Plug and Play Cultural Avatars for Training

Immersive training environments tend to be good at graphics, but lack realistic cultural content, especially in regard to highly interactive cultural characters. There is a growing gap created by the ever-increasing need for realistic cultural training and the lack of availability of interactive cultural training environments. There are a number of ongoing efforts aimed at building cultural models. However, these models often suffer from low fidelity expression of cultural behavior, are typically tied to particular virtual environment platforms, and lack natural modes of interaction with trainees. With our partners at Vcom3D, we have been developing portable cultural avatars that combine the cognitive and physical aspects of cultural behavior, can generate realistic facial expressions and cultural gestures, and which have been inserted in a number of virtual environments. We propose to extend this existing work to include (1) new input and output modalities (e.g., gesture and speech), (2) scenario-building tools, and (3) after-action review capabilities, to simultaneously increase the state of the art in interactive cultural training while decreasing the cost to build and deploy those models in realistic training scenarios.