Retinal neurocomputing principles for real time track identification

Important theoretical advances in quantifying retinal neurocomputing principles lead to strong predictions on the architecture and functionality of the visual pathway. Spatial, binocular and time decorrelation as well as redundancy reduction principles lead to efficient multiscale information coding and transmission through the multilevel network. Predicted functionalities like the on-center, off-center cells, even- and odd-symmetric simple cells can be abstracted to be used for real time track-vertex identification on particle physics data.