An adaptive fuzzy expert system to evaluate human visual performance

摘要

In this work, we present a fuzzy expert system to efficiently evaluate the performance of human observers in a visual detection task. With this system, based on fuzzy logic and the receiver operating characteristic theory, we reduced the duration of visual perception psychophysical experiments, which normally takes too long, to approximately one fourth. We used as input variables the partial performance indexes obtained along each experiment together with the corresponding increments of them. The output variable is the energy of the signal to be detected, which is adjusted adaptively by our system so that the observer reports approximately 80% of correct responses. With shorter adaptive experiments, we eliminate the effects of fatigue and loss of attention that limit considerably the applicability of the related theories and techniques. We also show that our results are equivalent to those obtained through long-lasting traditional psychophysical experiments.