• 지능정보연구
• /
• 제16권2호
• /
• pp.95-108
• /
• 2010

본 연구에서는 동영상으로부터 동적 수신호 패턴을 효과적으로 인식하기 위한 방법론으로서 복합형 신경망 모델을 제안한다. 제안된 모델은 특징추출 모듈과 패턴분류 모듈로 구성되는데, 이들 각각을 위하여 수정된 구조의 CNN 모델과, WFMM 모델을 도입한다. 또한 목표물의 움직임 정보에 기초한 시공간적 템플릿 구조의 데이터표현을 소개한다. 본 논문에서는 우선 수신호 패턴 데이터에서 특징점의 시간적 변이 및 공간적 변이에 의한 영향을 보완하기 위하여 3차원 수용영역 구조로 확장된 CNN 모델을 제시한다. 이어서 패턴분류 단계를 위하여 가중치를 갖는 구조의 FMM 신경망 모델을 소개하고, 신경망의 구조와 동작특성에 관해 기술한다. 또한 제안된 모델이 기존의 FMM 신경망에서 중첩 하이퍼박스의 축소과정에서 발생하는 학습효과의 왜곡현상을 개선할 수 있음을 보인다. 응용으로 가전제품 원격제어 문제를 전제하여 간략화된 수신호패턴 인식 문제에 적용한 실험결과로부터 제안된 이론의 타당성을 고찰한다.

• 한국정보통신학회:학술대회논문집
• /
• 한국해양정보통신학회 2005년도 춘계종합학술대회
• /
• pp.285-288
• /
• 2005

신경회로망은 지능제어알고리즘 중의 하나로 학습능력을 가지고 있다. 이러한 학습능력 때문에 많은 분야에서 널리 사용되고 있으나, 지능제어의 단점인 안정도 문제를 수학적으로 증명하기 어렵다는 문제점을 갖고 있다. 본 논문에서는 신경회로망의 한 종류인 RBFN과 적응제어기법을 이용하여 로봇 매니퓰레이터 궤적 제어기를 구성하고 자 한다. 본 논문에서는 RBFN의 파라메터들을 적응제어기법을 이용하여 수학적으로 구하였고, 시스템의 안정도를 수학적으로 UUB를 만족한다는 것을 증명하였다. 그리고 수평다관절로봇 매니퓰레이터 궤적제어기에 적용하였다.

• 전기학회논문지
• /
• 제60권4호
• /
• pp.862-870
• /
• 2011

In this paper, we introduce a new topology of Radial Basis Function-based Polynomial Neural Networks (RPNN) that is based on a genetically optimized multi-layer perceptron with Radial Polynomial Neurons (RPNs). This study offers a comprehensive design methodology involving mechanisms of optimization algorithms, especially Fuzzy C-Means (FCM) clustering method and Particle Swarm Optimization (PSO) algorithms. In contrast to the typical architectures encountered in Polynomial Neural Networks (PNNs), our main objective is to develop a design strategy of RPNNs as follows : (a) The architecture of the proposed network consists of Radial Polynomial Neurons (RPNs). In here, the RPN is fully reflective of the structure encountered in numeric data which are granulated with the aid of Fuzzy C-Means (FCM) clustering method. The RPN dwells on the concepts of a collection of radial basis function and the function-based nonlinear (polynomial) processing. (b) The PSO-based design procedure being applied at each layer of RPNN leads to the selection of preferred nodes of the network (RPNs) whose local characteristics (such as the number of input variables, a collection of the specific subset of input variables, the order of the polynomial, and the number of clusters as well as a fuzzification coefficient in the FCM clustering) can be easily adjusted. The performance of the RPNN is quantified through the experimentation where we use a number of modeling benchmarks - NOx emission process data of gas turbine power plant and learning machine data(Automobile Miles Per Gallon Data) already experimented with in fuzzy or neurofuzzy modeling. A comparative analysis reveals that the proposed RPNN exhibits higher accuracy and superb predictive capability in comparison to some previous models available in the literature.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2012년도 학술발표회
• /
• pp.39-39
• /
• 2012

Data assimilation techniques have received growing attention due to their capability to improve prediction in various areas. Despite of their potentials, applicable software frameworks to probabilistic approaches and data assimilation are still limited because the most of hydrologic modelling software are based on a deterministic approach. In this study, we developed a hydrological modelling framework for sequential data assimilation, namely MPI-OHyMoS. MPI-OHyMoS allows user to develop his/her own element models and to easily build a total simulation system model for hydrological simulations. Unlike process-based modelling framework, this software framework benefits from its object-oriented feature to flexibly represent hydrological processes without any change of the main library. In this software framework, sequential data assimilation based on the particle filters is available for any hydrologic models considering various sources of uncertainty originated from input forcing, parameters and observations. The particle filters are a Bayesian learning process in which the propagation of all uncertainties is carried out by a suitable selection of randomly generated particles without any assumptions about the nature of the distributions. In MPI-OHyMoS, ensemble simulations are parallelized, which can take advantage of high performance computing (HPC) system. We applied this software framework for several catchments in Japan using a distributed hydrologic model. Uncertainty of model parameters and radar rainfall estimates is assessed simultaneously in sequential data assimilation.

• 유통과학연구
• /
• 제13권1호
• /
• pp.7-18
• /
• 2015

Purpose - This study integrates a corporate competition system with a bird predation system to examine how organizational strategic adjustment capacity influences firm performance. By proving the prominent effects on performance, a financial vector is constructed to represent corporate strategic adjustment results, and an operation capacity vector is constructed, which can be categorized as a parameter for locating birds. All these works help us to propose a new method of investment, the portfolio decision model based on the strategic adjustment capacity. Research design, data, and methodology - Strategic adjustment capacity can be decomposed into three aspects: the organizational learning capacity from the top firms, the extent to which firms maintainor rely on the best operational capacity vector in history, and the ability to eliminate the disadvantages or retain the advantages of the operation capacity vector from the previous year. The method of solving cyclic equations is designed to evaluate strategic adjustment. Firms manufacturing specialized equipment are chosen to test the effects of the strategic adjustment capacity on three aspects of firm performance. Results - There is a positive correlation between the capacity to learn from the best firms and performance improvement. The relationship between the dependence or maintenance of a firm's advantages and performance improvement is a U-shape curve, and there is no significant effect of inertial control on performance improvement. Conclusions - A firm's competition system is a sophisticated adaptation, and competitive advantage and performance can be investigated based on the principles of competition in nature.

• 품질경영학회지
• /
• 제47권3호
• /
• pp.497-508
• /
• 2019

Purpose: The purpose of this study is to overcome limitations of conventional studies that to predict Baltic Dry Index (BDI). The study proposed applications of Artificial Neural Network (ANN) named Long Short-Term Memory (LSTM) to predict BDI. Methods: The BDI time-series prediction was carried out through eight variables related to the dry bulk market. The prediction was conducted in two steps. First, identifying the goodness of fitness for the BDI time-series of specific ANN models and determining the network structures to be used in the next step. While using ANN's generalization capability, the structures determined in the previous steps were used in the empirical prediction step, and the sliding-window method was applied to make a daily (one-day ahead) prediction. Results: At the empirical prediction step, it was possible to predict variable y(BDI time series) at point of time t by 8 variables (related to the dry bulk market) of x at point of time (t-1). LSTM, known to be good at learning over a long period of time, showed the best performance with higher predictive accuracy compared to Multi-Layer Perceptron (MLP) and Recurrent Neural Network (RNN). Conclusion: Applying this study to real business would require long-term predictions by applying more detailed forecasting techniques. I hope that the research can provide a point of reference in the dry bulk market, and furthermore in the decision-making and investment in the future of the shipping business as a whole.

• 한국건축시공학회지
• /
• 제21권5호
• /
• pp.397-408
• /
• 2021

건설업은 업무상 재해 발생빈도와 사망자 수가 다른 산업군에 비해 높아 가장 위험한 산업군으로 불린다. 정부는 건설 현장에서 발생하는 산업 재해를 줄이고 예방하기 위해 CCTV 설치 의무화를 발표했다. 건설 현장의 안전 관리자는 CCTV 관제를 통해 현장의 잠재된 위험성을 찾아 제거하고 재해를 예방한다. 하지만 장시간 관제 업무는 피로도가 매우 높아 중요한 상황을 놓치는 경우가 많다. 따라서 본 연구는 딥러닝 기반 컴퓨터 비전 모형 중 개체 분할인 YOLACT와 다중 객체 추적 기법인 SORT을 적용하여 다중 클래스 다중 객체 추적 시스템을 개발하였다. 건설 현장에서 촬영한 영상으로 제안한 방법론의 성능을 MS COCO와 MOT 평가지표로 평가하였다. SORT는 YOLACT의 의존성이 높아서 작은 객체가 적은 데이터셋을 학습한 모형의 성능으로 먼 거리의 물체를 추적하는 성능이 떨어지지만, 크기가 큰 객체에서 뛰어난 성능을 나타냈다. 본 연구로 인해 딥러닝 기반 컴퓨터 비전 기법들의 안전 관제 업무에 보조 역할로 업무상 재해를 예방할 수 있을 것으로 판단된다.

• 문화기술의 융합
• /
• 제5권1호
• /
• pp.77-82
• /
• 2019

인공지능은 기술개발 속도가 가속화되어 생활, 의료, 금융 서비스 및 자율자동차 등 산업 전반에 적용되고 있다. 4차 산업혁명 시대의 핵심기술로 자리 잡고 있는 인공지능 경쟁력 확보를 위해 선진국들은 국가적 역량을 집중하고 있다. 반면 IT강국으로서의 인프라와 인적자원을 보유한 한국은 미국, 캐나다, 일본, 등 전통적인 인공지능 선진국뿐만 아니라 지능형 기술집약 산업 육성에 총력을 기울이는 후발주자 중국에도 뒤쳐져있는 상황이다. 지능정보 사회의 고도화에 따라 인공지능은 향후 국가의 산업경쟁력을 좌우할 기반기술인바, 국가적인 관심과 역량 결집이 필요하다. 또한 인공지능 기술의 종속을 막기 위하여 자체 기술개발 노력과 함께 선두업체와의 공동 개발이 중요하다. 이에 더하여 인공지능 시장 저변 확대를 위하여 제도 개선과 법률적 기반 마련이 시급하다.

• 한국군사과학기술학회지
• /
• 제24권4호
• /
• pp.374-381
• /
• 2021

Raman spectrometers are studied and developed for the military purposes because of their nondestructive inspection capability to capture unique spectral features induced by molecular structures of colorless and odorless chemical warfare agents(CWAs) in any phase. Raman spectrometers often suffer from random noise caused by their detector inherent noise, background signal, etc. Thus, reducing the random noise in a measured Raman spectrum can help detection algorithms to find spectral features of CWAs and effectively detect them. In this paper, we propose a denoising autoencoder for Raman spectra with a loss function for sample efficient learning using noisy dataset. We conduct experiments to compare its effect on the measured spectra and detection performance with several existing noise reduction algorithms. The experimental results show that the denoising autoencoder is the most effective noise reduction algorithm among existing noise reduction algorithms for Raman spectrum based standoff detection of CWAs.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권7호
• /
• pp.2390-2406
• /
• 2022

Scene text recognition has important application value and attracted the interest of plenty of researchers. At present, many methods have achieved good results, but most of the existing approaches attempt to improve the performance of scene text recognition from the image level. They have a good effect on reading regular scene texts. However, there are still many obstacles to recognizing text on low-quality images such as curved, occlusion, and blur. This exacerbates the difficulty of feature extraction because the image quality is uneven. In addition, the results of model testing are highly dependent on training data, so there is still room for improvement in scene text recognition methods. In this work, we present a natural scene text recognizer to improve the recognition performance from the feature level, which contains feature representation and feature enhancement. In terms of feature representation, we propose an efficient feature extractor combined with Representative Batch Normalization and ResNet. It reduces the dependence of the model on training data and improves the feature representation ability of different instances. In terms of feature enhancement, we use a feature enhancement network to expand the receptive field of feature maps, so that feature maps contain rich feature information. Enhanced feature representation capability helps to improve the recognition performance of the model. We conducted experiments on 7 benchmarks, which shows that this method is highly competitive in recognizing both regular and irregular texts. The method achieved top1 recognition accuracy on four benchmarks of IC03, IC13, IC15, and SVTP.