A Neurocomputational Model of Figure-Ground Discrimination by Relative Motion

A neurocomputational model is presented for gure-ground discrimination by relative motion. In the model, the correlation type of elementary motion detectors (EMDs), an oscillatory neural network for motion perception and a selective attention module are involved. It is shown that through the oscillatory amplitude and frequency encoding and selective synchronization of phase oscillators, a perfectly camou-aged textured gure can be successfully discriminated from its identically textured background by relative motion alone. network (ONN), multilayer network of phase oscillators.