We propose that the low-pass characteristics in the temporal and velocity domain of area 17 cells are generated by the abundant excitatory connections between cortical neurons. We have incorporated this anatomical feature in a model circuit in which simple cells' firing is initiated by geniculocortical excitatory synaptic input, with a short time course, and firing is maintained by feedback corticocortical excitatory synapses, which have a longer time course. The low-pass performance of the model is demonstrated by computing the model simple cells' velocity response curves (VRC) elicited by moving bars, and comparing these to those of its LGN (lateral geniculate nucleus) inputs. For the same parameter set, the VRCs of sustained and transient LGN cells are transformed into VRCs typical of central area 17 and central area 18 cells, respectively.

We trained a multilayer perceptron with backpropagation to perform stimulus orientation discrimination at multiple references using biologically plausible values as input and output. Hidden units are necessary for good performance only when the network must operate at multiple reference orientations. The orientation tuning curves of the hidden units change with reference. Our results suggest that at least for simple parameter discriminations such as orientation discrimination, one of the main functions of further processing in the visual system beyond striate cortex is to combine signals representing stimulus and reference.