http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
- 中文 을 입력하시려면 zhongwen을 입력하시고 space를누르시면됩니다.
- 北京 을 입력하시려면 beijing을 입력하시고 space를 누르시면 됩니다.
RISS 인기검색어
- 검색키워드
- 저자 : 전효경 (검색결과 4 건)
- 무료
- 기관 내 무료
- 유료
-
근로환경 변화에 따른 딥러닝 클라우드 서비스 기술을 활용한 캐릭터 애니메이션 리타겟팅에 관한 연구
전효경,조동민 한국멀티미디어학회 2023 멀티미디어학회논문지 Vol.26 No.6
The development of deep learning cloud service technology, along with the post-COVID-19 phenomenon of "mass retirement," has led to new changes in labor management. In the field of game production, it is necessary to study the improvement of the animation production process system according to the change in working methods. This study compares and analyzes the effects of retargeting three animation actions using deep learning cloud services DeepMotion and Mixamo and 3D production tool Blender by measuring manipulated variable factors and quality evaluation items. By comparing and analyzing the 3D animation production process in the same work-from-home environment, the results showed that the deep learning platforms provided higher quality output for the same amount of time, resulting in higher efficiency and productivity. In addition, each platform has its own advantages and disadvantages, and if utilized well, it can be an opportunity for companies and workers to introduce new work systems to promote employment stability in a changing work environment. The purpose of this study is to propose improvements to the working environment and work process system by introducing a new work system suitable for the working environment, solving problems in the existing character animation production process and increasing work efficiency and productivity for workers and companies developing game content.
-
전효경,조동민 한국만화애니메이션학회 2024 만화애니메이션연구 Vol.- No.74
There are many different ways to create character animations,depending on the purpose. Recently, the number of AI platform serviceshas increased exponentially, and utilizing AI services in 3D characteranimation is challenging. These challenges include four factors. First,there is a wide variety of characters. Second, there is a need to reducethe cost and time of animation production. Third, the need for humancharacteristics and multiple behaviors, and fourth, the limitations ofutilizing existing animation data. By adopting AI technology to respondto these challenges, the efficiency and versatility of character animationcreation, transformation, and application can be improved. Thepurpose of this study is to explore the extent to which AI animationretargeting technology can provide a more efficient process thantraditional methods. As a methodology, we created animations on threedifferent platforms to compare and analyze the differences betweenusing AI and not using AI. To verify this, we used the SPSS statisticalprogram to compare and analyze the difference in the average of thethree platforms. As a result, statistically significant differences werefound in how to create Model Upload, Character Setup, AnimationRetargeting or Keyframe depending on the platform. (p<0.05) Therefore,the null hypothesis was rejected and the alternative hypothesis wasaccepted, indicating that there is a difference in the mean betweenthe three platforms. In terms of animation retargeting and keyframeanimation creation time, Mixamo had the highest mean difference of 0.03 compared to Blender’s 28.72. Therefore, the most significantdifferences in animation production across platforms were found tobe the difference between animation retargeting and keyframing. Inother words, the use of AI’s animation retargeting technology providesa new model for improving the animation workflow and creating anew pipeline for future content production. This can be positioned as aconvergence content of human-machine cooperation in the future, andthe development and dissemination of such technology will lower thebarrier to entry for animation production technology and create a newjob group. 캐릭터 애니메이션의 제작방법에는 목적에 따라 다양한 방법이 있다. 최 근 AI 플랫폼 서비스가 기하급수적으로 늘어나고 있고 3D 캐릭터 애니메이 션 분야에서도 AI 서비스를 활용하는 것은 도전적인 측면에 기인한다. 이러 한 도전에는 4가지 요소가 포함되어 있다. 첫째, 다양한 캐릭터가 존재하고 둘 째, 애니메이션 제작 비용과 시간을 절감하는 효과가 있다. 셋째, 인간적인 특 성과 여러 가지 동작이 필요하며, 넷째, 기존의 애니메이션 데이터 활용의 한 계가 있다. 이러한 도전에 대응하기 위해 AI 기술을 도입함으로써, 캐릭터 애 니메이션의 생성, 변환, 적용에 대한 효율성과 다양성이 향상될 수 있다. 본 연 구목적은 AI 애니메이션 리타겟팅 기술을 활용했을 때 기존의 작업방식보다 얼마나 효율적인 프로세스를 제공할 수 있는지 탐구하고자 한다. 연구방법으 로는 AI를 사용한 경우와 사용하지 않은 경우 차이점을 비교 · 분석하기 위해 3개의 플랫폼으로 애니메이션을 제작하였다. 이를 검증하기 위해 SPSS 통계 프로그램을 활용하여 3개 플랫폼의 평균의 차이를 비교 · 분석하였다. 그 결 과 플랫폼에 따라서 모델 업로드(Model Upload), 캐릭터 셋업(Character Setup), 애 니메이션 리타겟팅 및 키프레임(Animation Retargeting or Keyframe) 제작 방법에서 통계적으로 유의한 차이가 파악되었다. (p<0.05) 따라서 귀무가설 기각, 대립 가설이 채택되어 3개의 플랫폼 간의 평균 차이가 있는 것으로 나타났다. 애니 메이션리타겟팅 및 키프레임 애니메이션 제작 시간에서는 믹사모가 0.03 블 렌더가 28.72로 평균의 차이가 가장 많이 났다. 따라서 플랫폼의 따른 애니메 이션 제작에 가장 큰 차이를 보이는 요소는 애니메이션리타겟팅과 키프레임 방식의 차이로 파악되었다. 즉, AI의 애니메이션 리타겟팅 기술의 활용은 애 니메이션 작업공정의 문제점을 개선하고 향후 콘텐츠 제작의 새로운 파이프 라인을 형성하는데 새로운 모델을 제시해 준다. 이것은 미래의 인간과 기계협 력이라는 융합콘텐츠로 자리매김할 수 있고 이러한 기술의 발전과 보급은 애 니메이션 제작기술의 진입장벽을 낮추어 새로운 직업군을 형성하는 계기가 된다.
-
강혜영,김미선,김정효,김혜숙,박미화,박영주,송남호,이난희,전효경 전남대학교 간호과학연구소 2001 Nursing and Health Issues(NHI) Vol.6 No.1
The purpose of this study was to investigate public nurses' image from the perspectives of girl ’s high school teacher (n=53), broadcaster (n=50) , legal officer (n=30) and administrative officer (n=27) at G city from June 20th to July 20th, 2001. The instrument for this study was the 4-point Likert scale using 30 items about nurses' image developed by Jahang (1998) and the scale was divided into four areas: temperament as a nurse (Cronbach’s α=.72), role performance (Cronbach’s α =.55), community participation (Cronbach’s α=.74), interpersonal relationship (Cronbach’s α=40). The data were analyzed by percentage, t-test and ANOVA using SPSS/WIN program 10.0, and the results could be summarized as follows: 1. The most influencing factor on nurses' image building was nurses working in hospital (75.5%). The average score of image was 2.66 out of 4.00 points and 'temperament as a nurse' was 2.80, ’community participation' 2.64 and 'role performance' 2.54 points in subcategories. 2. According to the backgrounds of subjects, the scores on nurses’s image were significantly different: to career, teacher 2.78, administrative officer 2.64, broadcaster 2.58 and legal officer 2.57 points (F=5.47, p=.OOl); and to age, in the fifties was 2.84, the forties 2.75, the twenties 2.65 and the thirties 2.56 points in order (F=5.23, p=.002). And the scores on nurses' image were not significantly different according to other characteristics of the subjects. ßased on these results, the authors recommended as follows: 1. Follow-up study on nurses' image from the perspectives of inpatients, outpatients and their families in various hospital settings and clients in other nurse-working fields. 2. Follow-up study on nurses' image from the perspectives of teachers working at from kindergarten to high school; and a study to identify age-specific nurses' image of the public. 3. To develop strategies for affirmative self image building of nurses as a health professional and for mass media monitoring.
-
RISS연관 검색어 추천
이 검색어로 많이 본 자료
활용도 높은 자료
