You are viewing the SNL 2017 Archive Website. For the latest information, see the Current Website.

Poster A48, Wednesday, November 8, 10:30 – 11:45 am, Harborview and Loch Raven Ballrooms

Using Kinect technology to assess word learning

Andrés Méndez1, Rossana Guerra1, Leonel Gómez1;1Universidad de la República

Behavior has been often underestimated when it comes to understand complex cognitive functions, such as language acquisition. From an embodied and dynamic perspective, however, the structure of behavior and social interactions in space and time are key to understand how words become relevant to children. This does not mean that complex linguistic structures cannot emerge independent from experience, as classical cognitive theories describe. Recent evidence from embodied perspectives show the need to study how much variance and information there is in the interaction of infants with their environment, and therefore, how far we can go without the need to describe language as a symbolic-like process that happens all inside the child’s head. This proposal demands the development of studies that analyzes behavior as a function of interaction and time, which usually result in large amount of data and time-consuming coding procedures. Kinect technologies are well suited to solve the demanding challenge of quantifying complex behavior with high temporal resolution. Our study assesses the power of the most recent Kinect to provide reliable data on infant´s behavior. This was done by comparing data from the Kinect to data obtained through manual coding. Of specific interest was the detection of hand and head movements while both parent and child interact freely in object-naming conditions. The results show that optimizing this technology is a potential road to the design of automatic, remote, non-invasive and reliable technology for the study of word learning - and cognitive development in general - from an embodied perspective.

Topic Area: Methods

Back to Poster Schedule