Earticle

현재 국내 교량 구조물은 지속적으로 증가 및 대형화되고 있으며 그에 따라 공용된 지 30년 이상 된 노후 교량도 꾸준히 늘어나고 있다. 교량 노후화 문제는 국내뿐 아니라 전 세계적으로도 심각한 사회 문제로 다루어지고 있으며, 기존 인력 위주의 점검 방식은 그 한계점을 드러내고 있다. 최근 들어 딥러닝 기반의 영상처리 알고리즘을 활용한 각종 교량 손상탐지 연구가 이루어지고 있지만 교량 손상 데이터 세트의 한계로 인하여 주로 균열 1종에 국한된 교량 손상탐지 연구가 대부분이고, 이 또한 Close set 분류모델 기반 탐지방식으로서 실제 교량 촬영 영상에 적용했을 시 배경이나 기타 객체 등 학습되지 않은 클래스의 입력 이미지들로 인하여 심각한 오인식 문제가 발생할 수 있다. 본 연구에서는 균열 포함 5종의 교량 손상을 정의 및 데이터 세트를 구축해서 딥러닝 모델로 학습시키고, OpenMax 알고리즘을 적용한 Open set 인식 기반 교량 다중손상 인식 모델을 개발했다. 그리고 학습되지 않은 이미지들을 포함하고 있는 Open set에 대한 분류 및 인식 성능평가를 수행한 후 그 결과를 분석했다.

Currently, the number of bridge structures in Korea is continuously increasing and enlarged, and the number of old bridges that have been in service for more than 30 years is also steadily increasing. Bridge aging is being treated as a serious social problem not only in Korea but also around the world, and the existing manpower-centered inspection method is revealing its limitations. Recently, various bridge damage detection studies using deep learning-based image processing algorithms have been conducted, but due to the limitations of the bridge damage data set, most of the bridge damage detection studies are mainly limited to one type of crack, which is also based on a close set classification model. As a detection method, when applied to an actual bridge image, a serious misrecognition problem may occur due to input images of an unknown class such as a background or other objects. In this study, five types of bridge damage including crack were defined and a data set was built, trained as a deep learning model, and an open set recognition-based bridge multiple damage recognition model applied with OpenMax algorithm was constructed. And after performing classification and recognition performance evaluation on the open set including untrained images, the results were analyzed.

Mobile learning allows for an interactive way of learning through devices like smartphones. However, current methods usually rely on pre-set situations and struggle to recognize new contexts when they come up during testing. To solve this, we suggest the Open-set Learning Context Recognition Model (OLCRM). This model uses data extracted from smartphone sensors to identify whether a learning context is known or unknown. It also uses a Dual Discriminator Generative Adversarial Network (DDGAN) to create high-quality fake examples, which helps improve the accuracy of recognizing contexts. Experimental results demonstrate the effectiveness of OLCRM in open-set learning context recognition problems.

열린 집합 인식 방법론은 테스트 데이터의 클래스를 학습 시에 모두 파악할 수 없는 경우에 대한 인식 방법론이다. 따라서 열린 집합 인식 방법론은 분류와 유효성 검증의 절차를 필요로 한다. 이러한 연구는 얼굴 인식 모듈의 상용화를 위해 필수적이지만 지금까지 국내에서 연구 결과들이 거의 발표되지 않았다. 우리는 두 개의 검증 단계를 가지는 열린 집합 얼굴 인식 방법론을 제안한다. 첫 번째 단계에서는 학습 클래스 외에 더미 클래스들을 설정하고 희소표현 기반 분류를 수행한다. 이 때 테스트 데이터가 더미 클래스로 분류되면 무효 데이터로 판별하고, 유효한 클래스로 분류되면 다음 유효성 검증 단계로 넘어간다. 두 번째 단계에서 제안하는 네 가지 특징을 추출하고, 확률분포에 기반을 둔 판별함수를 통해 유효성 검증을 수행한다. 우리는 실험을 통해 열린 집합 인식 방법론의 시뮬레이션 방법을 제안하였고 제안하는 방법론의 성능을 제시하고, 희소기반 분류 방식에서 널리 사용되는 SCI 지표를 이용한 유효성 테스트보다 높은 성능을 보임을 입증할 수 있었다.

The open set recognition method should be used for the cases that the classes of test data are not known completely in the training phase. So it is required to include two processes of classification and the validation test. This kind of research is very necessary for commercialization of face recognition modules, but few domestic researches results about it have been published. In this paper, we propose an open set face recognition method that includes two sequential validation phases. In the first phase, with dummy classes we perform classification based on sparse representation. Here, when the test data is classified into a dummy class, we conclude that the data is invalid. If the data is classified into one of the regular training classes, for second validation test we extract four features and apply them for the proposed decision function. In experiments, we proposed a simulation method for open set recognition and showed that the proposed validation test outperform SCI of the well-known validation method

The purpose of this paper is to estimate visitor expenditures of the open set of TV drama, Haesin in Wando-gun and to analyse economic impacts by input-output model. On-site survey was conducted for the visitors in May of 2005. It is revealed that visitor's per capita expenditure is 50,134 won and 27,160 won was expended in Wando. The number of tourists visiting Haesin drama set for 2005 is 949,331. They expended 37.9 billion won nationwide and 20.5 billion won in Wando. The regional input-output analysis for Wando- gun indicates that total expenditures, 20.5 billion won generated about 29.7 billion won as output impact, 6.9 billion won as income impact and 315 equivalent full-time jobs, respectively. Total expenditures from 2005 to 2008 is 40.5 billion won and 58.7 billion won is expected as output impact, 13.7 billion won as income impact and 621 equivalent full-time jobs. The input-output model indicates that 37.9 billion won was expended nationwide for 2005 generated 68.6 billion won as output impact, 15.5 billion won as income impact, 32.2 billion won as value-added impact and 602 equivalent full-time jobs, respectively. Total expenditures from 2005 to 2008 is 74.8 billion won and 135.4 billion won is expected as output impact, 30.5 billion won as income impact, 63.6 billion won as value-added impact and 1,187 equivalent full-time jobs, respectively. This paper is the first study to visitors' expenditure and economic impacts of TV drama set in Korea and the result is to be a help for the assesment of subsidy by local government to the drama set.

최근 많은 이미지 데이터 셋들은 일반적인 특성을 추출하기 위한 다양한 데이터 클래스와 특징을 가지고 있다. 하지 만 이러한 다양한 데이터 클래스와 특징으로 인해 해당 데이터 셋으로 훈련된 물체 검출 딥러닝 모델은 데이터 특성이 다른 환경에서 좋은 성능을 내지 못하는 단점을 보인다. 이 논문에서는 하위 카테고리 기반 물체 검출 방법과 오픈셋 물체 검출 방법을 이용하여 이를 극복하고, 강인한 물체 검출 딥러닝 모델을 훈련하기 위해 능동 준지도 학습 (Active Semi-Supervised Learning) 을 이용한 다중 분기 트리 구조를 제안한다. 우리는 이 구조를 이용함으로써 데이터 특성이 다른 환경에서 적응 할 수 있는 모델을 가질 수 있고, 나아가 이 모델을 이용하여 이전의 모델보다 높은 성능을 확보 할 수 있다.

Recently there are many image datasets which has variety of data class and point to extract general features. But in order to this variety data class and point, deep learning model trained this dataset has not good performance in heterogeneous data feature local area. In this paper, we propose the structure which use sub-category and openset object detection methods to train more robust model, named multi-branch tree using ASSL. By using this structure, we can have more robust object detection deep learning model in heterogeneous data feature environment.

Background and Objectives：The aims of this study were 1) to assess the development of open-set speech perception in pre-lingualy deaf children who had multichannel cochlear implants implanted and 2) to document the outcomes according to age at implantation and the device type. Subjects and Method：received a Nucleus 22 or Nucleus 24 multichannel cochlear implants and who had been followed-up for a period of 1-4 years at Dong-A University Hospital. They had no other disabilities. Open-set speech perception was evaluated pre-operatively, postope-ratively for 6 months and then on a yearly basis using the following assesment materials：Phonetically Balanced Kindergarten monosyllabic word lists, bisyllabic word lists, and Glendonald Auditory Screning Procedure(GASP) sentence subtest. Paired t-diferences in the mean scores between two groups divided by age at implantation and the device type. Results：Open-set speech perception for all subjects was improved to mean 84% for monosylabic phonemes and 89% and 88% for bisyllabic phonemes and everyday sentences respectively at 4 years postimplantation. Open-set speech perception ability was better for children who received implantation before 5 years of age than those who did it after 5 years of age at 3 years interval and 4 years interval. Conclusion：The prelingualy deaf children in our study showed significant open-set speech perception abilities at the time of the last postoperative evaluation. Rapid improvement was noted in open-set spech perception for the children who had Nucleus 24M implanted before 5 years of age. In order to maximize the effects of access to auditory stimuli provided by the cochlear implant on the development of open-set speech perception, a structured postoperative schedule of programing and rehabili-tation is recomended in addition to the optimal educational environment.

This paper gives an extensive study of ideal topological space and introduce two new types of set with the help of local function. Several characterizations of these sets will also be discussed through this paper and finally gives new representation of ${\alpha}$-sets and semi-open sets.

Background and Objectives：In this study, we evaluated the effect of bimodal hearing in the speech perception test and the increasing level of bimodal hearing over cochlear implantation in speech perception score regarding residual hearing. Subjects and Method：Nineteen prelingually deaf patients, who had used bimodal hearing over a period of 8 months, were divided in two groups in accordance to their low frequency residual hearing. The children were tested in open-set speech perception under +10 SNR and +20 SNR. The scores of bimodal hearing and unilateral cochlear implantation were compared by the Wilcoxon signed-rank test. Also, the increased level of speech perception scores of bimodal hearing over cochlear implantation alone were calculated using the formula, [bimodal score (%) - unilateral cochlear implantation (%)], and was compared between groups. Results：In pure tone audiometry under silent condition, the result of bimodal hearing was similar to unilateral cochlear implantation in magnitude in both groups. Under the noisy condition, both groups had better result in bimodal hearing in terms of speech perception. Also, the increased level was higher in bimodal hearing over unilateral cochlear implantation in the better residual hearing group. However, since the sample size was small enough, it is considered that there was no statistical significance. Conclusion：Cochlear implanted patients with residual hearing are expected to get better speech perception in noisy environment with bimodal hearing regardless of the level of residual hearing.

Because of difficulty of using Schauder's fixed point theorem to the polynomial-like iterative equation, a lots of work are contributed to the existence of solutions for the polynomial-like iterative equation on compact set. In this paper, by applying the Schauder-Tychonoff fixed point theorem we discuss monotone solutions and convex solutions of the polynomial-like iterative equation on an open set (possibly unbounded) in $\mathbb{R}^N$. More concretely, by considering a partial order in $\mathbb{R}^N$ defined by an order cone, we prove the existence of increasing and decreasing solutions of the polynomial-like iterative equation on an open set and further obtain the conditions under which the solutions are convex in the order.

오픈소스 커뮤니티들은 개발자들 간의 커뮤니케이션을 활성화시키고, 협업 할 수 있도록 환경을 제공하고 있다. 하지만 스마트TV 셋톱박스 요소 기술과 관련된 오픈소스프로젝트는 극히 미비한 수준이다. 이에 본 연구에서는 스마트TV 요소기술과 상위 6개의 오픈소스 커뮤니티의 기능을 조사, 분석하였으며, 이를 기반으로 스마트TV 셋톱박스 기반의 오픈소스 커뮤니티를 설계하였다. 본 논문에서 제안하는 스마트TV 기반의 오픈소스 커뮤니티는 스마트TV 셋톱박스 기반의 오픈 소프트웨어의 고도화를 가지고 올 수 있는 개발 환경을 제공 할 수 있을 것이다.

오픈소스 커뮤니티들은 개발자들 간의 커뮤니케이션을 활성화시키고, 협업 할 수 있도록 환경을 제공하고 있다. 하지만 스마트TV 셋톱박스 요소 기술과 관련된 오픈소스프로젝트는 극히 미비한 수준이다. 이에 본 연구에서는 스마트TV 요소기술과 상위 6개의 오픈소스 커뮤니티의 기능을 조사, 분석하였으며, 이를 기반으로 스마트TV 셋톱박스 기반의 오픈소스 커뮤니티를 설계하였다. 본 논문에서 제안하는 스마트TV 기반의 오픈소스 커뮤니티는 스마트TV 셋톱박스 기반의 오픈 소프트웨어의 고도화를 가지고 올 수 있는 개발 환경을 제공 할 수 있을 것이다.

An open-end winding permanent magnet synchronous motor (OEW-PMSM) fed by dual inverters has good application value in the drive system of electric vehicles due to the motor's advantages, such as high output power, wide speed range and excellent fault-tolerant ability. However, the symmetry of dual-inverter topology causes difficulty in identifying the specific faulty switch when an open-circuit fault occurs. Thus, this study proposes a switch open-circuit fault diagnosis strategy for OEW-PMSM based on finite-control-set model predictive control (FCS-MPC). With the known and unchanged switching state in each control period of FCS-MPC, the proposed strategy uses the predicted switching state to predict the phase voltage. According to the error of predicted voltage and measured voltage under switch failure, the faulty phase and faulty switch pair are identified. Finally, the relationship between switching states and error voltage is analyzed, and a diagnosis function is constructed to identify the specific faulty switch. The validity of the proposed diagnosis strategy is confirmed by the simulation and experimental results.

Three highbush blueberries (Vaccinium corymbosum cvs. Bluecrop, Coville, and Northland) grown in an open field and in a rain shelter were compared to evaluate the potential benefits of rain-sheltering systems on growth and fruit quality. Shrubs grown in rain shelter flowered 5-14 days earlier and were in full bloom earlier than those in the open field. All three cultivars set fruit by nearly 90% of flowers when they were grown in the open field, but was markedly reduced when they were grown in the rain shelter, to approximately 50% for 'Bluecrop' and 'Coville'. Fruit sets ratio of 'Northland' grown in the in the rain shelter was 14.5. Fruit from greenhouse-grown 'Northland' were larger, heavier, and had more seeds per berry. Soluble sugar content (SSC) of fruits varied both with the cultivars and growing conditions. SSC of fruits grown in the greenhouse was higher in 'Bluecrop' compared to that grown in the field, but this pattern was reversed in 'Northland'. Titratable acid was significantly higher in fruits from rain shelter-grown shrubs of both 'Bluecrop' and 'Northland' than in fruits from open field-grown shrubs (P < 0.05), although there was no difference in TA with respect to growing conditions for 'Coville'. The number of seed per fruit, fruit length and diameter, weight, SSC, and TA were similar in 'Coville' shrubs grown in the open field and the rain shelter. Fruit firmness of 'Bluecrop', 'Coville', and 'Northland' was higher in the rain shelter than in the open field. 'Coville' and 'Northland' fruits harvested in the open field had higher blue chroma than those harvested in the rain shelter. However, this pattern was reversed for 'Bluecrop'. On the whole, sheltering from rain affected most fruit characteristics of the three cultivars differently. Therefore, suitable blueberry cultivars for the rain shelter should be chosen by purpose.

Background and Objectives：An open-set monosyllabic speech perception test is necessary, not only to devise habilitation programs but to select candidates and estimate performance in pediatric cochlear implant program. The purpose of this study is to develop a validated open-set monosyllabic speech perception test for preschool children. Subjects and Method：Qualitative test development was based on vocabulary familiarity, syllabic and phonemic frequency, and phonetic balancing between the lists. The devised test lists were applied to 138 normal hearing children and 46 children with cochlear implant to examine the validity and reliability. Results：The final two test lists were made up of 100 items with 260 phonemics. The consequential validity of this devised test was proven by the applied results that the speech perception score of children with implanted devices was significantly lower than that of normal children, when controlling for the length of auditory experience. Conclusion：The developed monosyllabic speech perception test is the validated instrument for preschool children. It is sensitive to perceptual characteristics of articulatory function in cochlear implanted children. The validity of this test needs to be enhanced by clinical application and further item analysis.

This study aims to investigate the development of open-set speech perception abilities in children who received cochlear implantation at older age and to examine the preoperative variables contributing to the postoperative speech perception outcomes. Subjects and Method：Open-set speech perception abilities were assessed with the Monosyllabic Word and Common Phrases Tests for 37 prelingually deafened children who received implantation between the ages of 7 and 17 years. Their performance was compared to that of 60 children who received implantation before 7 years of age. To identify the factors contributing to the postoperative open-set speech perception abilities, we analyzed age at hearing aid fitting, age at implantation, duration of implant use, preoperative language and speech perception abilities, and pure tone thresholds in the better ear. Results： The children who received implantation after 7 years of age significantly improved during the first year of implant use, but reached a plateau by 1.5 to 2 years after implantation, reaching scores between 60% and 70%. Children who received implantation after age 7 showed higher speech perception scores than the younger children over the first year of implant use, but showed lower scores than the younger children after 2-3 years of implant use. The variables associated with the postoperative open-set speech perception abilities were speech perception ability, language ability, and hearing levels before implantation. Conclusion：Prelingually deafened children who received implantation at older age can obtain substantial open-set speech perception understanding, but tended to reach a plateau between 1.5-2 years after implantation. Speech perception, language abilities, and hearing level before implantation contribute to the outcome. (Korean J Otorhinolaryngol-Head Neck Surg 2008;51:130-5)

‘집합 제한 논리 언어’는 ‘집합 이론’을 프로그래밍에 도입한 언어이다. 본 논문은 A. Dovier 연구팀이 제안한 집합 제한 문제 풀이(solver) 절차를 소개하고, 이 절차가 논리 언어 Prolog 상에서 어떻게 구현 가능한 지를 보인다. 이 절차는 ‘다시쓰기 규칙(rewrite rule)’으로 표현되어 있는데 이 표현의 특징은 일반 프로그래밍 언어가 표현하기 힘든 비결정적 규칙 적용(nondeterministic rule application)과 수학적 변수(mathematical variable)를 사용한다는 점이다. 본 연구에서는 이들 특징이 Prolog 언어에서 제공되는 비결정적 제어 (nondeterministic control), 논리적 변수(logical variable) 및 리스트(list) 자료구조의 사용으로 쉽게 구현 가능함을 보인다. 본 연구의 구현은 다음과 같은 의의를 가지고 있다. 첫째 본연구는 이 언어의 모든 기능을 완전하게 구현하였다는 점이다. 둘째 본 연구에서는 이 언어의 구현 방법을 누구나 알기 쉽게 기술하였다는 점이다. 셋째 기존의 구현이 상업적 Prolog인 SICStus Prolog를 사용하여 구현한 것과는 달리 본 구현은GNU GPL(General Public License)을 가지는CIAO Prolog를 사용하여 구현하였기 때문에 누구나 자유롭게 사용할 수 있는 점이다. 넷째 본 연구에서 개발된 소스 코드는 공개 소프트웨어이기 때문에 누구나 자유롭게 사용, 수정 및 배포할 수 있다는 점이다.

Set constraints logic language is a language that adopts ‘set theory’ in programming. In this paper, we introduce the procedure for solving set constraints proposed by A. Dovier and show how the procedure can be implemented in logic language Prolog. The procedure is represented in ‘rewriting rules’ and this representation is characterized by having nondeterministic rule applicationsand mathematical variables that is difficult to be implemented in general programming languages. In this paper, we show that the representation can be easily implemented by using nondeterministic control, logical variables and data structure ‘list’ provided in Prolog. Our implementation has following advantages.First we have implemented the full features of the language. Second we have described the implementation detail in thisresearch. Third other used the commercial Prolog called SICStus, but we are using CIAO Prolog with GNU GPL(General Public License) and anyone can use it freely. Forth the software of our implementation is open source so anyone can use, modify, and distribute it freely.

최근 사회적 이슈로 부각되고 있는 클라우드 컴퓨팅, 빅데이터 시스템 등은 대부분 공개소프트웨어를 활용하여 개발이 진행되고 있으며, 이에 따라 공개소프트웨어의 활용이 급격히 증가하고 있는 추세이다. 그러나 저비용, 고효율성의 장점을 가진 공개소프트웨어의 성공적인 활용을 위해서는 공개소프트웨어 라이선스를 체계적으로 관리하고, 제품출시 이전에 라이선스 위반 여부를 명확히 확인하여야 향후 법정 분쟁으로부터 자유로울 수 있다는 이슈를 가지고 있다. 본 논문에서는 공개소프트웨어 라이선스에 대한 명확한 이해를 통해 안전한 공개소프트웨어 활용을 위한 공개소프트웨어 라이선스 설정방안을 제시하고자 한다.

최근 사회적 이슈로 부각되고 있는 클라우드 컴퓨팅, 빅데이터 시스템 등은 대부분 공개소프트웨어를 활용하여 개발이 진행되고 있으며, 이에 따라 공개소프트웨어의 활용이 급격히 증가하고 있는 추세이다. 그러나 저비용, 고효율성의 장점을 가진 공개소프트웨어의 성공적인 활용을 위해서는 공개소프트웨어 라이선스를 체계적으로 관리하고, 제품출시 이전에 라이선스 위반 여부를 명확히 확인하여야 향후 법정 분쟁으로부터 자유로울 수 있다는 이슈를 가지고 있다. 본 논문에서는 공개소프트웨어 라이선스에 대한 명확한 이해를 통해 안전한 공개소프트웨어 활용을 위한 공개소프트웨어 라이선스 설정방안을 제시하고자 한다.