The work described in this proposal will be a step towards the understanding of the development of visual form perception. The first step in understanding this process is to distinguish the necessary and sufficient stimuli (or the type of information) needed for this task in early infancy. Once the type of information used by the brain is determined, then the potential computational mechanisms allowed by the stimuli can be explored, and the developmental origins of the computational mechanisms can be analyzed and investigated. We propose a systematic method of attacking these issues using an integrated psychophysical infant experiments and computational modeling approach. In exploring each issue, we use the result of experiments in order to guide our analysis through modeling and simulation, and use the result of the simulations in order to guide the experiments. Such interplay between the two complementary approaches, we believe, is crucial for a methodological and rigorous study of a highly complex system such as the brain--in particular the developing brain. The knowledge of the role of various cues exploited by the visual system at different stages of development, and the type and process of learned functionalities will contribute to the medical applications aimed at facilitation and improvement of learning for visually-impaired infants, as well as educational applications enhancing the development of normal infants. While this research builds on previous work on the topic, it is the first rigorous investigation of issues such as the level of object form representation in infancy, the computational mechanisms and the developmental origins.
