Combining Recurrence Analysis and Automatic Movement Extraction from Video Recordings to Study Behavioral Coupling in Face-to-Face Parent-Child Interactions.

ABSTRACT: The analysis of parent-child interactions is crucial for the understanding of early human development. Manual coding of interactions is a time-consuming task, which is a limitation in many projects. This becomes especially demanding if a frame-by-frame categorization of movement needs to be achieved. To overcome this, we present a computational approach for studying movement coupling in natural settings, which is a combination of a state-of-the-art automatic tracker, Tracking-Learning-Detection (TLD), and nonlinear time-series analysis, Cross-Recurrence Quantification Analysis (CRQA). We investigated the use of TLD to extract and automatically classify movement of each partner from 21 video recordings of interactions, where 5.5-month-old infants and mothers engaged in free play in laborato

SUBMITTER: Lopez Perez D 

PROVIDER: S-EPMC5742271 | biostudies-literature | 2017

REPOSITORIES: biostudies-literature