engleski

Adaptive stimulation of emotional system based on virtual reality

Certain areas of human activity, like training for stressful tasks, the psychotherapy of stress-related disorders, or entertainment, seem to endorse targeted modulation of human emotional states. Modulation may be implicit (e.g. contained in the plot of a movie or a computer game), or actively conducted by the qualified domain experts (e.g. therapists in psychotherapy). Affective computing paradigm, in which the computers attempt to recognize and appropriately respond to the user’s emotions, is also potentially applicable to the modulation of emotional states in these areas. In order to facilitate the development of such affective computing applications, the thesis proposes a generic model and architecture of the so-called “adaptive stimulation of emotional system, ” i.e. a personalized delivery of multimedia or synthetic virtual stimuli to the user by the computer, based on various manifestations of the user’s emotional state (e.g. physiological, vocal and facial). Potential forms of collaboration between the domain experts and the computer that may be needed in different applications are structured on the architectural level. Following the application-independent architectural exposition, application-specific stimuli delivery algorithm driven by the user’s physiological signals is outlined, in the context of training for improving the user’s stress resilience. A preliminary application for the automated physiology-driven modulation of the user’s arousal is also described, which has been developed from the proposed architecture.

adaptive stimulation of emotional system; virtual reality; affective computing; stimuli; emotional state; valence and arousal; physiology; amygdala; stress inoculation training; stress resilience

nije evidentirano