Understanding human vision processing has been the ultimate goal for many vision researchers, including machine vision developers. However, successes in specific problem domains such as face detection and gait recognition have driven the machine vision community to invest more effort on a so-called flat architecture than on a human-like deep architecture. The former employs a number of simple feature detectors compensated by a large amount of training data. The latter employs a hierarchy of specialized feature detectors to extract essential information from a smaller amount of training data. Through an extensive review of primate vision systems, the authors try to shift the trend toward the deep architecture.

Compared to other surveys on primate vision systems, this work clearly and consistently delineates both structural and functional perspectives of each component in the primary visual pathway. The structural view identifies potential hardware and the functional view identifies potential software implementations in machine vision systems. These views also paint clear pictures of the current knowledge of primate vision processing without going deeply into biology and physiology. Thus, the paper is an excellent introduction to primate vision for machine vision researchers.

It should be noted that the information provided in the paper often oversimplifies vastly complex structures. The brain is holistic and plastic. Thus, we cannot model an entire primate vision system as a collection of independent and dedicated components. Treating a group of neurons as static operators ignores the highly adaptive nature of the biological system.