Recently, a virtual human, which provides natural interaction with real human, has acquired lifelike motion patterns with evolving motion capture devices. However, a variety of the motion patterns is restricted because of the difficulty of actual capturing processes. Also, the virtual human has insufficient presence because of a lack of physical effects represented by force feedback. These limitations make the virtual human an artificial humanoid. Therefore, in this research, we propose a reactive virtual human with human scale haptic interactions for triggering user’s natural reactions. This system also achieves a self-learning motion database for automatically getting various motion patterns from its users.
Unfortunately, ACM prohibits us from displaying non-influential references for this paper.
To see the full reference list, please visit http://dl.acm.org/citation.cfm?id=1179695.