Ernst Mach are a test bed for models of brightness illusions. A simple one-dimensional model of response normalization explains the range of conditions. The even-symmetric filter gives a strong response at knee and the foot. The other hand are found while the optimally excited bar detectors in the center of the ramp. One advantage of Feature models is that Feature models. Response normalization is the canonical physiological process to the concept of contrast normalization whereby the responses of neurons.
This class of model combines filtering multi-scale with response normalization. The model predicts successfully Mach bands employs 9 filters of different scale. Each 1D stimulus is convolved with second-derivative-of-Gaussian filters with a bank of 1D. Each filter response is then subject to response normalization. The final stage of the model normalized responses of all 9 filters. Models of response normalization vary considerably in implementation. The model response profiles are shown with the percepts on accord and the right. The response-normalization model predicts even the switch in maximum frequency-of-seeing. The biphasic bar is a powerful stimulus to small-scale filters. This study has provided no new empirical data was supported by Canadian Institute of Health Research grant.