Within the Wildflower schools, video cameras and Kinect sensors are set up in the space in order to track individual and group behavior over days and months. Using computer vision and machine learning algorithms, our video analysis software tracks the children’s motion, position, and gestures during the class period. It provides valuable statistics that help the Montessorians understand the children’s learning patterns.
For example, activity graphs show the activity at designated places in the school space. This is calculated by tracking limb movements within a defined region that is specified by the user in our software. The amount of movement is depicted by a single number, calculated by summing and normalizing the change in contour shapes that correspond to hands, feet, heads, and lower body parts.