PSY280H1 Lecture Notes - Lecture 3: Rhinitis, Outline Of Object Recognition, Machine Vision
13
PSY280H1 Full Course Notes
Verified Note
13 documents
Document Summary
This is where perceived representations are matched to encoded representations. This is oliver selfridge"s computational model of letter recognition, pandemonium model: perception by committee, https://oup-arc. com/access/content/sensation-and-perception-5e-student- Demons loosely represent neurons, each level represents a different brain area resources/sensation-and-perception-5e-activity-4-5: but for this to work, we have to have an encoded representation of what letter b looks like in order to match it. They are recognized by non accidental properties, so they can be identified from any angle/viewpoint. Viewpoint invariance theories propose that our representation doesn"t depend on our viewpoint. Our object recognition system is best at categorizing objects in different ways, being able to do this allows us to deploy diff strategies that are consistent w our conceptual understanding of the world. Our ability to carve nature at its joins = our ability to categorize things that we see. We can do this better or worse depending on what type/level of object we"re looking at.
