SEGMENTATION BY A NETWORK OF OSCILLATORS WITH STORED MEMORIES

We propose a model of coupled oscillators with noise that performs seg mentation of stimuli using a set of stored images, each consisting of objects and a background. The oscillators' amplitudes encode the spati al and featural distribution of the external stimulus. The coherence o f their phases signifies their belonging to the same object. In the le arning stage, the couplings between phases are modified in a Hebb-like manner. By mean-field analysis and simulations, we show that an exter nal stimulus whose local features resemble those of one or several of the stored objects generates a selective phase coherence that represen ts the stored pattern of segmentation.