The Hierarchical Evolution in Human Vision Modeling

Abstract

Computational models of primate vision took a significant advance with David Marr's tripartite separation of the vision enterprise into the problem formulation, algorithm, and neural implementation; however, many subsequent parallel developments in robotics and modeling greatly refined the algorithm descriptions into very distinct levels that complement each other. This review traces the time course of these developments and shows how the current perspective evolved to have its alternative internal hierarchical organization.

Keywords: Active vision; Bayes hierarchies; Human gaze behaviors; Marr's paradigm; Reinforcement learning; Visual routines.