Principles of Cognitive Vision reframes vision as an active, predictive dialogue between seeing and thinking. Instead of a black box that turns pixels into labels for later reasoning, cognitive vision couples perception with goals, memory, and action. The result is a system that anticipates what should be seen next, guides attention and movement, and learns by testing its own expectations in the world.
Drawing on psychology, neuroscience, and robotics, the book explains how cognitive maps, predictive coding, and active inference connect low level signals to high level understanding. It shows why embodied agents must fuse bottom up evidence with top down knowledge at a middle layer where perception and reasoning meet. Clear and rigorous, Principles of Cognitive Vision offers a blueprint for visual systems that do not just recognize scenes, they plan, predict, and purposefully engage with them.