Week 4 pt.2 : Object perception

Question 1

An object is more than the sum of its parts…

Answer 1

• interpreting objects requires bottom-up & top-down cues
• example Oliver sacks book ‘the man who mistake his wife for a hat’
• patient w/ object agnosia
• thought leather glove was a change purse
• this patient had damage to inferotemporal lobe + it impaired his ability to bring together the bottom-up & top-down cues necessary to perform everyday tasks

Question 2

Ability to detect objects must overcome 3 aspects…

Answer 2

1. Image clutter = to overcome, we must discern the object despite the overlapping presence of nearby objects
2. object variety = we must recognize a specific object as part of a certain class/category despite small/large differences
3. variable views = we must recognize an object despite is being in different orientations relative to our retinae (have to identify same object from different vantage points)

Question 3

Shape constancy…

Answer 3

• Object perception provides a great deal of flexibility & gives rise to shape constancy
• shape constancy is the idea that an object will be perceives to have the same shape, regardless of how its orientation changes on our retina
• it’s taken the world’s best software engineers decades to produce computer vision algorithms capable of distinguishing objects

Question 4

representation versus recognition

Answer 4

• representation = seeing someone as human
• recognition = their identity

Question 5

object recognition

Answer 5

• how an object and its parts are encoded by different areas of the brain
• the neural code for what the object acc is
• may include representations of object features (certain neurons) and object representations (conjunction of features)

Question 6

Object recognition

Answer 6

• association of the representation w/ those previously stored in memory to asses the identity, use/meaning of an object
• must recognize… specific objects objects as members of larger classes and objects as specific instances of that category

Question 7

Perceptual organization

Answer 7

• the process by which multiple objects in an environment are grouped, allowing us to identify objects in complex scenes
• 2 important processes… grouping & segregation

Question 8

grouping & segregation

Answer 8

• grouping = elements in a figure grouped together into a common unit/object
• segregation = distinguishing 2 objects as being distinct/discrete

Question 9

figure-ground organization

Answer 9

• we separate the figure from the ground
• we divide the world into 2 elements: figure/object of regard and ground/background
• example is the face-vase illusion
• from a bottom-up perspective, the physical features of an image can have an effect on how the figure-ground distinction is made (e.g. images near the bottom of a landscape are closer to us)
• this makes us decide objects in lower field = figure against the higher background

Question 10

stimulus component role in figure-ground relations

Answer 10

• stimulus components play a role too…
• we show a strong bias toward perceiving symmetrical objects as being figures interest
• also show bias toward objects with convex borders as being figures as opposed to concave borders

Question 11

Gestalt psychology

Answer 11

• Gestalt psychologists interested in understanding the rules by which perception picked out the whole from its parts
• they created the gestalt laws of perceptual grouping
• they’re a group of physical parameters present in stimuli that influence the likelihood/pattern by which they will be grouped into coherent visual objects

Question 12

Law of continuation

Answer 12

figures that have smooth edges are more likely to be seen as continuous than figures with edges with abrupt/sharp angles

Question 13

Law of proximity

Answer 13

elements close together usually perceived as a unified group

Question 14

Law of similarity

Answer 14

elements similar to one another perceives as a unified group

Question 15

law of symmetry

Answer 15

elements symmetrical to one another perceives as a unified group

Question 16

low of common fate

Answer 16

elements moving together perceives as unified group

Question 17

Perceptual interpolation

Answer 17

• visual cortex can interpolate (fill in) missing info based on physical features of the surrounding area & experience with the properties of physical objects
• visual experience allows the visual cortices to fill in the information missing from the representation produced by the retina & provide a complete perception of objects

Question 18

Edge completion & illusory contours

Answer 18

• edge completion = perception of a physically absent but inferred edge, lets us fully perceive a partially hidden object
• illusory contours = perceptual edges that exist cuz of edge completion but are not physically present (necker cube), processed at an early level in vision (V2)
• action potentials equal to actual edges

Question 19

Recognition by components theory

Answer 19

• previous things talked about are all top-down
• opposing theory that suggests bottom-up and that objects in our world may be regained as the conjunction of a small set of simple object (geons)
• 40 independent geons
• not rly liked cuz it fails to account for many phenomena in the real world

Question 20

Pathway to object recognition

Answer 20

• object processing = ventral pathway
• (1) V1 = orientation sensitivity, unilateral motion
• (2) V4 = contour sensitivity, colour processing
• (3) inferotemporal cortex = object sensitivity, spatially indiscriminate, has different areas specialized for the processing of objects of different forms

Question 21

Pathway to face recognition

Answer 21

• contour and colour information from V4 passes thru the occipital face area located at the border of occipital & temporal lobes
• it serves a gatekeeper function
• determines whether object in question is a far or not (if it is it goes to FFA)

Question 22

FFA

Answer 22

• where higher level processing related to identification, perception of emotion etc. occurs
• faces elicit more activity than non-face stimuli
• high levels of activity in this region are associated with correctly identifying faces

Question 23

prosopagnosia

Answer 23

damage to the FFA = significant impairment to able to identify even very familiar faces

Question 24

Kobe Cells

Answer 24

• FFA likely encodes identify using a distributed model in which the combined output across multiple neurons allows this area to encode the 100s/1000s of faces we are familiar with
• Some research which suggests there may be a region of the brain located medial to the FFA in which single neurons actually encode highly salient individual faces

Question 25

parahippocampal place area

Answer 25

responsive to locations of interest (landscapes & outdoor)

Question 26

extrastriate area

Answer 26

- neurons encode non-face parts of the body