Visual saliency maps for studies of behavior of patients with neurodegenerative diseases: Observer's versus Actor's points of view

We are interested in finding the relation between the visual saliency maps of the viewer of visual content and the actors (person executing the actions) in the context of studies of neurodegenerative diseases such as Alzheimer's disease. From results of eye-trackers worn by the actors and used when recording observers, and on the basis of hand-eye interactions from motor control studies we established a time shift between actor's and viewer's saliency maps. This time shift corresponds to the latency of hand-eye interaction. The method is based on adequate normalization of saliency maps and computation of similarity metrics for pixel based saliency. This finding gives good perspectives for automatic prediction of a normal actor saliency map from observer saliency map.