Abstract
We observe that everyday images contain dozens of ob jects, and that humans, in describing these images, give different priority to these ob jects. We argue that a goal of visual recognition is, therefore, not only to detect and classify ob jects but also to associate with each a level of priority which we call ‘importance’. We propose a definition of importance and show how this may be estimated reliably from data har- vested from human observers. We conclude by showing that a first-order estimate of importance may be computed from a number of simple image region measurements and does not require access to image meaning.