In this work we describe a general probabilistic framework that accounts for deployment of attention in complex natural images, scene recognition and object detection. The framework integrates into a unique model three factors for attention guidance: bottom-up saliency (based on low-level image properties), target driven search (model of the appearance of the target object) and global scene priors (top-down priors provided by the gist of the scene). Using a Bayesian framework, we show how to determine (1) which are the locations and scales in the image that are best candidates to contain objects of interest, (2) how the expected appearance of the target modulates the saliency of local image regions and, (3) scene priors: used to modulate the saliency of image regions early during the visual search task.
Most models of attention and object recognition rely on the definition of sets of local features. In the local pathway, each location is represented by a vector of features that describe local image properties. It could be a collection of templates (e.g., object detection) or a vector composed by the output of wavelets at different orientations and scales (e.g., saliency models of attention). In the global pathway, the entire image is represented by a unique set of features that summarizes the appearance of the scene without encoding specific objects or regions.
