Neuroarchitecture is the growing discipline that studies how manmade built environments shape our brain and our behavior. In recent years, researchers have started to uncover human preferences for a variety of architectural characteristics, such as curvature and openness, which are often associated with corresponding neural and physiological responses. In my talk, I will describe a series of studies investigating how people recognize visual scenes, focusing on how early the human brain extracts diagnostic perceptual information that can inform navigation and movement in the environment. These studies reveal that very early on, around 220ms after the onset of visual processing, the brain already processes key properties of the visual environment. These properties, such as naturalness, openness, and navigability are known as global scene properties, and can be used to capture the structure and function of real-world scenes. Potential implications and applications of these findings will be further discussed.