A lattice filter model of the visual pathway

Early stages of visual processing are thought to decorrelate, or whiten, the incoming temporally varying signals. Motivated by the cascade structure of the visual pathway (retina → LGN → V1) we propose to model its function using lattice filters signal processing devices for stage-wise decorrelation of temporal signals. Lattice filter models make predictions of neuronal dynamics which are consistent with physiological measurements in cats and primates. In particular, they predict various shapes of temporal receptive fields that agree with those reported for nonlagged and lagged cells in the LGN. Moreover, connection weights in the lattice filter can be learned using Hebbian rules in a stage-wise sequential manner reminiscent of the neuro-developmental sequence in mammals. In addition, visual processing in insects may be modeled by lattice filters. Therefore, lattice filter is a useful abstraction that may help unravel visual system function.