The temporal dynamics of brightness filling-in.

Karl Frederick Arrington

Vision Research,
1994; 34(24), p.3371-3387.

Download PDF -
Play movie

Abstract

The temporal dynamics of brightness filling-in were studied throughneural network simulation experiments conducted under visual maskingstimulus conditions. Grossberg and colleagues have specified afilling-in model called the Boundary Contour System / Feature ContourSystem (BCS/FCS). The BCS generates boundary segmentation, while theFCS fills-in surface feature within these segmentation boundaries.Simulation experiments demonstrate that the model accurately predictsthat area-suppression follows a U-shaped function of forward masking(Stoper & Mansfield, 1978) and demonstrate that the psychophysicalfindings of Paradiso and Nakayama (1991) which they regarded as beingagainst the BCS/FCS model, actually support the model.

Errata

In Equation 2 the depolarization term should be changed from (-e_ij) to (+i_ij), and the hyperpolarization term should be changed from (-i_ij) to (+e_ij). Note that Equation 2 is never actually used in the simulations, as explained at Equation 10. This correction to Equation 2 has been made in the above PDF file.

Key Terms

brightness, filling-in, masking, neural network, area-suppression

Return to CV.