ABSTRACT
Humans are exceptionally proficient at extracting and using statistics embedded in the visual world. In a single exposure, we can immediately perceive a summary of the general properties in a group of objects. Across multiple exposures, we can learn hidden regularities in how objects co-occur. This chapter reviews and synthesizes different forms of such statistical perception and statistical learning, respectively, as they operate over spatial information. We then explore how these foundational processes support more complex spatial behaviors, namely our ability to navigate through the world. Current research is converging on the realization that statistical learning and spatial navigation share cognitive and neural mechanisms. By integrating experiences and enabling prediction, this interdependence may make real-world behavior more efficient and rewarding.
