• 한국멀티미디어학회논문지
• /
• 제24권3호
• /
• pp.382-395
• /
• 2021

In this study, character recognition using deep learning is performed among the various defects in the PCB, the purpose of which is to check whether the printed characters are printed correctly on top of components, or the incorrect parts are attached. Generally, character recognition may be perceived as not a difficult problem when considering MNIST, but the printed letters on the PCB component data are difficult to collect, and have very high redundancy. So if a deep learning model is trained with original data without any preprocessing, it can lead to over fitting problems. Therefore, this study aims to reduce the redundancy to the smallest dataset that can represent large amounts of data collected in limited production sites, and to create datasets through data enhancement to train a flexible deep learning model can be used in various production sites. Moreover, ResNet model verifies to determine which combination of datasets is the most effective. This study discusses how to reduce and augment data that is constantly occurring in real PCB production lines, and discusses how to select coresets to learn and apply deep learning models in real sites.

• 한국컴퓨터정보학회:학술대회논문집
• /
• 한국컴퓨터정보학회 2013년도 제48차 하계학술발표논문집 21권2호
• /
• pp.31-32
• /
• 2013

본 논문에서는 영화의 등장인물의 성향을 파악하기 위해 시나리오의 대사로부터 감정어를 추출하고, 등장인물의 감정어들을 긍정, 부정, 중립의 3개로 단순화하여 등장인물의 성향을 가시화 시켜주는 방법을 제안한다. 대사로부터 감정어를 추출하기 위해 WordNet 기반의 감정어 추출 방법을 제안한다. WordNet은 단어 간에 상위어와 하위어, 유사어 등의 관계로 연결된 네트워크 구조의 사전이다. 이 네트워크 구조에서 최상위의 감정 항목과의 거리를 계산하여 단어별 감정량을 계산하여 대사를 30 차원의 감정 벡터로 표현한다. 등장인물별로 추출된 감정 벡터를 긍정, 부정, 중립의 3개의 차원으로 단순화 하여 등장인물의 성향을 표현한다.

• 한국농촌건축학회논문집
• /
• 제1권1호
• /
• pp.77-86
• /
• 1999

The object of this study is to look for how to develop the style of residence formed spontaneously on the steep area by looking into the street-network and resident character in residence area and analysing fundamental style of residence. Specially, It is not easy to find out a study of a sea-village in the steep area. So, it is urgent to be developed a road-system and a style of residence which are able to be adapted to special circumstances and situation of ground. Therefore, th found out characters that a steep area contains, contents that are analyzed in this study are as follows : First, the order and the style of road-net formed to be adapted to circumstances of ground. Second, the style of road which is formed by directions. Third, the style of alley. Fourth, the relations with which alley and Madang are confronted each other.

• 반도체디스플레이기술학회지
• /
• 제20권3호
• /
• pp.6-10
• /
• 2021

Recognition of characters printed or marked on the PCB components from images captured using cameras is an important task in PCB components inspection systems. Previous optical character recognition (OCR) of PCB components typically consists of two stages: character segmentation and classification of each segmented character. However, character segmentation often fails due to corrupted characters, low image contrast, etc. Thus, OCR without character segmentation is desirable and increasingly used via deep neural networks. Typical implementation based on deep neural nets without character segmentation includes convolutional neural network followed by recurrent neural network (RNN). However, one disadvantage of this approach is slow execution due to RNN layers. LPRNet is a segmentation-free character recognition network with excellent accuracy proved in license plate recognition. LPRNet uses a wide convolution instead of RNN, thus enabling fast inference. In this paper, LPRNet was adapted for recognizing characters printed on PCB components with fast execution and high accuracy. Initial training with synthetic images followed by fine-tuning on real text images yielded accurate recognition. This net can be further optimized on Intel CPU using OpenVINO tool kit. The optimized version of the network can be run in real-time faster than even GPU.

• International Journal of Advanced Culture Technology
• /
• 제11권2호
• /
• pp.403-408
• /
• 2023

The widespread discussion of AI-generated content, fueled by the emergence of consumer applications like ChatGPT and Midjourney, has attracted significant attention. Among various AI applications, AI painting has gained popularity due to its mature technology, user-friendly nature, and excellent output quality, resulting in a rapid growth in user numbers. Midjourney and Stable Diffusion are two of the most widely used AI painting tools by users. In this study, the author adopts a perspective that represents the general public and utilizes case studies and comparative analysis to summarize the distinctive features and differences between Midjourney and Stable Diffusion in the context of AI character illustration. The aim is to provide informative material forthose interested in AI painting and lay a solid foundation for further in-depth research on AI-generated content. The research findings indicate that both software can generate excellent character images but with distinct features.

• 한국정보과학회논문지:소프트웨어및응용
• /
• 제35권8호
• /
• pp.512-521
• /
• 2008

웹 기반 정보 서비스는, 웹 정보 서버와 상호작용하는, 다양한 인터넷 웹-브라우징 기기들과의 접근성과 사용성을 위해서 평가되어야 한다. 다양한 웹-브라우징 장치로(예: 풀부라우징 휴대폰) 접근 및 사용가능하며, 신뢰할 수 있는 유비쿼터스 웹 정보서버는 유비쿼터스 웹 정보 서비스뿐만 아니라 개인화된 광고기반의 비즈니스 모델을 위한 통합센터가 되어야 한다. 개인화된 유비쿼터스 웹 정보 서비스를 위한 웹 상호작용의 접근성과 사용성을, 실시간 평가를 위한 메트릭으로서, 연구하였다. 시험용 웹사이트 ('ktrip.net')와 1자 한글 도메인명(예: 김.net, 이.net, 박.net, 최.net, ㄱ.net, ㄴ.net ... ㅎ.net, ㅏ.net ... ㅔ.net, ㄱ.com, ㄴ.com ... ㅎ.com)을 사용하여, 구현 및 한국, 일본 및 중국에서의 실험을 바탕으로 한 실험적 결과를 소개한다.

• 디지털산업정보학회논문지
• /
• 제13권2호
• /
• pp.87-98
• /
• 2017

The advancements in scientific technology and information network in the 21st century allow us to easily acquire a desired knowledge. In the midst of today's informatization, globalization, and cultural diversification, adolescents experience emotional confusion while accommodating diverse cultures and information. This study aimed at examining three aspects of character suggested by the Ministry of Education, which are ethics, sociality, and emotion, and the actual sub-factors required for character education. To that end, a survey was conducted with adolescents who were at a character-building age, and social network analysis (SNA) was performed to determine the effect of character education on the sub-factors. The statistics program SPSS was used to investigate the general traits of the subjects and the validity of the research variables. The 2-mode data that were finally selected were converted to 2-mode data using NetMinder 4, which is a network analysis tool. Furthermore, a data network was established based on a quasi-network that represents the relationships between ethics, sociality, and emotion. The results of this study showed that the subjects considered honesty and justice to be the sub-domains of the ethics domain. In addition, they identified sympathy, communication, consideration for others, and cooperation as the sub-domains of the sociality domain. Finally, they believed that self-understanding and self-control were the sub-domains of the emotion domain.

• 정보처리학회논문지B
• /
• 제15B권3호
• /
• pp.219-228
• /
• 2008

실시간 소셜(Social) 네트워킹 서비스를 위해서, 사용 가능한-통합적-유비쿼터스 (U3) 웹 서비스에서의 모바일 상호작용을 연구하였다. 실시간 소셜 네트워크를 위한 편리한 모바일 HCI로서, 또한 유비쿼터스 웹 서비스에서의 메타데이터 정보 색인 키로서, 다국어 한 글자 도메인 명(예, 김.net, 이.net, 가.net, ㄱ.net, ㄴ.net, ㅎ.net, ㅏ.net, ㅔ.net, ㄱ.com, ㅎ.com) 은 소셜 정보를 검색하고 또한 등록할 때 편리한 모바일 인터페이스이다. U3 웹 서비스의 예로서, 실시간 소셜 네트워킹 서비스 구현과 함께 모바일 상호작용을 위해 스케치된 디자인 목표와 한국, 일본, 중국에서의 모바일 상호작용 경험을 소개한다. 또한, 소셜 네트워킹과 센서 네트워킹 서비스에서 통합 정보관리를 위해, IP 기반의 센서네트워크 (IP-USN)에서 메타데이터 디렉터리 서비스에의 응용 확장 가능성도 소개한다.

• Journal of Plant Biology
• /
• 제39권2호
• /
• pp.149-159
• /
• 1996

The order of Rhododendro-Quercetalia mongolicae representing temperate oak forests in Korea was characterized in terms of presence of taxa of different rank. 140 releves were analysed for percentage contribution of each species, genera and families in syntaxa under consideration as well as extracton of diagnostic genera and families for syntaxa by an ordination technique. The Rhododendro-Quercetalia is characterized by high diversity of tree and shrub species contributing 40% of the total floristic composition as well as by a high contribution of the genus Rhododendron and the absence of the genus Fagus, characteristic of the QuercoFagetea sensu lato. The character families for the Rhododendron and Quercus include Liliaceae and Compositae, whereas Acer, Carex, Viola, Rhododendron and Quercus are the most common among genera. Rhododendron and Quercus are regarded as the transgressive cheractergenera, whereas the families of Pinaceae and Ericaceae are considered companions (in sense of the terminology of the Braun-Blanquet syntaxonomy) for the order. Family appeared to be an inadequate rank for diagnoses of alliances and suballiances. On the other hand, genus was found to be the most effective rank in differentiating the alliances and suballiances. The Lindero-Quercenion shares the same character-genera with the order Rhododendro-Quercetalia. Character genera of the suballiances Callicarpo-quercenion are Carpinus, Styrax, Smilax and Callicarpa, and those of the Pino-Quercion list Euonymus, Saussurea and Tilia.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2003년도 하계종합학술대회 논문집 Ⅲ
• /
• pp.1363-1366
• /
• 2003

This paper introduces computerized version of Hangout(Korean Language) Standard Pronunciation Rule that can be used in Korean processing systems such as Korean voice synthesis system and Korean voice recognition system. For this purpose, we build Petri net models for each items of the Standard Pronunciation Rule, and then integrate them into the vocal sound conversion table. The reversion of Hangul Standard Pronunciation Rule regulates the way of matching vocal sounds into grammatically correct written characters. This paper presents not only the vocal sound conversion table but also character conversion table obtained by reversely converting the vocal sound conversion table. Making use of these tables, we have implemented a Hangeul character into a vocal sound system and a Korean vocal sound into character conversion system, and tested them with various data sets reflecting all the items of the Standard Pronunciation Rule to verify the soundness and completeness of our tables. The test results shows that the tables improves the process speed in addition to the soundness and completeness.