• Journal of KIISE:Databases
• /
• v.34 no.6
• /
• pp.528-545
• /
• 2007

Recently ontology has been an attractive technology along with the business strategy of providing a plenty of more intelligent services. The essential problem in application domains using ontology is that all members, agents, and application programs in the domains must share the same ontology concepts. However, a variety of mobile devices, sensing devices, and network components manufactured by various companies, a variety of common carriers, and a variety of contents providers make multiple heterogeneous ontologies more likely to coexist. We can see many past researches fallen into resolving this semantic interoperability. Such methods can be broadly classified into by-mapping, by-merging, and by-translation. In this research, we focus on by-translation among them which uses a translation rule directly made between two heterogeneous ontology data like OntoMorph. However, the manual composition of the direct translation rule is not convenient by itself and if there are N ontologies, the direct method has the rule composition complexity of $O(N^2)$ in the worst case. Therefore, in this paper we introduce the cascade composition of translation rules based on web openness in order to improve the complexity. The research result made us recognize some important factors in an ontology translation system, that is speediness of translation, and conveniency of translation rule composition, and some experiments and comparing analysis with existing methods showed that our cascade method has more conveniency with insuring the speediness and the correctness.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.15 no.3
• /
• pp.1724-1733
• /
• 2014

This paper presents a pattern recognition analysis of two spirals problem and optimization of Cascade Correlation learning algorithm using in combination with a non-monotone function as CosExp(cosine-modulated symmetric exponential function) and a monotone function as sigmoid function. In addition, the algorithm's optimization is attempted. By using genetic algorithms the optimization of the algorithm will attempt. In the first experiment, by using CosExp activation function for candidate neurons of the learning algorithm is analyzed the recognized pattern in input space of the two spirals problem. In the second experiment, CosExp function for output neurons is used. In the third experiment, the sigmoid activation functions with various parameters for candidate neurons in 8 pools and CosExp function for output neurons are used. In the fourth experiment, the parameters are composed of 8 pools and displacement of the sigmoid function to determine the value of the three parameters is obtained using genetic algorithms. The parameter values applied to the sigmoid activation functions for candidate neurons are used. To evaluate the performance of these algorithms, each step of the training input pattern classification shows the shape of the two spirals. In the optimizing process, the number of hidden neurons was reduced from 28 to15, and finally the learning algorithm with 12 hidden neurons was optimized.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.24 no.12B
• /
• pp.2407-2414
• /
• 1999

This paper introduces the concept of warped discrete cosine transform (WDCT) and an image compression algorithm based on the WDCT. The proposed WDCT is a cascade connection of a conventional DCT and all-pass filters whose parameters can be adjusted to provide frequency warping. In the proposed image compression scheme, the frequency response of the all-pass filter is controlled by a set of parameters with each parameter for a specified frequency range. For each image block, the best parameter is chosen from the set and is sent to the decoder as a side information along with the result of corresponding WDCT computation. For actual implementation, the combination of the all-pass IIR filters and the DCT can be viewed as a cascade of a warping matrix and the DCT matrix, or as a filter bank which is obtained by warping the frequency response of the DCT filter bank. Hence, the WDCT can be implemented by a single matrix computation like the DCT. The WDCT based compression, outperforms the DCT based compression, for high bit rate applications and for images with high frequency components.

• Annual Conference of KIPS
• /
• 2020.05a
• /
• pp.184-187
• /
• 2020

본 논문에서는 실시간 얼굴인식 기술을 활용한 인물 선별 송출 시스템을 제안한다. 실시간 동영상 안에 등장하는 다수의 인물 객체 얼굴을 검출하고 인식하기 위해 Haar 특징 정보 기반의 다단계(Cascade) 학습 알고리즘을 사용한다. 이 시스템은 다수의 인물을 인식하고 학습할 수 있으며 인식된 각 인물의 송출 여부를 사용자가 직접 선택할 수 있는데 이 모든 선택 송출 과정을 실시간으로 처리할 수 있다. 여기에서 제시한 기술은 다자간 화상 채팅이나 다자간 화상 회의에서 특정인의 프라이버시 보호를 위한 기술로 활용될 수 있다.

• Annual Conference of KIPS
• /
• 2015.10a
• /
• pp.1365-1368
• /
• 2015

컴퓨터 비전 분야의 활용영역과 시장성이 증대하면서 가장 많이 사용되는 객체인식 및 탐지 기술과 관련된 연구는 꾸준히 진행되고 있다. 최근에는 ADAS(Advanced Driver Assistance Systems)와 특징적인 객체를 인식 추적할 수 있는 지능형 감시시스템에서의 가장 핵심적인 기술로 자리 잡고 있다. 본 연구에서는 보행자 탐지에 사용하는 특징들 중에서 조명변화에 강건한 HOG와 Cascade-Adaboost를 기반으로 보행자 탐지 모델을 후보영역을 검출하고 검출된 영역에서 컬러정보를 추출하여 의사결정 트리에 적용시켜 최종 보행자를 탐지하는 시스템을 제안한다.

• Proceedings of the IEEK Conference
• /
• 1999.06a
• /
• pp.361-364
• /
• 1999

A 200 MHz CMOS transconductor using translinear cells is presented. It adopts a cascade circuit consisting of a voltage follower(VF), a resistor, and a current follower(CF). The translinear cell which has substancely high-frequency is used as the VF and CF, respectively Simulation results show that the proposed transconductors have the 3-㏈ frequency of a 200 MHz, and the dynamic range of $\pm$2.5 V for a supply voltage $\pm$3 V.

• Proceedings of the IEEK Conference
• /
• 1999.11a
• /
• pp.55-58
• /
• 1999

For the asymmetric digital subscriber line communications, the paper proposes a new echo canceller with the so-called CALAOI(CAscade of LAttice and Orthogonalized IIR) structure, which comprises of a lattice and an orthogonalized IIR structure. Through simulations, the CALAOI echo canceller was verified to have much fewer complexity of computations and has faster convergence speed than conventional FIR echo canceller. The CALAOI echo canceller is predicted to maximize performances of communication services in high speed communications such as VDSL, GIGA bit Ethernet, and so on as well as ADSL.

• Proceedings of the IEEK Conference
• /
• 2003.07e
• /
• pp.2248-2251
• /
• 2003

Communication between speech networks employing different speech codecs requires interoperability. The cascade connection of two different codecs, called tandem coding, not only degrades speech quality, but also produces high computational loads. These Problems can be solved by using the transcoding algorithm. This paper presents an effective algorithm for transcoding from G.729A to EVRC and its simulation results.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.30B no.8
• /
• pp.16-22
• /
• 1993

As a method of separating the signal arriving from a desired direction in the presence of noise and interfering signals, a linearly constrained adaptive beamformer based on the escalator (Gram-Schmidt) structure is presented. The weights are chosen to minimize local output power subject to the response constraint. The performances of the proposed escalator adaptive beamformer are compared with those of cascade adaptive beamformer via computer simulations.

• Proceedings of the Korean Institute of Communication Sciences Conference
• /
• 1983.10a
• /
• pp.56-60
• /
• 1983

There has been a great amount of interest in the design of digital filters with low sensitivity to coefficient variations. Especiaily the wave digital filter modeled after analog IC ladder filter has been studied to have low-cocfficient-sensitivity properties. This paper examined the design of the wave digital filter and how the sensitivity and roundoff noise porperty arises. As a result of computer simulation the implementation of the digital filter was possible with a lower coefficient word length comparing with the conventional cascade structure.