The paper employs fuzzy automata for emotion control by audio-visual stimulus. In the proposed design, emotions are represented by states, and transition of emotions is controlled by selective audio-visual stimulus. Experiments are undertaken to determine the fuzzy membership for state transitions in the automata. A fuzzy relational algebra has been used to determine the current membership distribution of the emotions from their previous value and fuzzy state transition memberships. Experiments with fifty audio-visual stimulus and one hundred subjects confirm that automatic control of emotion to a desired state from a given state can be attained with a success rate of approximately 86%.
展开▼
