V4_HMax
This model captures some aspects of the tuning of V4 neurons to
boundary curvature. It is very similar to that proposed by Cadieu et
al. (2007) to explain the responses of V4 neurons to simple closed
shape stimuli (Pasupathy and Connor, 2001). The architecture of the
model is a refinement of the hierarchical alternation between simple
(S) and complex (C) cell layers described by Riesenhuber and Poggio
(1999). This is a strictly feedforward model, and it has no temporal
processing. Thus, a single response values is associated with a given
input image.