REALTIME PARALLEL AND COOPERATIVE RECOGNITION OF HUMAN FACE FOR A NATURALISTIC VISUAL HUMAN INTERFACE

The concept of a Visual Software Agent (VSA), i.e., an anthropomorphic agent possessing a realistic human form and communication functions t hat create an interface with the user, is proposed. Research is procee ding on its implementation on a parallel computer. In this paper, we d escribe a method that is effective as the visual input function of the anthropomorphic agent. This method recognizes and extracts only one p erson (user) directly facing the camera in a moving image in which sev eral people occasionally enter and leave. The recognition method can b e executed in real time in a typical indoor environment. The backgroun d and the lighting do not have to be restricted and recognition can be achieved in real time by a parallel computer. In addition, this techn ique simulates the initial stage of human visual functions and is base d on the parallel extraction and integration of multiple image feature s. The movements of visual attention of cognitive psychology can also be simulated. An anthropomorphic agent derived from this approach succ essively looks at multiple human faces, fixes its gaze only on a speci fic person, and synthesizes agent behavior that approximate that of hu mans. These moving image recognition and synthesis processes were impl emented on a newly designed parallel computer.