COGS 2160 Lecture Notes - Lecture 9: Retina

• School: York University
• Department: Cognitive Science
• Course: COGS 2160
• Professor: Muhammad Ali Khalidi

Document Summary

Computational level: specify the information-processing problem: in terms of vision - input: images on retina are double, inverted, two dimensional output: single, non-inverted, three-dimensional. Hardware implementation level: identifies a physical realization of the algorithm. We can study algorithmic level w/o implementational: how does the device do this? i. e. according to what rules. Studies on patients /w brain lesions reveal double dissociations. Damage to region r1 disrupts function x while leaving function y intact; damage to region r2 disrupts y while leaving x intact. This tells us which functions work relatively independent. Right parietal lesions: patients recognize object in conventional view: they do not recognize pail from above, deny it could be a pail. Left parietal lesions: patients unable to name object or state its purpose: but they correctly perceive its geometry (shape) in unconventional view, by, matching two views. Representation of shape of object is stored in a different place from representation of use and purpose.