Eye tracking and computational approaches to understand the roles of maturation and experience in infant looking

Project Summary/Abstract Infants use their eyes to gather information from their environment and learn about the unique scenes and objects that surround them. As infants accumulate these experiences, their cortex is concurrently maturing to support more controlled attention and visual processes. However, it is unclear how much of the development of gaze control reflects cortical maturation versus individual experiences. Therefore, the proposed project seeks to evaluate three competing hypotheses regarding respective roles of maturation and experience. We will use both traditional eye tracking analyses as well as advanced computational modeling. Specifically, in Aim 1, we will record infant looking behaviors while viewing scenes and objects that range in familiarity and assess how stimulus properties (e.g., salience, meaningfulness) relate to looking patterns for familiar and unfamiliar images across the first year of life. In Aim 2, we will relate these looking behaviors to a convolutional neural network inspired by the mature brain’s visual system. This innovative approach will evaluate the relationship between infants’ looking behaviors and use of higher-level visual processing across contexts that vary in familiarity. Ultimately, results from this proposal will inform developmental theories of visual attention by characterizing the contributions of maturational versus experiential factors. The information gained from the results of this project may facilitate future assessments and interventions of clinical populations with atypical visual scanning and attention control. To successfully accomplish these aims, it is necessary that the applicant gains theoretical and methodological expertise in both infant and adult attention research. This integrative perspective will allow her to characterize infant attention and visual behaviors using sophisticated methodological and analytical approaches, which will ultimately generate a wide range of subsequent research questions using similar techniques. These aims will be most successfully completed under the supervision and mentorship of Drs. Lisa Oakes and Steven Luck at the University of California, Davis, who are experts on attention in infant and adult populations, respectively. The proposed training goals focus on expanding content knowledge of attention and gaining methodological expertise in convolutional neural networks and representational similarity analysis. Accomplishing these training goals will better prepare the applicant for a successful independent research career that pushes the field of attention development forward.