메뉴 건너뛰기
소속 기관 / 학교 인증
인증하면 논문, 학술자료 등을  무료로 열람할 수 있어요.
한국대학교, 누리자동차, 시립도서관 등 나의 기관을 확인해보세요
(국내 대학 90% 이상 구독 중)
고객센터 ENG
주제분류

논문 기본 정보

저자정보
(Chung-Ang University) (Chung-Ang University) (Chung-Ang University) (Chung-Ang University) (Chung-Ang University)
저널정보
중앙대학교 영상콘텐츠융합연구소MINT Moving Image & Technology (MINT) MINT: Moving Image & Technology, Vol.1, No.1
오류 신고하기

피인용 0

검색

    초록·키워드

    The human activity recognition (HAR) technique is important for human-computer interaction (HCI) systems. Current HAR systems commonly utilize deep learning algorithms. Although these technologies can achieve higher accuracy, they require numerous motion datasets for the training process. Thus, we propose a method that can generate motion data in a short period of time. Our system can generate a large amount of motion data and edit it for the user"s intended target. The system transforms the data based on Motion-sphere trajectory data. Our system requires the first input data from a motion-capture system. The data vary and change depending on the amount of motion data. We showed the motion generated by the Motion-sphere’s trajectory and the virtual avatar’s motion. In addition, we combined our system with an authoring method based on anatomical knowledge. We were able to generate hook motions that hit the user’s desired target by combining the two methods. Our system can be applied to generate motion data and edit character motions in animation fields.

    최근 본 자료 전체보기

      UCI(KEPA) : I410-ECN-0101-2022-688-001072320