ABSTRACT
Visual masking refers to the reduced visibility of one stimulus, called the target, due to the presence of another stimulus, called the mask. When the mask is presented before the target, masking is said to be forward masking. Multiple factors can influence visual masking. Unfortunately, due to limited computational power, spatial representations in quantitative models of masking have been quite limited. Mathematical and engineering results show that it is extremely difficult to control the dynamics of systems with positive nonlinear feedback. Many of the building blocks of mathematical neuroscience were already laid out in 1950s and 1960s. The RECOD model suggests that cortical networks avoid undesirable unstable behavior by operating in a succession of transient regimes. The RECOD model has been used to explain existing masking data as well as to make new predictions for empirical testing. Furthermore, since RECOD is expressed mathematically, it allowed quantitative simulations that can be directly compared to empirical data.
