PSY270H5 Lecture Notes - Lecture 3: Retina, Translational Symmetry, Reinforcement
Document Summary
From the previous lecture: human"s superiority in object recognition. Objects may appear anywhere on retina in any size or orientation. Objects may be occluded by other objects. Perceptual consistency: template theory vs feature analysis theory. Object is recognized when perceived info matches w/ a memorized templates of objects. How many templates do we need to have to. Doesn"t jive well - bc would need diff template for each orientation of the world. Fixed set of elementary properties analyzed independently in parallel across visual field. Does component analysis --> doesn"t require the perfect match w/ the template: feature vs conjunction search. Target is defined by a single elementary feature. Very efficient due to parallel processing of elementary features (c. f. , feature analysis theory) Target is defined by a conjunction of elementary features (e. g. color and orientation, etc) Not as efficient as feature search bc of the necessity to bind the elementary features to define the target.