Fig. 2From: Predicting students’ attention in the classroom from Kinect facial and body featuresHuman estimation and annotation of person’s attention level. By observing video footage (a), five human coders estimated attention level of each of the test persons. On graph (b), mean attention score \(A^{m}_{u}(t)\) is shown as a blue line and final three-level reference attention \(A^{r}_{u}(t)\) as a red lineBack to article page