ON COMPUTATIONAL TECHNIQUES FOR EXPLORING PARENT-INFANT DYNAMICS DURING SOCIAL INTERACTION
Burling, Joseph Michael
MetadataShow full item record
With the recent trend in developmental science toward new techniques and methods that involve sophisticated hardware and data processing, it is becoming more and more important to merge the many different subfields within psychology, computer science, and engineering. This current paper provides a new perspective on how the two fields of developmental science and computer vision may benefit from one another through a mutual interest in resolving the complex questions regarding human learning. Several well established techniques in cognitive development and image processing are outlined in terms of their importance in bridging the gap between the two research areas. The combined efforts help to formulate new sets of questions crucial to understanding human cognition, particularly in regards to the development of infant visual attention. The methods and techniques outlined in this paper only scratch the surface of what is possible when combining knowledge across multiple disciplines. Specifically, processing images extracted from third-person and first- person-view cameras and head-mounted eye trackers can aid in understanding the complex series of factors involved in parent-child social interactions. Advantages in being able to collective extensive amounts of data from these techniques as well as their potential limitations will be discussed.