Neural data and models have proposed that the brain achieves invariant object recognition by learning and combining several views of a three-dimensional object.
How are such invariant codes learned when active eye movements scan a scene, given that the cortical magnification introduces a large source of variability in the visual representation even for the same view of the object? How does the brain avoid the problem of erroneously classifying together parts of different objects when an eye movement changes its cortical representation from one to the other? How does the brain differentiate between saccades on the same object and saccades between different objects?
The space variant representation of the scene in the brain degrades information in the periphery and enhances it in the fovea.
Cortical magnification represents foveal stimuli with much higher resolution than those in the periphery (Eric Schwartz)
Our biologically inspired ARTSCAN model of visual object learning and recognition with active eye movements proposes answers to these questions. The model explains how surface attention interacts with eye movement generating modules and object recognition modules so that the views that correspond to the same object are selectively clustered together. This interaction does not require prior knowledge of the object identity. The modules in the model conform to modules in What and Where streams of the visual system. The What stream learns a spatially-invariant and size-invariant representation of an object, using bottom-up filtering and top down attentional mechanisms. The Where stream computes indices of object location and guides eyes movements. Preprocessing is assumed to happen in the primary visual areas, notably log-polar compression of the periphery, contrast enhancement and parallel processing of boundary and surface properties. The WHAT stream is a variant of ARTMAP classifiers designed by Gail Carpenter and Stephen Grossberg.
