• Proceedings of the Korean Information Science Society Conference
• /
• 2006.06b
• /
• pp.301-303
• /
• 2006

본 논문은 중립 표정과 같은 표정 측정의 기준이 되는 단서 없이 다양한 내적상태 안에서 얼굴표정을 인식할 수 있는 개선된 시스템을 제안한다. 표정정보를 추출하기 위한 전처리작업으로, 백색화(whitening) 단계가 적용되었다. 백색화 단계는 영상데이터들의 평균값이 0이며, 단위분산값으로 균일한 분포를 갖도록 하여 조명 변화에 대한 민감도를 줄인다. 백색화 단계 수행 후 제 1 주성분이 제외된 나머지 주성분들로 이루어진 PCA표상을 표정정보로 사용함으로써 중립 표정에 대한 단서 없이 얼굴표정의 특징추출을 가능하게 하였다. 본 실험 결과는 83개의 내적상태와 일치되는 다양한 얼굴표정들에서 임의로 선택된 표정영상들의 얼굴표정 인식을 수행함으로써 다양하고 자연스런 얼굴 표정인식을 가능하게 하였다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2004.11a
• /
• pp.571-574
• /
• 2004

디지털화 된 멀티미디어 데이터가 증가함에 따라 데이터의 효과적인 분류 및 검색 작업이 더욱 중요해 지고 있으며, 네트워크와 하드웨어의 발전으로 영상 정보를 검색함에 있어 기존의 웹상의 클라이언트/서버 기반 검색만으로는 부족하게 되었다. 이에 따라 영상 정보를 웹상에서 뿐만 아니라 모바일 같은 이기종간의 검색이 필요하다. 그러나 이기종 간의 플랫폼에서 동일한 정보를 얻기 위해서는 상호운용성에 문제가 있다. 웹 서비스(Web Service)는 분산 애플리케이션에 대한 언어-중립적(language-neutral)과 벤더-중립적(vendor-neutral)[5]을 제시하고 있어 기존의 서로 다른 플랫폼간의 정보 이전 문제를 해결할 수 있다. 따라서 본 논문에서는 영화 컨텐츠 검색 시스템을 웹 서비스화 하였다.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2008.11a
• /
• pp.35-36
• /
• 2008

• Journal of Digital Convergence
• /
• v.10 no.1
• /
• pp.49-57
• /
• 2012

The Internet succeeded because of the end-to-end principle which allowed anyone to add functionality to the network. However, as the internet is increasingly becoming the platform for smart IT applications such as VoIP, IPTV, Cloud Computing and Smart Phone, networks are now under increasing strain of traffic congestion and the absence of quality of service insurances. To date, the debate over internet rules has focused on network neutrality rather than network efficiency. This article emphasizes the well-functioning role of market mechanism for the efficient use and further development of the network. To maximize the value of the network, this article proposes a differential treatment to packets based on customer types, and a two-part tariff pricing rule to secure funding to expand and upgrade networks.

• Journal of the Society of Naval Architects of Korea
• /
• v.50 no.5
• /
• pp.314-323
• /
• 2013

Production simulation technology is beneficial to solve the complicated and fluctuated problems in a shipyard. It takes too much time and effort to build simulation models in the field, though. This research proposes a feasible method to reduce the difficulties related to simulation modeling for the factory or shop capacity analysis. In addition, a proposed neutral data format for production information is efficient to manage information acquisition for simulation modeling automation. A panel block shop model is contributed to comparison between the conventional technique and the automated one. The automation technique is highly recommended to run a rapid simulation in the shipyard problem.

• Journal of the Society of Naval Architects of Korea
• /
• v.54 no.4
• /
• pp.344-351
• /
• 2017

CAD systems used in various manufacturing industries have been rapidly switched from 2D to 3D. 3D-based design is becoming common in the shipbuilding industries too. However, there is still a demand for 2D drawings for ship production. So, it is necessary to continually improve 2D drawings. One way to improve 2D drawings is to add production information to the drawings. To do this, it is necessary to define the relation between the geometry information and the production information. Various types of drawings, which is inevitable because each drawing has its own purpose, make it difficult to define relations. This process can be simplified if the drawing information is neutralized. XML is a standard format for defining neutral data, and is already being used in various fields. In this study, a neutral augmented drawing has been defined by neutralizing the drawing information based on XML, and implemented for the allocation of their information with high quality of readability as well.

• Journal of Internet Computing and Services
• /
• v.24 no.2
• /
• pp.47-56
• /
• 2023

Hydrogen energy is an eco-friendly energy that produces heat and electricity with high energy efficiency and does not emit harmful substances such as greenhouse gases and fine dust. In particular, smart hydrogen energy is an economical, sustainable, and safe future smart hydrogen energy service, which means a service that stably operates based on 'data' by digitally integrating hydrogen energy infrastructure. In this paper, in order to implement a data-based hydrogen charging station demand forecasting model, three hydrogen charging stations (Chuncheon, Sokcho, Pyeongchang) installed in Gangwon-do were selected, supply and demand data of hydrogen charging stations were secured, and 7 machine learning and deep learning algorithms were used. was selected to learn a model with a total of 27 types of input data (weather data + demand for hydrogen charging stations), and the model was evaluated with root mean square error (RMSE). Through this, this paper proposes a machine learning-based hydrogen charging station energy demand prediction model for optimal hydrogen energy supply and demand.

• Annual Conference on Human and Language Technology
• /
• 2020.10a
• /
• pp.292-297
• /
• 2020

자연어 추론 모델은 전제와 가설 사이의 의미 관계를 함의와 모순, 중립 세 가지로 판별한다. 영어에서는 RTE(recognizing textual entailment) 데이터셋과 다양한 NLI(Natural Language Inference) 데이터셋이 이러한 모델을 개발하고 평가하기 위한 벤치마크로 공개되어 있다. 본 연구는 국외의 텍스트 추론 데이터 주석 가이드라인 및 함의 데이터를 언어학적으로 분석한 결과와 함의 및 모순 관계에 대한 의미론적 연구의 토대 위에서 한국어 자연어 추론 벤치마크 데이터 구축 방법론을 탐구한다. 함의 및 모순 관계를 주석하기 위하여 각각의 의미 관계와 관련된 언어 현상을 정의하고 가설을 생성하는 방안에 대하여 제시하며 이를 바탕으로 실제 구축될 데이터의 형식과 주석 프로세스에 대해서도 논의한다.

• Journal of Korea Multimedia Society
• /
• v.6 no.1
• /
• pp.1-14
• /
• 2003

E-commerce requires many different types of communications and an unprecedented amount of data changes hands. The many different Platforms and systems interacting require a platform neutral standard for data exchange. One of the technologies that can fill this niche is XML, the extensible markup language established as a standard by the W3C. By being standardized and platform neutral, XML to be factor in e-commerce application and using many systems In this paper, we designed XML document that is used in transaction between corporations, and implement the B2B integration system based on XML database system. Also we use XSLT that make efficient transformation XML documents for exchanging heterogeneous XML data between corporations. So, he or she can more easily and efficiently exchange XML documents between corporations using this system.

• The Journal of the Convergence on Culture Technology
• /
• v.6 no.4
• /
• pp.575-582
• /
• 2020

We proposes a method to estimate political preference of users who write comments on internet news. We collected and analyzed a massive amount of new comment data from internet news to extract features that effectively characterizes political preference of users. We expect that it helps user to obtain unbiased information from internet news and online discussion by providing estimated political stance of news comment writer. Through comprehensive tests we prove the effectiveness of two proposed methods, lexicon-based algorithm and similarity-based algorithm.