Disambiguating ambiguous figures by a model of selective attention.

Ambiguous figures are visual stimuli which are interpreted multiply by the human visual system. A model is proposed which disambiguates the ambiguous figures. The model was formulated based on the characteristics of visual information processing, accompanied with selective attention. In the ambiguous figure "my husband and my father-in-law", it was necessary to simulate visual information processing so that attention was directed to the multiple features in the figure to disambiguate the ambiguous figure. Pictures, obtained from the model, were examined as to whether they were interpreted unambiguously or not. Results show that the model, simulated selective attention, can disambiguate the ambiguous figures. This suggests that the image per se, viewed through selective attention, becomes unambiguous before the figure is interpreted in the higher level. Results also show that the computer simulation of selective attention would make it possible to examine factors affecting the initial interpretation of the figure.