This scaling variation method allows to implement fractal relationship between an image and subparts of that same image. Starting with noisy stimulation, this method is thought to constitute a mean of uncovering pareidolic structures hidden inside (overcomplex) information. My observations have led me to think that this process follow life emergence mechanisms. The first step of upsampling generates an image that resemble organic cells, which are at the very basis of life. When repeating the method, more complex structures emerge, which lead to animal-like and plant-like pareidolic perception. This second step tends to show that scaling variation method enables to pass from simple organic pareidolia to more complex organism pareidolia. When repeating the method again, landscapes and autonomous worlds seems to emerge, going in the same evolving complexity described before. It is thought that this process could be implemented by algorithms, which would need to be able to choose adequately subparts of the signal that show aesthetic potential. This method, conjugated with online signal processing techniques, could constitute a basis for a project of pareidolic-generative landscapes, based on biotic signals, and controlled by the user in real-time to navigate into objective representations of his own mental world.
Presentation at Festival de la Imagen 2019, Manizales, Colombia