Fig. 1.

Sample frames and analysis from Night at the Museum. (A) Eye tracking from all participants (adults and children) shown as colored dots. (B–E) Outputs of faces-based model (B), graph-based visual saliency (GBVS) model (C), GBVS model with motion and flicker components (GBVS+F+M) (D), and Itti, Koch, & Niebur model (E). The bottom panel plots normalized salience for the adult group as a function of time. Note that in some cases (e.g., Frame f₂), although a face is present, most eye movements followed the hand motion.