Contextual Priming for Object Detection

There is general consensus that context can be a rich source of information about an object's identity, location and scale. In fact, the structure of many real-world scenes is governed by strong congurational rules akin to those that apply to a single object. Here we introduce a simple probabilistic framework for modeling the relationship between context and object properties based on the correlation between the statistics of low-level features across the entire scene and the objects that it contains. The result-ing scheme serves as an eective procedure for object priming, context driven focus of attention and automatic scale-selection on real-world scenes. Keywords Context, image statistics, object recognition, focus of attention, automatic scale selection, object priming The authors would especially like to thank A. Oliva and W. Richards for fruitful discussions. The authors also want to thanks D. Kersten for comments about this work. This research was supported in part by funds from the Alfred P