability of decision tree and temporal break sequence modeling based on the Markov process. For this experiment, a text corpus tagged with parts-of-speech and three break strength levels is prepared and evaluated. The complex feature set, textual conditions, and prior knowledge are utilized; and chunking rules are applied to the search results. The proposed model shows an error reduction rate of about 11.6% compared to the conventional classification model.

• 한국정보과학회논문지:소프트웨어및응용
• /
• 제36권6호
• /
• pp.431-442
• /
• 2009

소프트웨어 제품 생산 라인(Software Product Line) 공학은 새로운 소프트웨어 개발 패러다임으로 각광받고 있다. SPL에 FORM(Feature-Oriented Reuse Method) 방법론을 적용하면, 휴대전화나 디지털TV 같이 공통점이 많은 제품군의 다양한 소프트웨어를 휘처 모델링을 통해 만들어진 재사용 가능하고 유연한 컴포넌트를 조합하여 생산해 낼 수 있다. 한편, MDA(Model Driven Architecture) 방법론은 PIM(Platform Independent Model) 을 통해 다양한 개별 플랫폼을 위한 소프트웨어를 생산할 수 있게 하는 새로운 기술을 제공한다. 위 두 가지 방법론의 장점을 조합하면 공통점을 공유하면서 다양한 플랫폼에서 동작하는 제품군의 소프트웨어를 생산하는데 도움이 된다. 이 논문에서는 FORM 방법론과 MDA 방법론을 조합하기 위해 먼저, 프로파일링 기법을 통해 UML2.0을 확장하여 FORM 아키텍처와 Parameterized Statechart 모델링이 가능하게 한다. 다음으로, 휘처가 휘처 모델과 Parameterized Statechart사이에서 일관성 있게 element의 형태로 위치하고 있는지 검증하는 일관성 규칙을 제공한다. 몇 가지 규칙은 FORM 아키텍처와 Parameterized Statechart 사이의 일관성을 검사하기 위해 고안되었다. 마지막으로, 엘리베이터 시스템의 사례연구를 통해 이 논문에서 제안하는 모델링 기법과 일관성 검사 법칙의 유효성을 제시한다.

• 음성과학
• /
• 제12권2호
• /
• pp.171-182
• /
• 2005

An efficient GMM (Gaussian mixture modeling) method based on LPCA (local principal component analysis) with VQ (vector quantization) for speaker identification is proposed. To reduce the dimension and correlation of the feature vector, this paper proposes a speaker identification method based on principal component analysis. The proposed method firstly partitions the data space into several disjoint regions by VQ, and then performs PCA in each region. Finally, the GMM for the speaker is obtained from the transformed feature vectors in each region. Compared to the conventional GMM method with diagonal covariance matrix, the proposed method requires less storage and complexity while maintaining the same performance requires less storage and shows faster results.

• 한국정밀공학회:학술대회논문집
• /
• 한국정밀공학회 1997년도 춘계학술대회 논문집
• /
• pp.410-415
• /
• 1997

Recently, The manufacturing companies are introducing the CAD/CAM systems to solve problems for the lack of experts, the higher cost of manufacturing and the difficulties of process. Knowledge engineering approach makes it possible to change a know-how of experts to computerized information effectivly. The proposal of this paper is the development of an interactive knowledge-based CAM system to disign and manufacture the mold with non-expert engineers used easily. This system is composed of two functional parts. One is the geometric modeler that used the technique of a feature modeling. The other is the expert system module that composed inference engine and databas which contains characteristics of materials and cutting tools setc.

• Journal of Power Electronics
• /
• 제17권2호
• /
• pp.411-421
• /
• 2017

An approach based on a 2D lumped model is presented to quantify the voltage transfer gain (VTG) in power converter low power planes. The advantage of the modeling approach is the ease with which typical noise reduction devices such as decoupling capacitors or ferrite beads can be integrated into the model. This feature is enforced by a new modular approach based on effective matrix partitioning, which is presented in the paper. This partitioning is used to decouple power plane equations from external device impedance, which avoids the need for rewriting of a whole set of equation at every change. The model is quickly solved in the frequency domain, which is well suited for an automated layout optimization algorithm. Using frequency domain modeling also allows the integration of frequency-dependent devices such inductors and capacitors, which are required for realistic computation results. In order to check the precision of the modeling approach, VTGs for several layout configurations are computed and compared with experimental measurements based on scattering parameters.

• ETRI Journal
• /
• 제36권4호
• /
• pp.662-672
• /
• 2014

Human action recognition is used in areas such as surveillance, entertainment, and healthcare. This paper proposes a system to recognize both single and continuous human actions from monocular video sequences, based on 3D human modeling and cyclic hidden Markov models (CHMMs). First, for each frame in a monocular video sequence, the 3D coordinates of joints belonging to a human object, through actions of multiple cycles, are extracted using 3D human modeling techniques. The 3D coordinates are then converted into a set of geometrical relational features (GRFs) for dimensionality reduction and discrimination increase. For further dimensionality reduction, k-means clustering is applied to the GRFs to generate clustered feature vectors. These vectors are used to train CHMMs separately for different types of actions, based on the Baum-Welch re-estimation algorithm. For recognition of continuous actions that are concatenated from several distinct types of actions, a designed graphical model is used to systematically concatenate different separately trained CHMMs. The experimental results show the effective performance of our proposed system in both single and continuous action recognition problems.

• 한국멀티미디어학회논문지
• /
• 제7권5호
• /
• pp.617-632
• /
• 2004

비디오 데이터에는 시간의 변화에 따라 공간적인 위치가 변화하는 움직임 객체를 포함하고 있다. 이 논문에서는 비디오 데이터의 움직임 객체에 대한 새로운 모델링 방법을 제안한다. 제안하는 모델링은 움직임 객체를 효과적으로 검색하기 위해 시간의 변화에 따라 공간적인 위치와 크기 변화를 표현한다. 또한 객체의 시간에 따른 시각적 특징 변화와 객체의 방향, 거리 그리고 속도를 고려한 궤적을 표현한다. 따라서 움직임 객체의 시각적인 특징 유사도 검색, 거리 유사도 검색, 제적 유사도 검색을 수행할 수 있다. 또한 이들을 통합한 가중치 검색이 가능하도록 한다.

• 정보처리학회논문지A
• /
• 제8A권1호
• /
• pp.63-69
• /
• 2001

In this paper, we present an interactive 3D character modeling and animation including a data management tool for editing the animation. It includes an animation editor for changing animation sequences according to the modified structure of 3D object in the object structure editor. The animation tool has the feature that it can produce motion data independently of any modeling tool including our modeling tool. Differently from conventional 3D graphics tools that model objects based on geometrically calculated data, our tool models 3D geometric and animation data by approximating to the real object using 2D image interactively. There are some applications that do not need precise representation, but an easier way to obtain an approximated model looking similar to the real object. Our tool is appropriate for such applications. This paper has focused on the data management for enhancing the automatin and convenience when editing a motion or when mapping a motion to the other character.