The article by D. Marr and T. Poggio explores the cooperative computation of stereo disparity, a process essential for depth perception in the visual system. They analyze the computational structure of stereo-disparity extraction, which involves selecting a particular location in one image and identifying it in the other, followed by measuring the disparity between corresponding points. The authors derive a cooperative algorithm that implements this computation, demonstrating its effectiveness on random-dot stereograms. The algorithm is based on two key constraints: uniqueness (each item from each image can be assigned at most one disparity value) and continuity (disparity varies smoothly almost everywhere). The cooperative nature of the algorithm allows for the implicit representation of global order, which is crucial for solving the correspondence problem. The authors also discuss the implications of their findings for psychophysics and neurophysiology, particularly regarding the biological mechanisms underlying stereopsis. They suggest that the algorithm's success in solving stereo-disparity problems may provide insights into how the brain processes visual information.The article by D. Marr and T. Poggio explores the cooperative computation of stereo disparity, a process essential for depth perception in the visual system. They analyze the computational structure of stereo-disparity extraction, which involves selecting a particular location in one image and identifying it in the other, followed by measuring the disparity between corresponding points. The authors derive a cooperative algorithm that implements this computation, demonstrating its effectiveness on random-dot stereograms. The algorithm is based on two key constraints: uniqueness (each item from each image can be assigned at most one disparity value) and continuity (disparity varies smoothly almost everywhere). The cooperative nature of the algorithm allows for the implicit representation of global order, which is crucial for solving the correspondence problem. The authors also discuss the implications of their findings for psychophysics and neurophysiology, particularly regarding the biological mechanisms underlying stereopsis. They suggest that the algorithm's success in solving stereo-disparity problems may provide insights into how the brain processes visual information.