Exploring costume-avatar interaction in digital dance experiences

Abstract

Dance in the digital world is an area that has been explored vividly the last years with applications in education/learning, cultural heritage preservation, archiving and performance. Digital costume is a concept complimentary to dance that also evolves as technology grows, with virtual fashion shows, digital-only clothing, virtual showrooms, and recreation of historic costumes to make their appearance more and more often. However, their intersection hasn’t been quite explored yet although it has a lot to offer in areas like digital fashion design, animation, games and artistic performances. We envisage that combining realistically simulated clothes along with motion data can add a significant potential to movement research. Specifically for learning or exploring movement in digital environments, the representation of the dancer (3d avatar) is an important aspect to be examined, for example, specific digital clothes might enhance movement understanding and thus affect learning, in some cases. Having the hypothesis that different avatar and/or cloth visualizations might lead to different results in movement, an experiment with users has been conducted in an immersive Virtual Reality environment. We setup a testbed environment, where avatars of different forms (abstract, anthropomorphic, robot) and with different costumes could perform dances. The users were immersed in the environment and had to repeatedly follow the dance steps with all avatar-costume combinations. Their performance was then rated by experts regarding the general quality of movement, the rhythm, and the steps. The analysis of these results provided insights on the effect of costumes and avatar types on the above parameters and along with qualitative aspects, wishes to initiate a study around the limitations and opportunities of combining costume and dance in digital, interactive environments.