2016 ANNUAL REPORT Graduate School of Culture Technology
2016 ANNUAL REPORT Graduate School of Culture Technology미래의 문화를 만들어 가는 문화기술대학원문화기술이란 많은 사람들에게 수용되어, 인간 생활에 영향을 주고 생활 방식을 변모/발전시켜, 궁극적으로 한 시대의 문화 현상으로까지 자리잡게 할 수 있을 만큼 파급력을 가진 기술이라고 할 수 있습니다. 정보의 중요성을 근간으로 기술을 개발하는 것이 IT 라면, 인간에 대한 이해를 근간으로 우리가 속한 사회에 직접적인 영향을 주고 더 나아가첨단의 문화를 예측 할 수 있는 잠재력을 가진 기술이 CT 입니다. 문화기술 대학원은 이러한 기치아래, 21세기의 발전된 문화를 정의해 나갈 핵심적인 인재들을 양성 하기 위해 매진하고 있습니다.문화기술대학원은 문화 콘텐츠 산업의 핵심 인력을 키워내고 관련 산업 분야를 한국의 중심 산업으로 육성시키기 위한 목적으로, 첨단 과학과 기술의 요람인 카이스트에 2005년 설립되었습니다. 지난 10년간 석사 295명, 박사 34명등 총 329명의 졸업생을 배출하였으며, 2016년 석사과정 51명, 박사과정 88명으로 총 139명의 재학생을 교육하고있습니다. Digital Content Technology, Socio-Cultural Data Science, Digital Humanities, Human-ComputerInteraction 의 키워드로 분류되는 교과과정을 통해 첨단 문화기술의 지식을 습득하고 Digital Art & Entertainment,Ambient Communication, Interactive Media & Space 의 세 분야를 기반으로 각 영역을 가로지르는 융합연구를통해 세계를 선도하는 다양한 연구 결과물을 만들어 내고 있습니다.구성원들의 부단한 노력에 힘입어, 2016년에는 100편의 논문이 발표되었고, 그 중 32편의 논문이 국내외의 저널에 실렸으며, 22억 원 규모로 23건의 수탁과제 연구가 수행되었습니다. 또한, 28건의 지적 재산권이 출원·등록 되었으며, National University of Singapore와 파리시립미술관을 포함하여 6곳의 국내외 기관과 MOU를 맺으며 폭넓은 교류활동이 이루어지고 있습니다. International Workshop on Science and Culture 2016을 비롯하여 ISUVR2016 & ISDCH 2016을 개최하여 문화기술대학원을 세계적으로 알리는 계기를 마련한 것도 2016년의 큰 성과라 할수 있습니다.앞으로도 문화기술 대학원은 인간 사회에 대한 이해와 사람들의 삶의 질을 향상시키기 위한 근본적인 고민을 통하여문화 콘텐츠 산업에 필요한 원천/응용 기술들을 개발하고 이에 필요한 핵심적인 인력들을 배출할 것입니다.이를 통해, 발전된 미래의 문화를 예측하고 제시하는 선도적인 역할을 하겠습니다. 노 준 용2017년 12월 문화기술대학원 학과장
Table of Contents 7Institute 조직 및 현황 문화기술의 학술적 연구
17 63 123 167Appendix 재학생 및 졸업생 관련 학회 및 학술지 Activity 수상 및 특허 데모데이 콜로퀴움 대학원행사 Project 학술 활동 창작 활동 Study 과목 프로젝트 학위 논문 및 졸업연구
Institute조직 및 현황문화기술의 학술적 연구
Institute 조직 및 현황 문화기술대학원은 설립초기 학사조직 상 학제학부 단위로 설치되었으나, KAIST의 ‘융합기술’분야 육성·강화를 위 한 적극적인 정책추진에 따라 2006년 8월 문화과학대학의 설립과 동시에 문화기술대학원은 인문사회과학부와 함께 ‘인문사회융합대학’ 내 소속되었다. 2016년부터 문화기술대학원은 본격적으로 KAIST 내 일반학과로 성장·정착하였 으며 새롭게 정립된 비전 달성을 위해 교육·연구분야의 조직운영체계를 재정비하여 성과있게 운영하고 있다. 학과교수회의 문화기술대학원의 최고 의사결정 조직으로서 대학원의 주요운영과 발전방향에 대해 심의, 의결하는 전체교수회의이 다. 전임 및 초빙, 겸임교수 전체가 참여하며 매월 첫째주 월요일 정기회를 개최하고 사안에 따라 비정기회의를 병행 하고 있다. 비전위원회 대학원 중.장기 비전 및 발전계획을 수립하고 정책적 운영방향 정립을 위한 협의체로서 역할을 수행한다. 정책연구과 제 제안, BK운영에 따른 세부규정 마련, 학위·비학위과정 개설 등 다양한 형태로 대학원 발전방향을 제시한다. 위원장을 포함하여 전임교수 전원을 위원으로 구성한다. 인사심의회 대학원 관련 교원인사 행정에 관한 기본계획 수립 및 기준설정, 교원 인사에 관한 제규정의 검토 및 제정의뢰, 소속교 원의 임용, 승진, 재계약, 명예교수 추천, 교원포상 등 전반적인 주요 교원인사 사항에 관한 심의 및 제안활동을 수행 한다. 학사·교과과정위원회 대학원의 학사운영 및 교과과정 신설.변경.폐지, 졸업이수요건 등에 관련된 규정을 제.개정하는 역할을 수행한다. 학사책임교수가 위원장으로 주요 교육 및 학사운영 사항에 관한 심의 및 제안활동을 수행한다. 연구기획위원회 대학원의 공동연구분야 도출, 신규사업 발굴 및 기획 등 비전달성에 걸맞는 R&D 정책방향을 검토 논의하고 대내외 과제제안활동 등을 주도적으로 수행한다.8
Institute 조직 및 현황 조직도 문화기술대학원 학과교수회희연구기획위원회 인사심의회 비전위원회 학사·교과과정위원회 행정팀 교육분야 (Keywords) 연구분야 Content Technology Visual Content Technology (콘텐츠 기획, 창작 및 공유 기술) (시각 콘텐츠 기술) Cultural Science Sound & Music Technology (문화과학) (음향 및 음악 기술)Interaction and Design Computing: Interactive Technology (상호작용 및 디자인 컴퓨팅) (상호작용 기술) Computational Design (컴퓨테이셔널 디자인) Cultural Complexity Science (문화 복잡계) 9
Institute 조직 및 현황 ■ 교육분야(Education) 콘텐츠 기획, 창작 및 공유 기술 (Content Technology) 21세기에 접어들면서 문화 콘텐츠는 빠르게 발전하는 다양한 기술들의 접목을 통해 디지털 영역으로 급속히 확산되 고 있다. 아날로그적으로 생성되고 공유되던 콘텐츠가 이제는 디지털화되어 극장, TV 는 물론 각종 모바일, 웨어러블 디바이스를 통해 한층 더 쉽게 다양한 정보 및 오락을 제공하며, 진화하는 디지털 미디어 환경에 맞게 주제 및 소재, 포맷과 표현 방식을 확장하고 있다. 본 대학원의 디지털 콘텐츠 기술의 대표적인 분야에는 영상 제작, 편집, 시각특수 효과 등을 다루는 영상 콘텐츠 기술; 음악 및 음향의 생성, 가공, 해석 등을 다루는 음성 콘텐츠 기술; 가상 세계 구축 방법론 및 몰입적 경험 전달 등을 다루는 가상현실 기술; 신체 움직임을 주축으로 하는 체감형 게임과 다양한 산업분 야와 결합하여 사회적 파급력을 가지는 기능성 게임; 디지털 콘텐츠를 위한 스토리텔링 및 포맷 개발 등이 있다. 문화과학 (Cultural Science) 데이터란 자연계, 기술계 등 모든 계(=시스템)를 이해하기 위한 과학 활동의 시작이 되는 관찰의 기록이다. 급속도로 발전하고 있는 디지털 기술에 힘입어 문화계 데이터 또한 대규모로 쌓이고 있으며, 창작자와 소비자의 상호작용을 관 찰할 수 있는 형태로 진화하고 있다. 카이스트 문화기술대학원에서는 자연과학, 사회과학, 공학에 기반한 학제간 데 이터 과학을 통하여 미래 문화의 모습을 예측하고 유용한 데이터 응용 소프트웨어를 개발하는 데 역점을 두고 있다. 구체적인 주제로는 문화 창작 및 소비자들의 네트워크 데이터 분석을 통한 미래 창작의 지형 규명, 스토리 구조의 동 역학 이해를 통한 미래형 내러티브 구축 방법, 소셜미디어를 통한 문화적 선호도 발굴과 응용이 있다. 상호작용 및 디자인 컴퓨팅 (Interaction and Design Computing) 최근 현실공간과 이에 대응하는 가상공간을 3차원적으로 서로 연결하고, 부가적인 실감 콘텐츠와 정보로 관심 공간/ 장소/대상물을 증강하며, 양방향으로 상호작용하도록 하는 등 인간의 공간 인지와 경험을 확장시키는 연구가 주목 받고 있다. 이에 따라, 가상과 현실이 연동된 미래의 지능 공간은 다양한 방식으로 인간의 육체적 능력, 지적 능력, 그 리고 사회적 능력을 확장하게 될 것이다. 본 대학원에서는 지능 공간의 계산 인지, 가상현실, 증강현실과 증강휴먼 등 의 토대위에서, 게임, 디지로그 문화유산, 관광, 전시 및 공연 등 미래형 문화산업 분야에서 필요로 하는 지능공간에서 의 다양한 2D/3D 상호작용을 교육한다. 디자인 컴퓨팅 혹은 컴퓨테이셔널 디자인(Computational Design)은 디자인 분야에서 나아가, 사용자경험(UX), 문 화요소, 창의성과 같이 인간의 두뇌로 해결하기 어려운 추상적 개념으로부터 도출된 어려운 문제들을 컴퓨터를 통하 여 탐구한다. 이러한 지적 탐구는 알고리즘과 휴리스틱 계산 방법을 사용한 향상된 계산성을 토대로 컴퓨터 기반의 프레임워크 및 시스템 개발에 기여한다.10
Institute 조직 및 현황 교육과 연구분야■ 연구분야(R&D)시각 콘텐츠 기술 : Visual Content Technology실시간으로 초사실적인 콘텐츠를 제공하고, 사용자에게 몰입감을 제공하는 기술To create hyper-realistic content in real-time and to provide the audience with immersive experience음향 및 음악 기술 : Sound & Music Technology계산적인 접근을 통한 음향 콘텐츠 이해, 모델링, 생성으로 청각적인 경험과 음악활동을 향상시키는 기술To enhance auditory experiences and musical activities by understanding, modeling, and generatingaudio content through computational approaches상호작용 기술 : Interactive Technology스마트 공간에서 2D/3D상호작용을 제공하기 위한 계산적 인지, 인간 중심 경험 디자인, 가상현실, 증강 현실, 증강 휴먼 기술To provide various 2D/3D interactions in smart space by exploring integrated research in the areas ofcomputational perception, human-centered experience design, virtual reality, augmented reality andaugmented human컴퓨테이셔널 디자인 : Computational Design제조업과 서비스 산업에서 새로운 접근법인 문화 정보를 활용한 대량 맞춤형 시스템 구축을 위한 기술To build mass customization systems while integrating cultural information which are the new frontierin both manufacturing and service industries문화 복잡계 : Cultural Complexity Science문화시스템의 복잡한 역학을 해독하고 문화산업을 위한 새로운 과학적 방법론을 개발하는 기술To decipher the complex dynamics of cultural systems and develop novel scientific methodologies forcultural industries 11
Institute 2016 재학생 조직 및 현황 분포 현황 재학생 현황 졸업생 현황 (2016.9월 기준) 재학생 : 총 139명 석사과정 51명 / 박사과정 88명 ■ 석사과정 : 과기원(문화부)장학생 43명, 일반 장학생 4명, 외국인 2명 ■ 박사과정 : 과기원(문화부)장학생 62명, 일반 장학생 13명, 외국인 1명 * 재학생 학부 전공비율 공학/자연계열 : 107명(77%), 인문/사회계열 : 23명(17%), 예술/디자인계열 : 9명(6%) 2005-2016 재학생 현황 2007-2016 졸업생 현황 2007 - 2016 졸업생 현황 2005 - 2016 재학생 현황12
Institute 조직 및 현황 졸업생 진로 현황졸업생 : 총 329명석사 295명/ 박사 34명■ 졸업생 취업현황- 민간기업 (Weta Digital, Universal Studios, Samsung, LG, Kakao, NHN, SK, KT, Nexon, KT&G, CJ E&M, TmaxSoft 등) : 129명- 연구기관 및 단체 (프랑스 국립과학원, 삼성경제연구소, ETRI 융합기술생산센터, 영화진흥위원회 영상산업정책연구소, LG전자 CTO Convergence연구소, 대한민국예술원) : 37명- 교원 및 공무원 (전임: 성균관대학교, 서강대학교, 국민대학교, 아주대학교, 단국대학교, 홍익대학교, 오산대학교, 연세대학교, 강사: 고려대학교, 강원대학교, 동덕여자대학교, 공무원: 통계청) : 16명- 국내/국외 대학원 진학 (취리히 공대, 뉴욕대 MBA, Michigan State University, University of Maryland, 일리노이 대학, Ohio State University, KAIST 동대학원 진학 (41명), 그 외 서울대, 연세대) : 85명- 공공기관 (국립과천과학관, 경기도어린이박물관, 한국문화콘텐츠진흥원, KBS, 마케팅공사) : 10명- 창업 (레진엔터테인먼트, ㈜파란오이필름, 엔트리코리아, 카이스튜디오, 아이앰랩, CatchItPlay, 앨리스원더랩, 보다미디어그룹,클래스팅, 어반플레이, ㈜스피어그램, 위브랩, 안바도비디오, 버니버닛, 어딕션) : 20명- 프리랜서 (영화감독, 미디어아트 작가, 사이언스 아티스트, 사진작가, 가수) : 7명- 기타 : 25명2007 - 2016 졸업생 진로 현황 13
Institute 조직 및 현황 논문/창작/수탁 연구 실적 현황 창작, 기획 및 학술연구 운영실적 ■ 학술논문 발표실적 (2016. 1. 1 - 2016. 12. 31) 구분 국제저널 국내저널 국제학술 국내학술 계 2008 10 2 13 8 33 2009 6 6 31 15 58 2010 8 11 26 16 61 2011 15 9 28 21 73 2012 15 14 45 30 104 2013 23 12 63 17 115 2014 22 18 46 24 110 2015 19 9 45 7 80 2016 24 8 40 28 100 총계 142 89 337 734 166 수탁연구 수주실적 ■ 23과제, 2,255백만원 (2016. 1. 1 - 2016. 12. 31) 구분 금액 (단위: 백만원) 수탁과제 건수 2008 1,082 14 2009 1,798 24 2010 1,633 22 2011 2,665 30 2012 3,144 26 2013 2,831 30 2014 3,264 24 2015 2,244 24 2016 2,255 23 총계 20,916 217 총 게재논문 수 수탁연구 수주 금액 (단위: 백만원)14
Institute 조직 및 현황 창작활동 현황주요 연구 결과 현황 • 대형 영상 상영관 기반 실감영상 및 상영 시스템 개발 • K-Culture Time Machine : 시공간 연결형 문화콘텐츠 생성 및 제공 기술개발 • 모델예측제어에 기반한 데이터 구동 물리 기반 동작 생성 • 디지털 문화유산 영상관 전시 기본계획 및 공간구성 연구 • 대화 상황과 감정 인지형 인공지능 대화 시스템 개발 • 패턴, 프레이즈, 모티프 기반의 지능형 국악 가상악기 개발 • 3D 글래스 환경에서 기하관계 중심의 스마트폰 기반 상호작용형 저작 기술 • Virtual Experience向 Avatar animation 기술 개발 • 촬영 각도/거리 기반 제품형상 왜곡보상 기술 개발 • 맞춤형 스포츠 기어 신속제작을 위한 인체-스포츠기어 통합 역학 시뮬레이션 소프트웨어 개발(3차년도/총3차) • 인공지능형 빅데이터 기반 원자력 관련 국민 여론 추이 판단 시스템 연구 • 원격지 인간간 접촉 인터랙션을 매개하는 가상 아바타 애니메이션 기술 창작(출판/전시/영상/공연/기획) 실적■ 5건 (2016. 1. 1 - 2016. 12. 31) 구분 전시 공연 출판 영상 기획 기타 계 2008 1 5 8 - - - 14 2009 12 14 6 5 6 - 43 2010 20 11 4 2 - - 37 2011 11 8 - 1 4 7 31 2012 12 9 4 9 1 3 38 2013 10 5 3 0 3 1 22 2014 8 3 2 1 14 2015 2 4 6 2016 1 4 5 총계 76 59 27 17 14 12 205창작활동 현황 2520151050 2009 2010 2011 2012 2013 2014 2015 2008 전시 공연 출판 영상 기획 기타 15
Institute 문화기술의 학술적 연구 CT의 연구방법 • CT연구는 예술, 교육, 과학기술, 의식주, 제조산업 등 기존의 인간 삶의 방식과 양식에 대 한 이해에서 출발 • 문화콘텐츠산업에 대한 연구는 해당 영역의 좁은 시야 안에서만 이루어지는 것은 한계 가 있음 • 산업의 계층 그림에서 봤을 때, 창조산업의 이해를 바탕으로 문화산업으로, 문화산업의 이해를 바탕으로 문화콘텐츠산업에 접근하는 연구방법이 적용되는 것이 바람직함 • 넓은 영역을 포괄하는 통찰력을 바탕으로 밖에서부터 접근하는 연구방법을 적용하여야 기존의 것을 뛰어넘는 새로운 도약을 도모할 수 있음 산업 창조산업 문화산업 문화콘텐츠산업 콘텐츠기술 문화기술 예술표현기술 경험기술 고유영역 디지털디자인 플랫폼응용기술 문화기술 문화 경제경영/정책 디지털 기술 디지털 인문사회학 기초영역 Aesthetics, Cognitive science, Communication, 인접학문 Economy, Management, Statistics & Probability, 2D/3D design, Web design & programming16
Institute 문화기술의 학업적 연구 CT의 고유기술문화기술의 학술적 연구범위에서 고유기술에 해당하는 기술의 연구범위 기술명 연구범위 콘텐츠기술 예술표현기술 • 문화콘텐츠 창작,제작을 지원하고 체계적인 경영을 가능하게 하는 방법론 경험기술 • 문화콘텐츠 유통 및 보호 기술을 아우르는 연구 분야로서 영상기술, 게임기술, 디지털디자인 모바일콘텐츠기술 등 문화콘텐츠 각 장르별 기술군으로 구성플랫폼응용기술 • CT 기술 분야 중 문화콘텐츠산업을 직접 지원하는 분야 (미디어아트, 건축, 전 시,정보디자인 등) • 각종 예술 장르에서 표현력을 극대화하거나 작품 창작/제작의 효율성을 높이 는 데 기여하는 기술 • 디지털 음악/음향,디지털 공연,시각/공간예술 지원 기술 등 으로 구성 • 디지털기반 기술로서 사물의 문화적 가치를 향상시키거나 인간의 문화적 생활 을 함양하는데 적용되는 기술 • 전시기술,관광기술,스포츠기술,문화재관련 기술 등으로 구성 • 디지털기반 기술로서 사물의 문화적 가치를 향상시키거나 인간의 문화적 생활 을 함양하는데 적용되는 기술 • 전시기술,관광기술,스포츠기술,문화재관련 기술 등으로 구성 • 인터넷, 모바일 등 다양한 플랫폼을 기반으로 하는 문화콘텐츠의 유통, 소비를 활성화하는 기술 • 웹을 통한 문화콘텐츠 유통의 활성화로 문화콘텐츠의 유통에 기반이 되는 기술 • 다양한 분야로 확장이 가능하며, 유통되는 문화콘텐츠와 플랫폼의 속성에 대한 근본적인 연구가 이루어져야 함 • 소셜컴퓨팅,IPTV기술 등이 이에 속함 17
Study과목별 성과학위 논문 및 졸업 연구
Study과목별 성과 문화기술론Team Seoul | 유정은, 용상언, 김용한Title | Journey Across the Universe with The Little Prince어린왕자 이야기를 기반으로 사람들에게 잊고 살던 가치를 일깨워줄 수 있는 모바일 기반 미니 게임을 제안하였다. 사용자들은 모바일 디바이스를 가지고 걸어다니며 지구를 비롯한 행성을 여행하는 게임 시나리오를 체험할 수 있으며, 미션을 수행하는 동안 나타나는 자신이 과거에 찍은 사진과 인터랙션하며 추억을 회상하고 잊고 살던 가치를 떠올릴 수 있는 것이 특징이다.Team Ecuador | 신재은, 이현진, 이종인, 전진우Title | PLANETOON: A Collaborative Webtoon Platform Exploring the World of Interdisciplinary Studies사용자들이 제안한 이야기를 수집하고 가장 적절한 이야기를 고른 후에 만화 형식으로 사용자들에게 컨텐츠를 제공하는 크라우드 소싱 기반 웹툰 제작 어플리케이션을 제안하였다. 사용자들은 웹과 모바일 어플리케이션을 사용하여 제시된 주제에 이어질 아이디어나 이야기를 자유롭게 제시할 수 있다. 이때 시스템은 사람들이 제안한 다양한 이야기를 취합하고 사용자들의피드백을 콘텐츠에 반영하여 하나의 이야기를 완성하며 웹툰 형식으로 제공한다.
Study 과목별 성과 문화기술론Team Paris | 박흥석, 최지정, 박지영Title | Bucket PrinceAre you still dreaming about your travel? Why don’t you plan now? 사용자가 원하는 여행지를 입력하면 빅데이터 기반의 관련 여행 장소, 여행 명소, 숙소 등의 콘텐츠를 제공하는 여행 계획 가이드 어플리케이션을 제안하였다. 사용자는 어린왕자 스토리를 기반으로 어린왕자와 함께 보다 쉽고 친근하게 여행 계획을 세울 수 있는 것이 특징이다.Team Oslo | Amartuvshin Narangerel, 강슬기, 김성실Title | The Voyage of a Fox어린왕자 이야기 중, “What is essential is invisible to the eye” 구절에서 영감을 받아, 사용자가 촉각과 청각을 사용하여“Seeing with your heart”를 실현할 수 있는 모바일 퍼즐 게임을 제안하고 제작하였다. 어린왕자 이야기를 기반으로 여우가 어린 왕자를 찾기 위해 여정을 떠나는 주제로 게임 시나리오를 재각색하였다. 사용자는 여우가 어린왕자를 찾을 수 있도록 단서를 찾고 퍼즐 게임을 풀어야한다. 게임은 사용자가 시각적 요소가 아닌 소리나 진동, 감각에 의존하여 풀도록 디자인한 것이 특징이다.
Study과목별 성과 문화기술론Team Guatemala | 김상훈, 진샘, 최순범Title | Lonesomnia: Save my little prince어린왕자가 진실된 친구를 만나고 다양한 행성을 여행하는 어린왕자 원작을 바탕으로 인터랙티브 프로젝션 기반의 보드게임을 제안하고 제작하였다. MBTI 테스트를 기반으로 사용자의 게임 캐릭터를 설정하고 보드 게임의 아날로그적 상호작용을 디지털 화면으로 표현한 것이 특징이다. 캐릭터와 게임 진행 상황에 따라 관련 장면과 콘텐츠가 보드게임 위에 프로젝션 된다.
Study 과목별 성과 문화기술론Team Pompeii | Shengen Lim, 이기쁨, 최동혁Title | Remember어린왕자의 “중요한 것은 눈에 보이지 않는다 (What is essential is invisible to the eye)”라는 구절에서 영감을 받아 한국 사회의 문제를 다룬 4개의 미디어 아트 작품을 구상하였다. 첫번째 “the witness”는 위안부 문제를 interactive webtoon을 이용하여 문제의식을 전달하였다. 이 작품은 위안부 소녀상의 손을 어루만지는 것으로 웹툰과 인터랙션을 시도하였다. 두번째“the remained”는 Aqua display를 이용하여 세월호 사건의 아픔을 위로하고자 했다. Aqua display 밑에서 떠오르는 희생자들의 실제 메시지를 관람객이 응시하며 그곳에 남아있던(remained) 사람들과 지금 여기에 남아있는(remained) 사람들의연결을 시도한다. 세번째 “the vip”는 외모지상주의를 비판한 작품으로, AR을 이용해 실제 이미지와 스크린에 보이는 이미지의 편차가 누가 중요한 사람인지 역설적으로 드러내준다. 네번째 “the voyeur”는 탈북자들을 관찰한 비디오 클립을 편집한비디오아트로 그들을 바라보는 한국인의 시선을 관음(voyeur)에 빗대어 표현한다.
Study 지도교수 학위논문 졸업연구 남주한 석사과정 금상은 Singing Melody Extraction Using Multi-Column Deep Neural Networks While the music market has been growing, the need for new service has also been increasing, such as cover song identication and query by humming. These services use a melody to search songs and so extracting melody, particularly from singing voice, is important to implement the systems. In this thesis, we focus on algorithms to extract the singing melody from audio signals. Singing melody extraction is a task that tracks pitch contour of singing voice in polyphonic music. While the ma- jority of melody extraction algorithms are based on computing a saliency function of pitch candidates or separating the melody source from the mixture, data-driven approaches based on classication have been rarely explored. In this thesis, we present a classication-based approach for singing melody extrac- tion using multi-column deep neural networks. In the proposed model, each of neu- ral networks is trained to predict a pitch label of singing voice from spectrogram, but their outputs have dierent pitch resolutions. The melody contour is inferred by combining the outputs of the networks. We conduct the Viterbi decoding based on hidden Markov model to capture long-term temporal information. Our system also includes a singing voice detector to select singing voice frames using an additional deep neural network. It is trained with labels of singing voice activity and the output of deep neural networks for melody extraction. In order to take advantage of the data-driven approach, we also augment training data by pitch-shifting the audio content and modifying the pitch label accordingly. We use the RWC dataset and part of the MedleyDB dataset for training the model and evaluate it on the ADC 2004, MIREX 2005 and MIR-1k datasets. Through several settings of experiments, we show incremental improvements of the melody prediction. Lastly, we compare our best result to those of previous state-of-the-arts.24
지도교수 Study 학위논문 원광연 졸업연구 석사과정 김건기 An Experiment on Augmenting Information Visualization with Real Objects: Design and Evaluation by the Coloring TypeInformation visualization is the study of visual representations of abstract data toreinforce human cognition. Some people unfamiliar with information visualizationmay have some difficulty to understand because of the visual representation of data.However, displaying real objects could augment visualization to help user’s betterunderstanding.The purpose of this thesis is to explore the potential of using real objects in informa-tion visualization. For this, this thesis defines augmented visualization as visualiza-tion with real object display for more realistic representation, and then explains itsnecessity.To investigate the effects of displaying real objects in information visualization,augmented visualizations were designed with three coloring types of real objects(screen color, color projected, colored) and user evaluation was implemented.Quantitative evaluation results indicated that displaying objects of real color wasgenerally the most effective way to augment visualization while other types haverelatively poor or negative effects in some evaluation items. Qualitative evaluationresults showed merits and demerits of each type. Designers should select a suitablecoloring type of real objects for their visualization in consideration of the merits anddemerits.In conclusion, this thesis may facilitate systematic research about augmented visu-alization and help people who want to design augmented visualization. 25
Study 지도교수 학위논문 졸업연구 차미영 석사과정 김승희 Discourse on Nuclear Safety after Fukushima Disas- ter among Korean Twitter Users In the era of hyper-connected world, a catastrophe in one country is broadcasted live and can aect policies in other countries. The 2011 Fukushima Daiichi nuclear disaster in Japan is one such prominent case that in uenced many countries. Based on analyzing 158,964 tweets generated over a 4-year period, this paper presents a model of risk perception on radiological accidents in Japan’s nearest neighbor, Korea. We explain the process through which continual media broadcasts of foreign risks (i.e., disaster in Japan) become perceived by Korean social media users as internal risks and hence call for agenda setting and participation at a societal level. Our model empirically presents a mechanism that could transmit perception and awareness towards risk across cultural boundaries.26
지도교수 Study 학위논문 김정화 졸업연구 우운택 석사과정 김하연 Practical Use of KCHDM Ontology Standard in Devel- opmentof Cultural Heritage Mobile AR Information Service -With a Case of Injeongjeon and the VicinitySystems, having combined Augmented Reality with the Semantic Web technologies,enable visitors of the heritage site to query and consume cultural heritage data onthe web. However, these systems usually serve as data providers rather than provid-ing support for cultural heritage interpretation. The interpretation of heritage sitesshould explore the significance of a site in its multifaceted historical, political, spir-itual, and artistic contexts. Therefore, this research studied the application method ofKCHDM ontology for the development of a mobile AR information service to ag-gregate and provide cultural heritage data contextually. With a case of Injeongjeonand the vicinity of Changdeokgung Palace, we extracted data from five Korean cul-tural heritage web databases and designed information through three mapping phas-es. We implemented the application and conducted user studies to provide directionsthat will be useful for designing information to support visits to heritage sites, anddiscussed how the AR technology could improve the application. 27
Study 지도교수 학위논문 졸업연구 우운택 석사과정 김형일 Smartwatch-assisted Robust 6-DOF Hand Tracking System for HMD-based Augmented Reality. In this paper, we propose a smartwatch based sensor fusion approach to robustly track 6-DOF hand movement in head mounted display (HMD) based augmented reality (AR) environment, which can be used for robust 3D object manipulation. Our method uses both wrist-worn smartwatch and HMD-mounted depth sensor to robustly track 3D position and orientation of user s hand. We introduce HMD-based augmented reality platform with smartwatch, and method to accurately calibrate orientation between smartwatch and HMD. We also implement natural 3D object manipulating system using 6-DOF hand tracker with hand grasping detection. Our system is easy to use, and doesn’t require any hand held devices. User test shows that our method shows much higher performance in 6-DOF virtual object manipula- tion than existing vision based hand tracking method.28
지도교수 Study 학위논문 이성희 졸업연구 Key Pose-Based Ballet Scoring Using Benesh 석사과정 Movement Notation Syste 김혜지Extracting key frame in the animation is crucial for archiving, retrieving, and alsofor making more optical posture or movement of the character. In this sense, ex-tracting key frames of ballet works is critical to abstract the main points of bal-let story and motion information. Ballet is composed of very artistic and scientificmovement based on limb actions: extension, flexion, rotation and abduction. Thus,it is meaningful to focus on the ballet specialized method and to develop more ap-propriate way to inform the ballet movement information. This paper addresses newapproaches to find a ballet specialized key frame using ballet key poses and theballet notation system: Benesh Movement Notation system, and also speculates aneffective way to present them in the end.To extract ballet key poses, 168 poses are selected by the Vaganova technique. 168postures, including basic ballet pose: 5 feet and 3 arm poses are selected and all ofthese poses are exported to bvh file format by the DanceForms software. Theseselected 168 poses are then considered as input data to compare ballet key poses inreal ballet works(1309 frames). To make a graphic readable platform to change 168ballet key poses into numerical format, Benech movement notation principles areapplied. BMN puts the human body on five staves followed by head, shoulder, waist,knee, and foot levels. Consequently, each pose is expressed as relative joint coordi-nates in that BMN grid is a mathematical understructure to calculate relative jointposition. To extract the main key frames in the whole sequence, the Euclidean Dis-tance was used to calculate the pose similarity distance between original ballet keyposes and poses in 1309 frames, and the results were presented as musical score typebye the DanceForms, in the end. Consequently, this paper shows the key pose-basedscoring using ballet notation system is efficient to catch the ballet points, in the end. 29
Study 지도교수 학위논문 졸업연구 우운택 석사과정 박진우 Instant Radiosity-based Global Illumination in AR Considering a Real-world Environment. Graduate School of Culture Technology This thesis deals with a novel method for performing real-time photo-realistic ren- dering of the virtual objects in augmented reality environment. With the develop- ment of the new AR devices such as HoloLens from Microsoft, augmented reality technology is expected to change everyday life. However, because of the unrealistic appearance of virtual objects, users have difficulties in concentrating on the AR con- tents. For resolving this problem, it is necessary to render visually well-matched vir- tual objects by reflecting the information of real environment. The proposed system is composed of three technical steps: 1) Obtaining 3D voxel data of real environment by using an RGBD sensor 2) Randomly sampling radiances from the main light source and detecting collisions with the surface voxels 3)Creating virtual point lights and calculating global illumination in real-time Through this process, we can make photo-realistic rendering of augmented objects in real-time and enlarge users’ immersion in AR contents.30
지도교수 Study 학위논문 노중용 졸업연구 이성희 석사과정 김재현 Study on Spontaneous Interactive Animation for Virtual Agents based on Iterative Predictions and Action-Difference LearningSpontaneous reactions is assumed to play a vital role in making realistic human-agent or agent-agent interaction. For the spontaneity, the importance of abilities topredict action and to control reaction speed were investigated. The suggested data-driven approach used action-reaction pairs that are temporal skeleton informationof two persons captured from a depth camera. The reactions synchronized with orfaster than actions were made by learning the data with artificial neural networks.One part of networks predicted action pose at a time step, and the other created aninteraction representation, corresponding to the action pose, which is the differencefrom the action pose to a reaction pose. The results showed that the synchronizedand faster reaction with a few steps of valid action prediction could afford a virtualagent a certain extent of spontaneity. 31
Study 지도교수 학위논문 졸업연구 이성희 석사과정 상종희 Gait Generation with Low Dimensional Embeddings Locomotion of human-like character locomotion plays important role in computer games and animations as well as simulations with virtual environment. Nowadays, many researchers are focused on the environment adaptive motion as requirement of users. Also, since the human motion data is an articulated object with a high number of degrees of freedom, nonlinear latent variable models are useful method to reduce dimensionality of motion data. In this paper, we propose a novel approach for generating gait of biped figure to adapt the environment and efficient search for reduced space. When human goes to the destination and interact with the environment, he/she is more interested in the position of his/her foot than the joint angle of his/her knee. In this sense, one appealing approach to generate a rich set of character gait with nonlinear latent variable model is to use end-effector of motion as features. Moreover, in order to efficiently accomplish the locomotion in user-specified environments, we use tree search for reduced space. Given start and target positions with respect to stance foot in a virtual environment, our scheme gives an optimal solution in the reduced space to step from the start to the goal. The proposed method consists of following four parts: feature extraction, learning of GPLVM, search for reduced space, and constructing lower limb. The experimental results show that the proposed method can generate highly realistic gait trajectory and the extracted features are reliable for motion generation. We demonstrate our approach through the user-specified environment, interactive input and compare our method to others.32
지도교수 Study 학위논문 김정화 졸업연구 석사과정 서주희 A Study on Constructing Semantic Structure of His- toric Site Information by Implementing KCHDM: Fo- cusing on Achasan Koguryo Historic SiteWith vast amounts of data being generated in the field of digital cultural heritage, ex-tensive studies have been conducted on how to utilize such data and provide relatedservices. One challenge in providing cultural heritage information is enabling usersto experience content in a meaningful and coherent manner. Institutions special-izing in digital heritage information services are more focused on the provision ofinformation, and fall short of user expectations due to a lack of support in informa-tion access and browsing. Historical site information is especially heterogeneousand exhibits difficulty in such management and service. Against this backdrop, thisstudy proposes the development of a semantic structure. First, key relations to sup-port access and interpretation of historical site information were identified based onan analysis of the information characteristics of historical sites. Second, using theAchasan Koguryo Historical Site as a case study, the Korea Cultural Heritage DataModel (KCHDM) was applied to model the relations through a two-way approachof top-level mapping and bottom-level mapping. Third, a knowledge-based platformprototype was constructed as one way of visualizing the semantic structure, and auser test of the platform was conducted in order to understand how users experiencethe semantic path. The significance of this study lies in suggesting a semantic struc-ture for the management and service of heterogeneous historical site information,which field has been largely neglected in research and in implementing KCHDM foractual historical site information applicable to the existing database. 33
Study 지도교수 학위논문 졸업연구 노준용 석사과정 서진 Aesthetic Evaluation on Projected Wall versus Real Wall in Home Environment Displays are getting larger, thinner and flexible. In this point of view the wall in the home can be replaced to the wall-size display in the near future. As projectors have advantages in high resolution, wide field of view and relative cost, major- ity of related studies developed wall size display using projection system. Accord- ingly, in this paper virtual wall is defined as a wall-size display using projection system. However, it is not clear so far whether the projected wall is replaceable to the real wall especially in the aesthetic aspect. The aim of this study is to develop the aesthetic response user study on the virtual and the real wall and to figure out the differences between two walls. The results showed that people find out the visual differences between the virtual wall and the real wall. The virtual wall was judged more aesthetically than the real wall in this study. Moreover, there were differences according to three colors related to the color temperature.34
지도교수 Study 학위논문 이지현 졸업연구 석사과정 안성은 Identifying the in uence of design features on vehi- cle style recognition using contour deletionIn the automobile industry, maintaining recognizable brand style is known to beparamount for realizing market competitiveness. To develop and maintain style rec-ognition, understanding how consumers recognize car styles and investigating thevisual signicance of each design feature of a car design are necessary. The objectiveof this research was to identify the quantitative visual in uence of design featureson car style recognition, focusing on shapes of design features. To do so, we rst de-composed car design into design features, applied contour-deletion methods to eachdesign feature, and tested each pair of each design feature at three contour-deletionlevels for conjoint analysis. We conducted the consumer survey in three phases. Firstwas the pretest to test the eect of contour-deletion method. Second was the mainsurvey to estimate the visual signicance coecients through conjoint analysis. Thirdwas verication survey for verifying the coecients derived from the main survey. Allsurveys were done through Amazon Mechanical Turk. The visual signicance resultswere all dierent for brands, but all brands had high visual signicance coecients forthe grill. The main nding of this paper is that design features with a high visual signi-cance have a more important impact on consumers’ perception of the brand styles ofthe car design. Thus, the primary implication of this is that designers and othersinvolved in the formal evaluation of product appearance are able to reason and makejudgments concerning brand recognition from designs less subjectively. 35
Study 지도교수 학위논문 졸업연구 김정화 석사과정 안재연 A Research on developing KCHDM for Semantic En- richment of History Huseum Collection Information : HTC-based Navigation In the era of digital technology, cultural heritage institutions have been integrating their heterogeneous dataset to overcome the problem of lack of meaning in col- lection data. In this paper, framework for HTC-ONP (Historical Thinking Concept based Navigation Path) was proposed as a new way of browsing history collection data. It suggests navigation path using ontology triples as a path. This framework use KCHDM (Korean Cultural Heritage Data Model) as ontology model which is developed with the goal of integrating Korean heterogeneous databases with ontol- ogy based approach. And as a theoretical base of HTC-ONP, Historical Thinking Concept of UBC was applied which was developed for the benchmarks of historical thinking and history performance assessments in history education. The framework suggested in this study was applied to design virtual museum with case of the Korean Independence Movement of 1919. With this application case, usability test and interviews with history museum experts were conducted to inves- tigate the potentiality of this framework. The result of this study shows HTC-ONP would be a useful tool for various application for history museum. This study pre- sents practical possibility of ontological approach in history museum information.36
지도교수 Study 학위논문 이원재 졸업연구 석사과정 여찬구 The Effect of Network Closure in Collaboration Network on Performance: Time-series Network Analysis in <Leagueof Legends>This study tests a hypothesis that the performance of <League of Legends> gamers isinfluenced by the structural position within their collaboration networks. From ‘OP.GG’,we constructed ego-centric networks of 130 gamers for 50 days. We divided the networksinto seven-day moving windows for time-series analysis. We computed clustering coef-ficient to measure the degree of how closely gamers are connected, ‘embeddedness’, andestablished a panel regression model. Results demonstrated that the gamers embedded inclosed networks were more likely to improve their tier. 37
Study 지도교수 학위논문 졸업연구 이지현 석사과정 이상은 Color and Shape Recommendation System for CI/BI Logos Based on Corporate/Brand Identity and Pub- lic images As the competition among corporates/brands becomes more and more intensified, the user-friendly and positive corporate/brand image has emerged as an essential is- sue in order to increase the value of the corporate/brand. Corporates and brands are making continuous efforts to increase their values and to form desired images for increasing the corporate/brand recognition by the consumers. For the image forma- tion of such corporates and brands, corporate identity(CI) or brand identity(BI) is getting important. Out of these CIs and BIs, the logo which is a visual element is a valuable asset and a strategic means to differentiate themselves from competitors, and in reality, it has an immediate and vital influence on consumers. In spite of the importance of the logo, it is difficult for non-experts to design it since designing logo is considered as domain expert’s work. In this study, we proposed a method that allows non-experts to design logos easily using the knowledge which has been accumulated in successful CI/BI logos. We have identified the public im- ages of existing corporates or brands, and analyzed their logo designs in terms of color and shape, and developed a novel color and shape recommendation system for CI/BI logos. To build our recommendation system, we first collected and analyzed color and shape characteristics of the existing 500 logos, extracted the image adjectives with respect to the public corporate/brand images of these logos through the survey, and then constructed a CI/BI logo database. Next, we analyzed the correlation between the public response about logos and the color and shape characteristics of the CI/ BI logos using the k-nearest neighbor (k-NN) and decision tree algorithm, and then finally implemented the color and shape recommendation system for designing CI/ BI logos. We have verified our system using the accuracy of recommendation. In the case of color recommendation, we measured how accurately the actual colors of existing logos coincides with the recommended colors by our system. Experimental results show that the first color recommendation achieved around 80% accuracy while the second color recommendation achieved 78.26% accuracy, respectively. For shape recommendations, we generated the shape recommendation model using C5.0 algo- rithm using the training data set, and verified the accuracy of our model using the test data set. As a result, the accuracy of symbol mark recommendation was 71.07% while that of word mark recommendation was 81.59%, respectively, indicating the considerable level of accuracy.38
지도교수 Study 학위논문 도영임 졸업연구 석사과정 이석우 Mobile Game Design Strategies for Enhancing En- joyment andSelf-Efficacy to Promote Daily Physical ActivityThe importance of performing physical activity in daily life has gained continu-ous attention. In this research, we propose mobile game design strategies throughpervasive game play, symbolization, achievement and self-regulation for enhancingenjoyment and self-efficacy to promote daily physical activity. We present an im-plementation of the game prototype which is designed by using the proposed fourstrategies. The game was inspired by the old traditional play, the ground of conquer.This paper conducted user studies twice 1) to see how much real users are satisfiedwith pilot test application of the initial game concept and 2) to correct and completeprototype properly. At the first user study, 6 people tested the prototype and 9 peopleparticipated in second one. User studies consist of user test in the fild and focusgroup interview. In order to verify effectiveness of the prototype, additional devel-opment was conducted for control group.After twice user studies, the result was evaluated by the interview. I found severalinsigts and possibilities. Participants had an enjoyment experience by goal setting(conquer territory, obtain items in time), achievement of misson and record boardfor self-regulation. Through the result of effectiveness, the propotype was verifiedby questionnaire with five veriables (perceived usefulness, reliability, enjoyment,continuous use, self-efficacy) among six veriables. Finally, we could say the gamethrough the design strategies we poposed plays an important role in trigger of pro-moting physical activity and being motivated. 39
Study 지도교수 학위논문 졸업연구 원광연 남주한 석사과정 이소형 A Study on the Interaction Model for the Interpre- tive Exhibition Design As exhibition plays a role for communication method between visitors and exhib- its, visitor’s experience become more important in exhibition design. In addition, to convey contents and meanings behind exhibits more effectively, systematic ap- proaches for interpretive exhibition are needed. In this study, I proposed an interac- tion model appropriate for interpretive exhibition and possibility of its application was reviewed. For this purpose types of visitors were analyzed and interactions in exhibition were studied. Exhibition that reflected suggested interaction model was designed and it was used for user test. From this I could recognize preferred pattern of interaction differs depending on type of visitor. The result of this study can be utilized to induce more diversified participation of visitors in designing interpretive exhibition based on the interaction model.40
지도교수 Study 학위논문 전봉관 졸업연구 석사과정 이수지 Study on the Method of Rewarding Participants for Enhancing the Interactivity between TV Programs and ViewersRecently, TV is losing the loyalty of viewers because of the competition with othermedia. Because of this issue, TV producers planned the interactive TV programsfor interaction between TV programs and viewers through viewer participation.However, these interactive TV programs generally gave material rewards for theirparticipants, but very few winner were able to receive the rewards and the other par-ticipants became dropouts. This study points out this limitation and suggests a com-plementary method through gamification rewards which can be produced infinitelyand provided to all participants. To make these gamification rewards become valid,they should satisfy the participants’ psychological needs. Based on these ideas, thisstudy conducted two procedures. In the first, a survey was designed based on PLEXframework which is a theory about playfulness of game, and the four elements ofparticipants’ common needs were extracted by the survey. In the second, a experi-ment was conducted and the subjects got three gamification reward types after par-ticipating in TV programs. After then, there was a analysis progress to figure outhow each reward types enhance four elements of playfulness in viewer participat-ing situation. Based on the result, this study suggests a guideline for rewarding allparticipants. This method is expected to contribute to help TV producers to build aenhanced interactivity when they want to plan interactive TV programs. 41
Study 지도교수 학위논문 졸업연구 도영임 석사과정 이슬 Phenomenological Approach to Parent-Child Con- flict in Relation to Adolescents’ Smartphone Use: Deriving Conflict Structure Model and Designing Conflict Reduction Activities Adolescents’ abusive smartphone use is becoming a growing social concern because it has potential in causing problems in adolescents’ health, development and social relationships. However, smartphone is a major communication channel for youth and gives them diverse benefits and opportunities to achieve developmental tasks. Thus, creating environment that can help users to build healthy smartphone usage habits is necessary, rather than treating it as problematic use. Parent-child relation- ship is an important factor influencing adolescents’ smartphone use. In this research, we aim to understand the parent-child conflict in relation to smart- phone use and derive the conflict reduction activities as a method to enhance rela- tionship between adolescents and their parents. We built an intervention program including conflict reduction activities based on a focused group interview upon mother-child dyadic that experience smartphone use conflict. The core activity was making family smartphone use rules through discussion and keeping the rules among all family members. Lectures about adolescents’ cyber culture, parental mediation methods and positive use of smartphone were included. Nine pairs participated in the program. We derived a structure model for parent-child conflict in relation to smartphone use analyzing the participants’ experience data with qualitative meth- ods. Also, the behavior directions for parent-child conflict reduction and implica- tions for intervention program design were suggested. This study is meaningful in that view the parent-child conflict as an intervention point for improving parent-child relationship and promoting adolescents’ beneficial smartphone use. Also, deriving the structure model for parent-child conflict about smartphone use through phenomenological approach and designing conflict reduc- tion activities based on experience analysis is significant. The structure model for parent-child conflict in relation to smartphone and behavior directions for conflict reduction can be used to design intervention programs. However, limitations of this study are due to its small-scale qualitative research method, which is difficult to generalize the results. The validation of the results with additional research and quantitative analysis is required. Furthermore, developing analysis instrument about awareness and behavior changes of participants, a follow- up study to evaluate effectiveness of the intervention and the influence of other fac- tors such as siblings and peers are important issues for further studies.42
지도교수 Study 학위논문 박현욱 졸업연구 석사과정 이용진 Mammographic Mass Detection by Adaptive Thresh- olding and Region GrowingThis research starts with a simple curiosity about spatial patterns of users in homescreens. Home screens that are accessible via displays of smartphones are a kindof mediated spaces and those are experienced most frequently by users nowadays.Therefore, this research aims to investigate the spatial pattern of how the smartphoneusers conceptualize and categorize the home screens, as a mediated space. Due to itsmobility, a smartphone has a relatively small sized screen and the number of iconsarrangeable is limited. Users contemplate how they will arrange their home screensby making folders and etc., for efficient access to the applications. In other words,home screens reflect users’ mental models and therefore seem to be a meaningful re-search subject. As a sort of mediated spaces, home screens are studied through prox-emics research methodologies. The proxemics methods, which were used by Hall tofind out people’s spatial patterns in real spaces, were also applied in this research inorder to investigate those in the home screens. Prior to the application of the meth-odologies, the differences between a real space and a mediated space are concernedand reflected into reconstruction of the methodology framework. 43
Study 지도교수 학위논문 졸업연구 이지현 석사과정 이지호 A Comparative Study on Thematic Building Colors based on Image Analysis – Focusing on Regional and Cultural Difference in Global Theme Parks. Gradu- ate School of Culture Technology The notion of glocalization has become widely spread in various fields, and ac- cordingly, it is increasingly more important to take account of indigenous cultural characteristics in each field. An aspect of achieving glocalization, globalization with local consideration, is to consider the difference of color usages and environmental atmosphere between distinct cultures. This study aims to find cultural and locational differences with the color usages under the condition of equivalent Disney themes but different cultural circumstances—eastern (Tokyo Disneyland) and western (Disneyland Paris), and identify whether if there is a relationship between the used colors and the local urban environment. This study suggests an approach to investi- gate the color differences between eastern and western cultures with the case analy- sis of building façade colors in Disneyland Paris and Tokyo Disneyland. To do this, digital image based k-means clustering algorithm, CIELAB color space and Natural Color System (NCS) was used to explore color characteristics. Accordingly, Tokyo and Paris urban color usage characteristics were analyzed to investigate cultural connections. This study found out that there is a difference with the color usages between eastern and western located theme parks and in addition, there is a con- nection between the colors used in local theme parks and the local surrounding en- vironment in terms of color hue and nuance. Disneyland Paris used relatively calm and deep colors while Tokyo Disneyland used relatively active and bright colors for their attractions. In the same vein, environmental colors of the urbanscape also ap- peared similar features and color usage characteristics. Whether it is intended or not, both of the theme parks considerably showed correspondence with the overall urban environmental color usages. The major motivation of this paper was to reflect the atmosphere and the mood of the space that can be easily felt but not readily express- ible into a cultural color palette. By finding the color differences, this study suggest a strategy for designers to consider local colors from environmental clues when designing themed environments to adapt into their themed environment design.44
지도교수 Study 학위논문 원광연 졸업연구 석사과정 정영헌 User Interaction and Interface Design in Collabora- tive Environment using Tiled-displayFor the past few decades, computing power technology has seen a remarkable in-crease in performance, but display resolution technology has not advanced at thesame rate. So, there has been research about tiled display system which can be usedas a screen by combining multiple displays in specific configuration for overcom-ing limitations of single display in size and resolution. However, tiled display sys-tem has been used mainly by expert people in restricted environments because ofcomplex architecture of it. These days applicability of tiled display is increased inseveral area such as education, medical facilities, exhibition, etc. In addition, severalkinds of mobile devices become popular and then, there has been active researchabout interaction user and tiled display through mobile devices in environment tointegrate heterogeneous mixture of displays.In this paper, this research considers tiled display system as an environment whichensures users can interact with others as well as a platform to visualize informationand suggests a method of interaction among members in collaborative environment.Also, this research aims to verify usefulness of this interaction method through usercase study using prototype application. For that, this research classified interactionpatterns into three categories depending on the ways of sharing information in col-laborative environment using tiled display. After that, this research extracted interac-tion elements, wrote scenarios to reflect them through analyzing related studies andusage of tiled display, and designed interaction model.Interaction among multiple members who use several devices in collaborative en-vironment that includes heterogeneous displays are studied in recent research. Inthis point, this research classified members of collaborative environment accordingto roles and designed user interface and method of interaction, and furthermore,suggested research direction for efficient information visualization on large display. 45
Study 지도교수 학위논문 졸업연구 조경민 석사과정 노준용 Outfielder Trajectory Synthesis based on Psycho- logical Models In this paper, we propose a novel model to generate a realistic outfielder trajectory. To achieve this, we modeled the perception of outfielder based on the simplified vi- sion model. Also, to catch the details of human behavior, we modeled the behavior algorithm based on the result of experimental psychology researches. In the result, we show that the crude perception of the outfielder can generate similar trajectory that is shown in real world and thus, improve the realism. Unlike previous research, our model can generate various type of trajectory including the trajectory waiting at the landing position of the ball, depending on the accuracy of perception.46
지도교수 Study 학위논문 전봉관 졸업연구 석사과정 조수연Affective and Cognitive Dimensions Correlated toAbstract Movement Features Extracted from Mod-ern Dance This study aims to define online memorial museum and visitors’ museum experience by investigating the effect between visitors’ satisfaction, characteristic of exhibition, and their visiting motivation. The definition of online memorial museum and characteristics of exhibition and visitors motivations are proposed based on previous related research on online museum, online memorial, and memorial museum. With the refined definition, four characteristics of exhibition — authority, timeliness, flexibility and co-production — and variables for classifying types of visits — personal interest, specific research, remembrance, pilgrimage and casual browsing — are selected based on related researches. In order to see the reliability of the above factors, visitors of ‘e-museum of the Japanese military sexual slavery’ operated by ministry of gender equality and family republic of Korea were asked to participate in a survey. Analysis from the survey result of 209 visitors shows that visitors’ satisfaction is affected by theses characteristics of exhibition directly. Furthermore the result states that the visiting types can be clustered by motivations — explorer, commemorator, recharger and researcher — and these types do play a role as a moderator in some of the relationship between the characteristic of exhibition and visitors’ satisfaction — timeliness and flexibility. Through such empirical research, this study provides a more specific room for researches on online memorial museum and its visitors. In addition, this study can give practical advice when operating online memorial museums. 47
Study 지도교수 학위논문 졸업연구 이성희 석사과정 진태일 A Motion Retargeting Method to Preserve Spatial Relationship of Skinned Characters Motion retargeting algorithm can be used for animating character motion sequences to various characters. Several motion retargeting methods retarget interaction mo- tions that contact interaction and close interaction. Among motion retargeting algo- rithms, representing spatial relationships between body parts of single or multiple articulated characters is an efficient approach to capture interaction motion’s scene semantics. This paper presents a new method for retargeting interaction motions. We introduce a simple mesh structure called a space mesh to represent surface-level spatial relationships. Using our method, we can retarget original character interac- tion motions to target character preserved surface-level spatial relationships while reducing the number of inappropriate character skin penetrations48
지도교수 Study 학위논문 김원준 졸업연구 석사과정 허경석The Dynamics of Online Music Industry: The Effectof Mobile Technology on Streaming and Download-ing Due to the rapid technological innovation, online music industry is undergoing sig- nificant changes in their music service platforms. Especially, due to the fast penetra- tion of smart devices and mobile service technology such as LTE, more and more music streaming services are wide spread in current music market. However, there is no study that examines how this introduction of new technology affects online music consumption, although new technology introduction is significantly affecting music industry’s value chain, pricing, market shares, etc. In this study, we examine whether music streaming service a substitute or complement for music download- ing service considering the moderating effects of mobile technology. We use data- mining techniques to collect unique dataset on online music market for 248 weeks in Korea. The results suggest that music downloading service is complement to music streaming, however, this complementary relationship is negatively moderated by the LTE service. In addition, major distribution and major genre music are positively related with streaming service. On the one hand, broadcasting system is positively related with downloading, but price policy decreased downloading. In other words, consumers listen to music what is exposed and promoted through the media. 49
Study 지도교수 학위논문 졸업연구 김원준 석사과정 Christopher John CLugston The Unintended Consequences of eBay v. MercEx- change: An Empirical Study of the Effects on Health-Related Industries The Federal Circuit has historically been seen as pro-patent. As such, the court de- veloped a general rule that injunctions should issue once infringement and validity had been established. However, around the turn of this century, the ease with which patent owners could obtain injunctions was coming under increasing scrutiny. In Congress, the Patent Reform Act was poised to limit the use of injunctions, and in the public sphere, the Blackberry case had raised awareness of the issue. In front of this backdrop, the eBay case worked its way through the judicial system. The Supreme Court would ultimately take up the issue, dramatically changing the law in the process. This article empirically evaluates the aftermath of the eBay decision, looking at nine years of post-eBay injunction decisions. Numerous early studies were done to evaluate the effect of the eBay decision on NPEs. This study confirms the findings in the earlier literature that NPE plaintiffs are significantly disadvan- taged. More interestingly, our results also reveal that firms in health related indus- tries are likewise granted injunctions at significantly lower rates.50
Search
Read the Text Version
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10
- 11
- 12
- 13
- 14
- 15
- 16
- 17
- 18
- 19
- 20
- 21
- 22
- 23
- 24
- 25
- 26
- 27
- 28
- 29
- 30
- 31
- 32
- 33
- 34
- 35
- 36
- 37
- 38
- 39
- 40
- 41
- 42
- 43
- 44
- 45
- 46
- 47
- 48
- 49
- 50
- 51
- 52
- 53
- 54
- 55
- 56
- 57
- 58
- 59
- 60
- 61
- 62
- 63
- 64
- 65
- 66
- 67
- 68
- 69
- 70
- 71
- 72
- 73
- 74
- 75
- 76
- 77
- 78
- 79
- 80
- 81
- 82
- 83
- 84
- 85
- 86
- 87
- 88
- 89
- 90
- 91
- 92
- 93
- 94
- 95
- 96
- 97
- 98
- 99
- 100
- 101
- 102
- 103
- 104
- 105
- 106
- 107
- 108
- 109
- 110
- 111
- 112
- 113
- 114
- 115
- 116
- 117
- 118
- 119
- 120
- 121
- 122
- 123
- 124
- 125
- 126
- 127
- 128
- 129
- 130
- 131
- 132
- 133
- 134
- 135
- 136
- 137
- 138
- 139
- 140
- 141
- 142
- 143
- 144
- 145
- 146
- 147
- 148
- 149
- 150