• Proceedings of the Acoustical Society of Korea Conference
• /
• 1994.06c
• /
• pp.407-410
• /
• 1994

In this paper, we describe the new unsuperivised learning algorithm, SASOM. It can solve the defects of the conventional SOM that the state of network can't converge to the minimum point. The proposed algorithm uses the object function which can evaluate the state of network in learning and adjusts the learning rate flexibly according to the evaluation of the object function. We implement the simulated annealing which is applied to the conventional network using the object function and the learning rate. Finally, the proposed algorithm can make the state of network converged to the global minimum. Using the two-dimensional input vectors with uniform distribution, we graphically compared the ordering ability of SOM with that of SASOM. We carried out the recognitioin on the new algorithm for all Korean phonemes and some continuous speech.

• Proceedings of the Korea Contents Association Conference
• /
• 2011.05a
• /
• pp.573-574
• /
• 2011

기존의 4-State 바코드 인식 방법은 정확한 바코드 인식을 위해 모든 윤곽선을 이용하여 외곽선을 추출하고, 학습에 따른 분석 모델을 적용하는 방식이었지만, 본 논문에서는 바코드의 최외각의 윤곽선을 이용하여 불필요한 연산 작업을 최소화하고 인식과정에서 3-line 화소 구분법으로 학습데이터의 사용의존도를 최소화하는 고속 우편물 인식 시스템 개발에 목적을 둔다.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1996.10a
• /
• pp.333-336
• /
• 1996

The sign language is a method of communication for deaf person. For sign communication, sign language and manual alphabet are used continuously. In this paper is proposed a system which recognize Korean sign language(KSL) and Korean manual alphabet(KMA) continuously. For recognizing KSL and KMA, basic elements for sign language, namely, the 14 hand directions, 23 hand postures, and 14 hand orientations are used. At first, this system recognize current motion state using speed and change of speed in motion by state automata. Using state, basic element classifiers using Fuzzy Min-Max Neural Network and Fuzzy Rule are executed. Meaning of signed gesture is selected by using basic elements which was recognized.

• Journal of Multimedia Information System
• /
• v.8 no.1
• /
• pp.31-34
• /
• 2021

The classical HMM is defined by a parameter triple �� = (��, A, B), where each parameter represents a collection of probability distributions: initial state, state transition and output distributions in order. This paper proposes a new stationary parameter e = (e1, e2, …, eN) where N is the number of states and et = P(|xt = i, y) for describing how an input pattern y ends in state xt = i at time t followed by nothing. It is often said that all is well that ends well. We argue here that all should end well. The paper sets the framework for the theory and presents an efficient inference and training algorithms based on dynamic programming and expectation-maximization. The proposed model is applicable to analyzing any sequential data with two or more finite segmental patterns are concatenated, each forming a context to its neighbors. Experiments on online Hangul handwriting characters have proven the effect of the proposed augmentation in terms of highly intuitive segmentation as well as recognition performance and 13.2% error rate reduction.

• Asian-Australasian Journal of Animal Sciences
• /
• v.24 no.5
• /
• pp.723-738
• /
• 2011

Contamination of agricultural crops by mycotoxins results in significant economic losses for grain producers and, when consumed, it can cause reduced growth and health in a wide range of animal species. Hundreds of mycotoxin producing molds exist, however each has a different frequency and pattern of occurrence, as well as differences in the severity of the diseases (mycotoxicoses) they cause. Among the mycotoxins considered to be major contaminates are aflatoxin, deoxynivalenol, fumonisin, ochratoxin, and zearalenone. Although a multitude of species can be harmed by consumption of these mycotoxins, swine appear to be the most commonly affected commodity species. The swine industry can thus experience great losses due to the presence of mycotoxin contamination in feeds. Subsequently, recognition and prevention of mycotoxicoses is extremely important and dependent on adequate grain sampling and analysis methods pre-harvest, as well as effective strategies post-harvest to reduce consumption by animals. The aim of this review is to provide an overview of the major mycotoxin contaminants in grains, to describe methods of analysis and prevention to reduce mycotoxicoses in swine and other animals, and finally to discuss how mycotoxins directly affect swine production.

• Earthquakes and Structures
• /
• v.14 no.2
• /
• pp.143-153
• /
• 2018

Correlation among different factors must be considered for selection of influencing factors in safety monitoring of high dam including positive correlation of variables. Therefore, a new factor selection method was constructed based on Copula entropy and mutual information theory, which was deduced and optimized. Considering the small sample size in high dam monitoring and distribution of daily monitoring samples, a computing method that avoids causality of structure as much as possible is needed. The two-dimensional normal information diffusion and fuzzy reasoning of pattern recognition field are based on the weight theory, which avoids complicated causes of the studying structure. Hence, it is used to dam safety monitoring field and simplified, which increases sample information appropriately. Next, a complete system integrating high dam monitoring and uncertainty prediction method was established by combining Copula entropy theory and information diffusion theory. Finally, the proposed method was applied in seepage monitoring of Nuozhadu clay core-wall rockfill dam. Its selection of influencing factors and processing of sample data were compared with different models. Results demonstrated that the proposed method increases the prediction accuracy to some extent.

• Journal of radiological science and technology
• /
• v.27 no.3
• /
• pp.51-58
• /
• 2004

The gel, which is stained on probe after ultrasonography, is a good circumstances for proliferation of microbe. This study is to investigate into the actual state of sanitary management, recognition degree and infection level of ultrasonographic probes. We had performed a question with telephone to 42 hospitals in Seoul area from December in 2003. We also cultured to obtained a sample from three ultrasonographic units to investigate infection level of the probes. Sanitary management of the probes was performed in 21 hospitals with alcohol cotton. Sanitary management was performed daily in 14 hospitals. Most hospitals used cotton towel for clearing of gel stained on probes. Preventive management against infection was performed in 32 hospitals with vinyl cover, surgical glove, or alcohol sterilization etc. In the recognition degree on infection, the response that using method of ultrasonographic probes is insanitary were in 78.6%(33 hospitals), and 54.8%(23 hospitals) responded that bacteria can be infected through the probes. In the results of germiculture, bacteria and fungi were detected too number of to count, but escherichia coli was not detected. In conclusion, The gel stained on probe after ultrasonography must be cleared completely, and it is necessary that change of recognition on sanitary management.

• Journal of the Institute of Convergence Signal Processing
• /
• v.8 no.3
• /
• pp.178-184
• /
• 2007

The EEG signal in general can be categorized as the Alpha wave, the Beta wave, the Theta wave, and the Delta wave. The alpha wave, showed in stable state, is the dominant wave for a human EEG and the beta wave displays the excited state. The subject of this paper was to recognize the stable state of EEG quantitatively using wavelet transform and power spectrum analysis. We decomposed EEG signal into the alpha wave and the beta wave in the process of wavelet transform, and calculated each power spectrum of EEG signal, using Fast Fourier Transform. And then we calculated the stable state quantitatively by stable state ratio, defined as the power spectrum of the alpha wave over that of the beta wave. The study showed that it took more than 10 minutes to reach the stable state from the normal activity in 69 % of the subjects, 5 -10 minutes in 9%, and less than 5 minutes in 16 %.

• Journal of Internet Computing and Services
• /
• v.12 no.2
• /
• pp.9-17
• /
• 2011

This paper proposes an algorithm that predicts the state of user's next actions, exploiting the HMM (Hidden Markov Model) on user profile data stored in the ubiquitous home network. The HMM, recognizes patterns of sequential data, adequately represents the temporal property implicated in the data, and is a typical model that can infer information from the sequential data. The proposed algorithm uses the number of the user's action performed, the location and duration of the actions saved by "Activity Recognition System" as training data. An objective formulation for the user's interest in his action is proposed by giving weight on his action, and change on the state of his next action is predicted by obtaining the change on the weight according to the flow of time using the HMM. The proposed algorithm, helps constructing realistic ubiquitous home networks.

• Smart Structures and Systems
• /
• v.14 no.2
• /
• pp.145-158
• /
• 2014

The objective of this study is to improve the survivability and reliability of the FBG sensor network in the structural health monitoring (SHM) system. Therefore, a model reconstruction soft computing recognition algorithm based on support vector regression (SVR) is proposed to achieve the high reliability of the FBG sensor network, and the grid search algorithm is used to optimize the parameters of SVR model. Furthermore, in order to demonstrate the effectiveness of the proposed model reconstruction algorithm, a SHM system based on an eight-point fiber Bragg grating (FBG) sensor network is designed to monitor the foreign-object low velocity impact of a CFRP composite plate. Simultaneously, some sensors data are neglected to simulate different kinds of FBG sensor network failure modes, the predicting results are compared with non-reconstruction for the same failure mode. The comparative results indicate that the performance of the model reconstruction recognition algorithm based on SVR has more excellence than that of non-reconstruction, and the model reconstruction algorithm almost keeps the consistent predicting accuracy when no sensor, one sensor and two sensors are invalid in the FBG sensor network, thus the reliability is improved when there are FBG sensors are invalid in the structural health monitoring system.