In this paper we describe an experiment aimed at determining the most effective and natural orientation of a virtual guide that gives route directions in a 3D virtual environment. We hypothesized that, due to the presence of mirrored gestures, having the route provider directly face the route seeker would result in a less effective and less natural route description than having the route provider adapt his orientation to that of the route seeker. To compare the effectiveness of the different orientations, after having received a route description the participants in our experiment had to ‘virtually’ traverse the route using prerecorded route segments. The results showed no difference in effectiveness between the two orientations, but suggested that the orientation where the speaker directly faces the route seeker is more natural.
|Name||ACL Workshop Proceedings|
|Publisher||Association for Computational Linguistics|
|Workshop||ACL Workshop on Embodied Language Processing 2007|
|Period||28/06/07 → 28/06/07|
- HMI-IA: Intelligent Agents