A computational model of dynamic *perceptual attention for virtual humans

Modeling interactive virtual humans has been one of the primary goals of immersive virtual environments for training. An important characteristic of a virtual human is the ability to direct its perceptual attention to objects and locations in a virtual environment in a manner that looks believable and serves a functional purpose. Not only must the virtual human pay attention to objects related to the tasks it is performing, but it must also be able to cope with sudden events that demand attention. It is often the case that the amount of information in the virtual environment far exceeds the processing abilities of the virtual human. In fact, only a small fraction of sensory information can be fully processed and assimilated into the cognitive model. A successful model of perceptual attention provides a way of both pruning the incoming sensory input and choosing the most salient information to focus on during the next step of a decision cycle. In this dissertation, we present a computational model of perceptual attention for virtual humans. This model extends a prior model of perceptual resolution [39][40] based on psychological theories of human perception. This model allows virtual humans to dynamically interact with objects and other individuals, balancing the demands of goal-directed behavior with those of attending to novel stimuli.