• 정보과학회 컴퓨팅의 실제 논문지
• /
• 제24권3호
• /
• pp.138-144
• /
• 2018

본 고에서는 비디오로부터 coherent story를 학습하여 비디오 스토리를 재현할 수 있는 스토리 학습/재현 프레임워크를 제안한다. 이를 위해 연속 이벤트 순서를 감독학습 정보로 사용함으로써 각 에피소드들이 은닉 공간 상에서 궤적 형태를 가지도록 유도하여, 순서정보와 의미정보를 함께 다룰 수 있는 복합된 표현 공간을 구축하고자 한다. 이를 위해 유아용 비디오 시리즈를 학습데이터로 활용하였다. 이는 이야기 구성의 특성, 내러티브 순서, 복잡도 면에서 여러 장점이 있다. 여기에 연속 이벤트 임베딩을 반영한 인코더-디코더 구조를 구축하고, 은닉 공간 상의 시퀀스의 모델링에 양방향 LSTM을 학습시키되 여러 스텝의 서열 데이터 생성을 고려하였다. '뽀롱뽀롱 뽀로로' 시리즈 비디오로부터 추출된 약 200 개의 에피소드를 이용하여 실험결과를 보였다. 실험을 통해 에피소드들이 은닉공간에서 궤적 형태를 갖는 것과 일부 큐가 주어졌을 때 스토리를 재현하는 문제에 적용할 수 있음을 보였다.

• International Journal of Aeronautical and Space Sciences
• /
• 제10권1호
• /
• pp.91-97
• /
• 2009

It is undertaken systematization of results of satellite and ground observation parameters characterizing a current condition and climatic variability of two selected geographical areas. One of them covers territory of Azerbaijan and another covers a wide area of Caspian See region. Average values and mean square deviations of following values are investigated: outgoing long wave radiation during a day and night (in nebulosity and cloudless). absorbed within a day of the stream of a sunlight of the system in "a terrestrial surface-atmosphere". degree of a covering by clouds of the selected areas during a day and at night, ground temperature values of air. pressure and speed of a wind. Monthly average values of corresponding parameters create a basis of suggested investigations. It has been presented features of a time course of investigated parameters for each month and year in the whole due to the continuously observations since 1982-2000. The scientific problem consists that there are no existed models which authentically would be cover the main aspects of a realities specified changes: they are identified by economic activities. growth of the population and other features of development of a human society or internal fluctuations of biogeophysical/climatic system. Possibilities of predictability of biosphere and climate changes depend on available timely supervision. adequacy of construction of appropriate models. understanding of mechanisms of direct and feedback influences in such complicated systems.

• 교육시설
• /
• 제11권1호
• /
• pp.39-46
• /
• 2004

The objective of this study is to propose the architectural basic data for the planning of kindergarten attached to the elementary school. For this study, we selected and investigated 23kindergartens in Gwangju city. The results of this study are as follows. 1) The elementary school in separated style had the advantage of securing educational space and reducing interference with students of elementary school. But it had the problem in using the facilities of elementary school. So we have to consider the kindergarten to be located near by dinning room, multi-purpose hall. 2) Concentrating the entrance, it makes the inside and outside space use easily. For the actual use, we have to establish facilities for convenience in outer-space. 3) For each room, we suggest guides as follow. The classroom and the playroom have to be combined for flexibility and supervision. In the planning of toilet, it must have sliding door for confirm the inside situation and reasonable height partition to consider the student's body-size. The shape of teacher's room must be open-type for supervising children but protect the entry of children. There is consideration for the material room and the sleeping room for all-day-long class.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권2호
• /
• pp.365-380
• /
• 2022

The target detection algorithm based on supervised learning is the current mainstream algorithm for target detection. A high-quality dataset is the prerequisite for the target detection algorithm to obtain good detection performance. The larger the number and quality of the dataset, the stronger the generalization ability of the model, that is, the dataset determines the upper limit of the model learning. The convolutional neural network optimizes the network parameters in a strong supervision method. The error is calculated by comparing the predicted frame with the manually labeled real frame, and then the error is passed into the network for continuous optimization. Strongly supervised learning mainly relies on a large number of images as models for continuous learning, so the number and quality of images directly affect the results of learning. This paper proposes a dataset STAR-24K (meaning a dataset for Space TArget Recognition with more than 24,000 images) for detecting common targets in space. Since there is currently no publicly available dataset for space target detection, we extracted some pictures from a series of channels such as pictures and videos released by the official websites of NASA (National Aeronautics and Space Administration) and ESA (The European Space Agency) and expanded them to 24,451 pictures. We evaluate popular object detection algorithms to build a benchmark. Our STAR-24K dataset is publicly available at https://github.com/Zzz-zcy/STAR-24K.

• 복식문화연구
• /
• 제24권2호
• /
• pp.132-149
• /
• 2016

With the increased importance of design creativity in global fashion marketplaces, there have arisen numerous promotional strategies for new fashion designers in Korea, yet little research has been done on this subject. This study examines the new fashion design promotion policies in Korea and their effects from the designer's perspective. First, we analyzed the contents of twenty-one strategies from thirteen organizations regarding the types of support from the strategies. As a result, four distinct types are identified: new designer recognition; information and consulting support; brand incubating; and marketing support. Next, we performed in-depth one-on-one interviews with eleven designers who had been awarded from one or more of the fashion design incubation policies. The results reveal that the incubating policies' cash rewards and space support are found to be most effective. In terms of the revised fashion and brand consulting policy, "brand operation" had an influence. In addition, fashion the new brand advertisement and marketing policy had positively influenced the "brand development." However, throughout the interview, designers note that the quality of education and information provisions need some improvement. Furthermore, the designers indicate the lack of supervision and professionalism from the marketing departments. Based on this content, this study emphasizes the necessity of an effective fashion design promotion policy, and strategic and consistent support for the whole fashion promotion team. At the minimum, an integrated and united supervision by the organization is necessary and should be taken into consideration.

• 한국콘텐츠학회논문지
• /
• 제17권9호
• /
• pp.619-636
• /
• 2017

본 연구는 수퍼비전을 통한 수퍼바이저와 수퍼바이지의 관계형성에 대한 내러티브적 묘사이다. 즉, 수퍼바이저 A의 삶과 수퍼바이지 B의 삶 속에서 맺고 있는 그들의 관계를 있는 그대로 기술 및 설명하였다. 본 연구로부터 얻은 결론은 진솔성과 개방성은 인간중심 상담의 핵심요소이며, 특히 수파바이저의 무조건적인 수용에 대한 노력은 수직관계 형성으로 '은폐' 방법을 사용하기 쉬운 수퍼바이지의 태도를 더욱 개방적으로 유도할 수 있었다. 이것은 수퍼비전의 관계는 상하관계라는 기존의 시각에 새로운 관점을 제시한다. 수퍼비전에 초보 수퍼바이지들을 적극적으로 참여 할 수 있도록 수퍼바이저들이 먼저 수용하고 신뢰할 수 있는 분위기와 라포(rapport)를 형성한다면, 수퍼비전의 관계는 수직에서 수평으로 더욱 진전될 수 있음을 본 연구를 통하여 알 수 있었다.

• Journal of Electrical Engineering and Technology
• /
• 제11권6호
• /
• pp.1872-1879
• /
• 2016

The techniques to recycle and reuse plastics attract public attention. These public attraction and needs result in improving the recycling technique. However, the identification technique for black plastic wastes still have big problem that the spectrum extracted from near infrared radiation spectroscopy is not clear and is contaminated by noise. To overcome this problem, we apply Raman spectroscopy to extract a clear spectrum of plastic material. In addition, to improve the classification ability of fuzzy Radial Basis Function Neural Networks, we apply supervised learning based clustering method instead of unsupervised clustering method. The conditional fuzzy C-Means clustering method, which is a kind of supervised learning based clustering algorithms, is used to determine the location of radial basis functions. The conditional fuzzy C-Means clustering analyzes the data distribution over input space under the supervision of auxiliary information. The auxiliary information is defined by using k Nearest Neighbor approach.

• Genomics & Informatics
• /
• 제18권2호
• /
• pp.20.1-20.4
• /
• 2020

Accessible negative results are relevant for researchers and clinicians not only to limit their search space but also to prevent the costly re-exploration of research hypotheses. However, most biomedical relation extraction datasets do not seek to distinguish between a false and a negative relation among two biomedical entities. Furthermore, datasets created using distant supervision techniques also have some false negative relations that constitute undocumented/ unknown relations (missing from a knowledge base). We propose to improve the distinction between these concepts, by revising a subset of the relations marked as false on the phenotype-gene relations corpus and give the first steps to automatically distinguish between the false (F), negative (N), and unknown (U) results. Our work resulted in a sample of 127 manually annotated FNU relations and a weighted-F1 of 0.5609 for their automatic distinction. This work was developed during the 6th Biomedical Linked Annotation Hackathon (BLAH6).

• 조명전기설비학회논문지
• /
• 제13권2호
• /
• pp.32-38
• /
• 1999

운송설비용 자율 주행 운반체는 인간의 지도없이 주어진 환경 내에서 장애물과 충돌을 회피하며 효율적으로 목표지점까지 주행할 수 있는 최적의 이동 경로를 생성해야 한다. 본 논문에서는 장애물과 충돌을 회피하는 전역 및 지역경로를 유전알고리즘을 이용하여 계획하였다. 본 논문에서는 제안한 운송설비용 자율 주행 운반체의 충돌회피 알고리즘은 전통적인 충돌회피 알고리즘에 비해 능률적임을 모의 실험을 통해 확인하였다.

• 한국정보전자통신기술학회논문지
• /
• 제3권3호
• /
• pp.62-72
• /
• 2010

As the applications for Mobile Ad-hoc NETworks (MANETs) have varied, performance analysis has become one of the main research areas. They commonly offer only simple radio propagation models that neglect obstacles of a propagation environment. The radio wave propagation model has a strong impact on the results of the simulation run. In this paper we present the new experimental results of the impacts of the various propagation models on MANETs' performance. Intensive simulations have been presented using the group mobility which models typical ad-hoc situations such as military movements or disaster recovery activities under the supervision of a group leader. Comparisons of conventional simple models with more complicated models, i.e., shadowing, Raleigh, and Ricean models, show that, in spite of the models' popularity, the free space and two-ray ground models are too optimistic in describing real ad-hoc group mobility situations.