Validation of Virtual Humanoid Intelligent Agents in Virtual Reality Systems

摘要

One of the great benefits VR systems offer is their ability tosimulate a number of virtual humans when their presence isneeded in the context of some learning or training experience.Being that the real humans may not be available to play differentroles and support virtual sessions, the ability of a system togenerate highly believable representations of autonomous virtualhumans - virtual intelligent agents - is vital in achieving specificlearning and training objectives. Eliminating the elements of thesystem that can cause a negative learning and training transfer is aparamount in those systems. We illustrate the results of two userstudies focused on validation of non-deterministic domainspecificbehaviors generated by our system (example: behaviorstypical for a well coordinated group of paramedics or militaryunit). The results and observations confirmed that when it comesto VR systems with stringent requirements and high expectationsfor positive learning/training transfer, we still need humans toevaluate and validate synthesized human-like agent behaviors.