When you think about the mechanism of human vision, i.e., "how we perceive objects and the environment," you might think that the information received by the eyes is processed sequentially through various stages. However, our vision is not accomplished by such a sequential flow alone. Various pieces of knowledge and assumptions about objects in the external environment and the environment itself are stored in the higher parts of the brain that reside beyond the primary visual cortex, which is responsible for early visual information processing. When an object is seen, the bottom-up information coming up from the eyes to the brain, and the top-down information that comes down from the database reside in the higher levels of the brain, meet, compete, and compromise with each other in one way or another. Only after such a compromise, an object can be perceived and understood.

For example, when you look at the inside of a mask with two eyes, it looks concave. However, when you look at it with only one eye, you see it concave at first, but if you keep looking, it eventually starts to appear convex. The depth information (binocular stereopsis information), which functions when looking with two eyes, does not function with only one eye and the depth information becomes ambiguous. In such occasions, the top-down information, that "face are convex" becomes predominant, and you perceive the inside of the mask convex although it is actually concave.

I am trying to understand the mechanism of the visual perception through psychological experiments. However, to explore the nature of such a complex mechanism, it is important to interact with researchers from various research fields such as neuro-physiologists who study brain mechanism, researchers in brain imaging, modeling people who study theory, and experts in computer vision who attempt to build artificial vision systems.