METHOD AND APPARATUS TO CONTROL RESPONSIVE ACTION BY A COMPUTER-GENERATED CHARACTER

摘要

A computer-generated character is described having the capability to respond to one or more present object images (933). In one embodiment, the character has a field of vision (e.g., a cone emanating from the character's eyes). Thus, the character will respond only to object images that are in the field of vision. The character's responses can be, for example, movement (i.e., orienting the character's eyes or body toward the identified object image) or displaying emotion (e.g., showing surprise or sadness) (935). If several object images are in the field of vision, an importance parameter (937) can be assigned to each object image so that the character responds to the image having the highest such parameter. The importance parameter can be, for example, proportional to the proximity of the object image to the character or proportional to the velocity of the object image.