Sounds like maybe this could be used for deep-learning networks to learn what to pay attention to. The memorable regions should be the salient or important regi

Sounds like maybe this could be used for deep-learning networks to learn what to pay attention to. The memorable regions should be the salient or important regions in the image. Basically the more you are paying attention, the more memorable something is.