• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2022.06a
• /
• pp.1311-1314
• /
• 2022

DNN을 사용하여 객체 인식 과정에서 객체를 잘 분류하기 위해서는 시각적 설명성이 요구된다. 시각적 설명성은 object class에 대한 예측을 pixel-wise attribution으로 표현해 예측 근거를 해석하기 위해 제안되었다, Scale-invariant한 특징을 제공하도록 설계된 pyramidal features 기반 backbone 구조는 object detection 및 classification 등에서 널리 쓰이고 있으며, 이러한 특징을 갖는 feature pyramid를 trainable attention mechanism에 적용하고자 할 때 계산량 및 메모리의 복잡도가 증가하는 문제가 있다. 본 논문에서는 일반적인 FPN에서 객체 인식 성능과 설명성을 높이기 위한 피라미드-주의집중 계층네트워크 (FPN-Attention Layered Network) 방식을 제안하고, 실험적으로 그 특성을 평가하고자 한다. 기존의 FPN만을 사용하였을 때 객체 인식 과정에서 설명성을 향상시키는 방식이 객체 인식에 미치는 정도를 정량적으로 평가하였다. 제안된 모델의 적용을 통해 낮은 computing 오버헤드 수준에서 multi-level feature를 고려한 시각적 설명성을 개선시켜, 결괴적으로 객체 인식 성능을 향상 시킬 수 있음을 실험적으로 확인할 수 있었다.

• The Journal of the Korea Contents Association
• /
• v.18 no.10
• /
• pp.40-52
• /
• 2018

Media can be considered as a type of play for kids in this new media era. Researches on media content for kids have mainly focused on educational effect or specific features such as design, but a great deal of kids content is created for fun and enjoyment. This study analyzes fun factors of the most popular kids content on Youtube: Kongsuni and Friends, Carrie and Toys, and Pinkfong. The result shows that seven fun factors can be grouped into three categories; 1) Self-determination and independence are related to the media environment; 2) Sensuous vividness, empathy, and physical dynamic are related to the content; 3) Self-expression and social interaction are related to associated activities. This study extends our perspective on kids content by analyzing fun factors in the comprehensive process of media consumption.

• Broadcasting and Media Magazine
• /
• v.14 no.3
• /
• pp.58-65
• /
• 2009

방송 통신 융합 환경의 도래는 뉴미디어 산업 발전의 기회이며 동시에 저작권의 위기이다. 반면 저작권은 오랜 기간 논의된 문제였음에도, 제도적으로 기술적으로 해결되지 못한 미완의 숙제로 남아있다. 이 글은 저작권 보호와 공정 사용이 문화발전이라는 공동의 목적을 가지고 있음을 논증하고 융합 환경에 대한 구조 분석을 시도한다. 이를 통해 저작권 보호 기술이 가지는 한계와 제약을 검토한다. 또한 OWL로 표현된 관계 기반의 저작물 추적 및 통제 기술을 새롭게 제안한다. 특히 접근 통제 기술인 DRM을 시맨틱하게 설계할 수 있는 저작권 보호 기술의 속성과 특성에 대한 분석을 소개함으로 향후 디지털 콘텐츠 관리 및 응용 기술 구현에 도움을 주고자한다.

• Korean Institute of Interior Design Journal
• /
• v.19 no.2
• /
• pp.117-125
• /
• 2010

Theories and concepts emerging in new media environment are influencing the field of current spatial design. Art and technology have always coexisted, not but they are differentiated in two parts. They show new communicational paradigm in terms of supplementary position each other. The fact that examining the expression method and the communicational structure of today's new media art will be an analytical tool which analyzes the concepts of contemporary spatial design. For this, we attempt to use the Bolter and Gusin's remediation theory as analytical tool. First of all, we consider its theoretical aspects, and then, classify them according to the creation method, and investigate the characteristics of expression according to their types. We prepared an framework of analysis using the Remediation theory and its traits. We reviewed the relations between new media and spaces concerning remediation. After researching, we could analyze the cognitive aspects between new media and spaces as immediacy, and recognize the operational aspects of visual image as hypermediacy. Through the analysis, we could recognize the traits that digital environmental simulacres, sense of telepresence, and non-linear hypermediacy etc. remediate with repeating of mediation of mediation.

• The Journal of the Korea Contents Association
• /
• v.13 no.3
• /
• pp.119-130
• /
• 2013

This paper researches the algorithm, whose materiality and expressiveness can be obtained through live coding. Live coding is an improvised genre of music that generates sounds while writing code in real time and projecting it onto a screen. Previous studies of live coding have focused on the development environment to support live coding performance effectively. However, this study examines the aesthetic attitude immanent in the realization of the algorithm through analyzing mostly used languages such as ChucK, Impromtu, and the visualization of live code and cases of "aa-cell" and "slub" performance. The aesthetic attitudes of live coding performance can be divided into algebraic and geometric attitudes. Algebraic attitudes underline the temporal development of concepts; geometric attitudes highlight the materialization of the spatial structure of concepts through image schemas. Such a difference echoes the tension between conception and materiality, which appears in both conceptual and concrete poetry. The linguistic question of whether conception or materiality is more greatly emphasized defines the expressiveness of the algorithm.

• Spatial Information Research
• /
• v.22 no.6
• /
• pp.91-103
• /
• 2014

The high popularity of geo-referenced mash-up improves the quality and availability of geo-referenced contents. Such contents which accurately represent the real world in street level as well as in aerial level have appeared. Despite the advent of geo-referenced contents of street level, it is still hardly used for mash-up. In this study, we thus find out its obstacles to spatial mash-up and solutions to overcome them to activate the mash-up based on street level contents. By analyzing the current geo-referenced content with respect to its spatial scope and the way of representation, we draw the limitations of current street level contents. Furthermore, we propose the 'new media geo-referenced content' as a solution to overcome those limitations. With 'new media geo-referenced content', it is able to determine the 3D coordinates of given object in real world. Using those coordinates, database linkage can be more closer, flexible and organic. Such 'new media geo-referenced content' can contribute to activation of geo-referenced contents in street level and its mash-up.

• Archives of design research
• /
• v.17 no.2
• /
• pp.451-462
• /
• 2004

The range of graphic design has expanded very widely as the digital era emerged, and many new areas of film design, including film, DVD, interactive TV, mobile and DMB, began to require the work of designers. This could be considered as an application as well as a case of digital film design as it has more developed new media environment as basis. The change in the expression techniques of film language is inevitable since major consumers of new media are the younger generation. The method of creating and distributing visual communication work is not standing still anymore. It has been diversified through dynamic movements and changes resulting from interaction. The major objective of this study is to find out and suggest the directions that film and digital media graduate school education of future by researching and analyzing how the nature of design, as an academic subject that keeps on changing and developing, is evolving in the new media environment. For this purpose, the methods of appropriate education system in higher design education, which are derived from case studies and previous studies on education system, are evaluated by examining the consumers. This study aims at finding the most effective training method in film and digital media design, and by integrating current design education with digital film design education establish a better fit education system that complies with the demands of today's society. It will suggest how an educational institute should see ahead of the time and provide what the society will demand in the future.

• Korean journal of communication and information
• /
• v.40
• /
• pp.45-79
• /
• 2007

This study explored three questions: (1) what are the motives of producing and using user-created contents (UCC, video clips which are produced and uploaded at various web sites by individuals), (2) what are the gratifications that users obtain from UCC, and (3) what are the differences of gratifications according to individual uses of UCC. A survey was conducted to 258 UCC users and data were factor-analyzed to find the dimensions of motives and gratifications. As for the motives of UCC uses, six factors were constructed: self-expression, private life, participation, information retrieval, rest, and entertainment, among which self-expression showed the largest variation. As for the gratifications, five factors were constructed: self-expression, rest, information retrieval, socializing, and entertainment.

• Archives of design research
• /
• v.16 no.4
• /
• pp.225-232
• /
• 2003

A German philosopher, Herbert Bolz predicted that the human way of thinking would be fundamentally and completely changed because of digital media and the 'Gutenberg-Galaxis', named by M.McLuhan which was symbolized of modern reason was doomed to be over. He thought that the limit of reason-centered European culture would be overcome by the up-to-date multimedia to revive the communication life. On this theory, he emphasized the emotional perception, 'aisthesis' which is original meaning of aesthetics. That is to say, he insisted on the restoration of communication media to enable the five senses' amusement condition mentioned by Kant. This thesis asserts that the representative hypermedia digital animation may play a key role to rehabilitate human sensibility pressed by reason centered modernism. Digital animation has the unique worth of Art that is firstly to deal with time and space and enable unlimited expressions and can communicate effectively as a characteristic synthetic medium which consists of intensive computer techniques. Based on the background, this thesis analyzes the possibility of the digital animation as a new medium. Especially, it is focused on the relations to the hypermedia theory of Norbert Bolz who is a media analyst and professor of a design college.

• Journal of Broadcast Engineering
• /
• v.28 no.2
• /
• pp.185-193
• /
• 2023

As many different over-the-top (OTT) services become ubiquitous, demands for high-quality content are increasing. However, high dynamic range (HDR) contents, which can provide more realistic scenes, are still insufficient. In this regard, we propose a new HDR video reconstruction technique using multi-exposure low dynamic range (LDR) videos. First, we align a reference and its neighboring frames to compensate for motions between them. In the alignment stage, we perform content-based alignment to improve accuracy, and we also present a high-resolution (HR) module to enhance details. Then, we merge the aligned features to generate a final HDR frame. Experimental results demonstrate that our method outperforms existing methods.