The topological perception theory proposed a precise description for how one organizes the visual world that eliminates the vagueness of subjective phenomenology. In addition, the topological theory challenges the dominant computational view on the part-to-whole hierarchy of visual information processing. Lines of evidence from adult psychophysics, brain imaging data, and even honeybee's behavior have supported the notion that the global topological properties are the very primitives of visual representation. However, the question of how the sensitivity to topological properties originates during development has not been explored much. In a previous study, the investigators found that 2- to 6 month old infants could reliably discriminate stimuli based on topological differences, but failed to do so based on geometric differences. Using familiarization/novelty preference procedure, the present study intends to explore the visual sensitivity for topological properties in newborn infants. Experiment 1 focuses on whether neonates can discriminate a disk (no hole) and a ring (with a hole) that are topologically different, and/or a disk (no hole) and a triangle (no hole) that are geometrically different. Experiment 2 focuses on whether neonates can detect a change in the number of holes and/or the size of the hole. If newborn infants are only sensitive to topological properties and not to geometric properties, this will be a strong proof for the claim that topological property is the very "primitive" visual representation at empirical as well as theoretical level.