>>14408977I suffer from long-term sleep deprivation which is keeping me from my research focus: evolutionary constraints upon visual area architecture. There are tons of interesting facets and phenomena that come into play.
The first stage of human visual analysis, V1, is structured kind of like an organic analogue to a TV screen. Instead of variations of combinations between red, blue and green, there are variations of combinations between shape orientations, ranging in various degrees between vertical, horizontal and diagonal. The very simplest of cells that are sensitive to a single orientation only, pool together their outputs to more complex cells that represent simple combinations of these orientations, which then combine to more compex combinations of simple combinations, on and on until you have actual objects.
The actual basic principles involved are insanely simple, MAX-like operations with weighted inputs, a distorted hexagonal lattice of orientation preference columns (structures of V1 cells with large clusters of specific orientation sensitive cells), projection coordinate translations between retinal data and the retinotopic map... When you look at it the structures are incredibly complex but they flow from very straightforward and logical principles.
I wish I knew more about the actual physics of how light itself behaves that I could incorporate into my studies, but first I need to find a way to sleep unfortunately.
