Midterm 2 Flashcards
1
Q
name of the two visual streams
A
ventral - temporal
dorsal - parietal
for vision and interaction in the environment
2
Q
how many ganglion cell types are there and what are their names
A
M ganglion
P ganglion
3
Q
M ganglion cells
A
Magnocellular layers in LGN (layers 1 and 2 - inside)
fucntion movement and low light vision
high conversion of rods goes through M ganglion cells
4
Q
P ganglion cells
A
Parvocellular layers in LGN (3, 4, 5 & 6 - outside) function colour, texture and depth
5
Q
classic visual pathway in the brain
A
M ganglion - magno LGN - V1 –dorsal— paritel lobe
P ganglion - parvo LGN - V1 —-ventral— temporal lobe
6
Q
how lesion / ablation studies work
A
animal trained to indicate perceptual ability
specific part of brain is ablated or removed
animal is retrained to determine which perceptual abilities remain
results reveal which portions of the brain are responsible for specific behaviours
7
Q
classic ablation study for what and where pathways
-set up
A
animals trained on object discrimination task
-monkey shown an object
-then present with two choice task
-reward given for detecting target object
also trained on spaticial landmark discrimination problem
-monkey is trained to pick the food well next to a cylinder - so there is a spatital relationship and no object discrimination
then temporal and parietal lesions
8
Q
classic ablation study
results
A
temporal lobe removed = problems in object discrimination task so what pathway
parietal lobe removed = problems in landmark discrimination task so where pathway
9
Q
behaviour of patient DF and conclusions
A
damage to ventral pathway die to carbon monoxide posioning
could not tell orientation (perceptual orientation matching) of the slot but could actively post a letter into it (visuomotor posting)
ventral stream = what
dorsal stream = how
really prfound deficit - brain knows what it is at some level but cnanot do it
10
Q
how did what where evolve
A
what and how
11
Q
what is the inverse projection problem
A
an image on the retina can be caused by an infinite number of objects
fundamentally ambigious at the level of the retinal image - need to impose additional constraints
so no specific right way object recognition must work
12
Q
what else makes object recognition tricky
A
huge variation - eg loads of pictures of dolphins, lots of inconsistencies but our brain still knows is a dolphin
13
Q
what does gestalt mean
A
german word
configuration or pattern
14
Q
according to gestalt perception…
A
is not built up from sensations but is a result of your brain imposing perceptual organization on incoming stimuli
15
Q
gestalt principles are known as..
A
heuristics = best guess rules
16
Q
6 gestalt organizing principles
A
good continuation proximity / similarity common fate common region uniform connetedness meaning
17
Q
good continuation
A
continuous shapes viewed as single segmented obejcts
helps us perceive a pile of rope as one continous object and not all broken up
18
Q
proximity / similarity
A
things that are bear to each other are grouped together
19
Q
common fate
A
things moving in the same direction are grouped together
20
Q
common region
A
elements in the same region tend to be grouped closer together
21
Q
uniform connectedness
A
connected a region of visual properties are peceived as a single unit
22
Q
meaning
A
interpret images in line with top-down knowledge
eg we see faces everywhere and in everything. us imposing top down knowledge on inanimate stuff
23
Q
explain gestalt laws competring
A
leads to an ambigous percept
shows there isn’t a correct way to process images
brain just tries to apply the heuristics to disambiguate the info
24
Q
charlie chaplin mask illusion
A
we have a lifetime experience seeing faces
never see them as concave - always projecting out
so we impose structure on the world
and so see both sides of the mask as pointing out
cannot see otherway round
top down imposition of knowledge by the brain on the bottom down world
25
Biederman recognition by components theory
simple computational model of object recognition
36 geons
each geon is uniquely identifiable from most viewpoints
objects can be identified if geons can be identified
26
evidence from Biederman
objects harder to recognise if geons are obscured, easy to recognise if can see geons
27
strengths of biederman
viewpoint invariance
| represents 3D structure
28
weaknesses of biederman
complexity of representation
doesn't easily represent subtle metric differences (ie distance between the eyes)
recognition is at the level of categories (chair vs table) rather than individuals (my chair vs office chair), also trump vs someone dressed up as trump
29
modern day version of recognition by componenets
deep neural networks
modern machine vision
architecture for performing object recognition
30
how deep neural networks work
series on interconnected layers of modelled neurons
node = neuron / pop of neuron
input
basic processing at ealry layers eg 100000000 instead of just geons, look a lot like primary visual cortex
as progress through the layers = more and more complex representations
even modern machine learning use same general basic approach
building blocks put together to descirbe objects in detail
note this is how AI works
31
meaningfulness or familiarity
things are more likely to form groups if the groups appear meaningful or similar
32
meaningfulness or familiarity experiment
how may toothbrushes in the image
if you show the image to a neural netowrk straight up 2, if shown to a person = 1 as second huge one only seen on inspection as our brain uses prior knowledge to constrain inputs coming in from the retina
33
neurons response in area IT
their responses are similar within each object accross viewpoints
so can recognise all dolphins
so respond to object irrespective of where it is in the visual field
population code and not a unit code
perceptual experience due to entrire distribution of activity is across the cortex and not one very loud neuron
converging inputs into IT
IT - inputs from whole bunch of areas, why neurons can be so selective in IT
34
IT an new objects
we have to be able to rapidly learn new infor but can't have brain reconfiguring everytime it sees something new. so need stability and flexibility
flexible microwire - implanted in brain and moves with the brain - will follow the same neuron in the brain for months and months at a time
stability - brain processes monkey face the same
despite stability responses are flexible during learning = plasticity
series of novel objects presented to mokey = cell goes all over the place. as animal learns this activity becomes sterotyped and constant
period of plasticity before stable response turns up
35
what do we mean by modules in the ventral pathway
specialised chunks of cortex perform somewhat different perceptual functions
based off single cell recording in monkies = specific cells for faces
36
Nancy Kanswisher study
fMRI
| face areas on left and right hemispheres and 2 for outdoor scenes
37
what brain area is for
| faces
fusiform face area
38
what brain area is for
| scenes/ landmarks
parahippocampal place area
39
what brain area is for
| intact over scrambeled objects
lateral occipital complex
40
are the FFA/PPA/LOC innate?
hard to answer but certainly develop early in life
41
Tsao's research on face patches
macaques in fMRI
localized face selective regions
lots of areas
then single cell recording in these areas
used scrambeled patterns as a base for 0
but only high decoding activity specifically for human faces over everything else in one area
pattern of activity is what matters
lots of areas process faces but only some are face specific but others firs for lots of objects, faces included
42
what happens when motion perception fails
akinesthesia
43
hubel and weisel discovery for motion perception -
direction selective cells in the cat V1
| almost all neurons in area MT are also direction selective
44
psychophysical evidence for motion detectors and why
motion after effect
provides evidence for direction and rotatioanlly selective neurons
exposure to moving stimuli causes the most responsive neurson to stop responding after a while
even when a stationary stimulu is presented these neurons aren't as active and neurons that prefer the opposite direction are more active and they bias perception
45
motion detectors
not simply a static visual field but temporal dimension
spatio-temporal receptive field
responds to a particular direction of motion (preferred direction) and not as much to other (null/anti-preffered) directions
only when stimulus MOVES in the receptive field
46
explain the appeture problem
is the consequence of a limited receptive field in V1
| so square moving diagonally right with a small appeture can look like vertical upwards movement
47
how is the appeture problem solved
cells later in the visual system integrate information from the appeture of multiple complex cells
many downstream neurons in MT can do this
so the receptive field of MT neurons will encompass the movement of the whole square
48
what experiments proposed the link between motion perception and MT
causation study
Newsome
correlated dot motion - degree of coherance manipulated
part 1 - determined a single neuron in MT can discriminate motion as well as the monkey. used left right choice and recorded activity of neurons that responded maximally to one direction, if neuron responded alot then experimenter guessed the motion was left, therwise right. accuracy of neuron computed by histogram
part 2 - ablated MT in monkeys - severe deficits following selective lesion of MT
part 3 - microstimulation
-mokey performed left vs right discrimination - pass trials of stimulating pulses into collumn of neurons that prefer left motoin = biased the monkeys to chose left even if 90% of dots moving right
49
explain the monkey training stage of using correlated dot patterns
trained monkeys to do as well as humans
graph x axis = degree of correltaion in movement of dots
y axis = behavioural performance
wiggly curve up (slow, steep, slow)
50
organization of MT
like ice cube tray model
51
what is biological motion
another example of kinetic depth effect
light point walker
shows male vs female, body size, affect etc
52
fMRI of bio motion
processed in the superior temporal sulcus STS (next to MT)
thought the STS bridges MT and IT - the two streams dorsal and ventral
this region is thought to be crticial for interpreting actions of others in social situations
53
what is visual neglect
disorder in which individuals are unaware of events in the space opposite their lesion
-can just be vision but tends to include other sensory or mototr deficits
importantly not dues to a sensory system problem eg abnormal vision
neglect is generally regarded as a selective lack of attention
54
two examples of visual neglect seen in class
house on fire = listen to go over lecture example
old lady shown bathroom picture only named items from one half of the image, did name object in the middle - they don't always do this
55
impact of visual neglect on a patient
live in a different world - won't eat food from one side of their plate, ignore people etc
may neglect parts of their own body - makeup and not accept ownership of left limbs
may even physically abuse their left side
56
visual neglect and the brain
occurs in 11-37% patients following brain damage
right inferior parietal lobe / superioral temporal gyrus = common areas
may includ other lesion sites
-left TPJ
-supramarginal or angular gyrus
-inferior and middle frontla gyri
also subcortical lesions to thalamus, basal ganglia and midbrain
57
can you recover from neglect
yes, often after an acute injury
| then left with extinction usually
58
what is extinction
failure to detec a stimulus contralateral to a lesion during simultaneous bilateral stimulation (so stimuli presented on both sides at the same time, would get right if just presented one at a time)
= competition
59
explain the mental imagery study
done in Milan with two locals
subject asked to imagine themselves in a popular plaza
recall the shops and buildings from the cathedral or the opposite side of the street
both patients only named shops on the right side
when asked to do it from the reverse perspective again only named right shops (just from this new perspective)
not sue to difficulty in recall obviously
but info cannot be accessed even when relying on mental imagery so its more than just don't act on info, they don't have the info in the first place
60
what is simultagnosia / balints syndrome
complete inability to attend to more than one feature regardless of the side of presentation - even in the fovea
when presented with two spatially conjoint objects (eg pen crossing a pencil)
-will report only one
-may report the other in other trials but only one per trial
61
brain area damage associated with simultagnosia
also known as balints
| bilateral damage to the parietal lobes
62
three types of cues for depth perception
oculomotor
monocular
binocular
63
define oculomotor cues
cues based on sensing the position of the eye and luscle tension
64
define monocular cues
cues that rely on only one eye
65
define binocular cues
cues that rely on both eyes
66
explain types of oculomotor cues
convergence - knowing the inward movement of the eyes when we focus on nearby objects
accomodation - feedback from changing the focus of the lens
67
explain the 2 broad types of monocular cues
pictorial cues - sources of depth information that come from 2D images liek pictures
movement produced cues
68
name 8 types of pictorial cues and explain when not obvious what they mean
oclusion - one object covers another
relative height - higher in visual field = more distant
relative size = if same size objects, closer one will take up more sapce on retina
familiar size - distance info based on knowledge of object size
perspective convergence - parallel line come together in the distance
atmospheric perspective - far away = blue tint
texture gradient = equal spcing appears more packed if further away
shadows
69
explain the two types of motion-produced cues
motion parallax - close objects in direction of movment glide rapidly past but objects in the distance move slowly (truck on long road example)
deletion and accretion (occlusion in motion) - objects are covered or uncovered as we move relative to them
70
what are vergence movements and explain the two tyes
eyes moving to keep the object of fixation on the fovea of both eyes
convergence = when an object gets closer
divergence = object gets farther
71
what is binocular disparity
your eyes get two different images of the world
72
what is a horopter
an imaginary circle that passes through the point of focus
objects on the horopter will have equal disparity in the two eyes
it is an imaginary line of zero disparity where point of light from a given object will fall in corresponding locations
73
define disparity
the difference in where light is on the retina of the focus point and the image
74
the closer the object to the eye relative to the point of fixation the ...... (disparity)
bigger distance the closer to the retina
75
what can binocular disparity tell us about
cue to depth
76
crossed disparity is when
th object is infront of the fixation point
77
uncrossed disparity is when
the object infront of you is behing the fixation point
78
when inside the horoptor (but not infornt of you), objects in the near eye have.... than the far eye
less disparity
79
when outside the horoptor (but not infornt of you), objects in the near eye have..... than the far eye
more disparity
80
what is stereopsis
depth information provided by bioncular disparity cues
- steroscope uses two pictures from slightly different vewpoints to give the impressionof depth and solidarity
- 3D movies use this principle
81
explain the physiology of depth perception
neurons have been found that respond best to depth perception
=bioncular depth cells or disparity selective cells
= response best to a specific degree of disparity between the images on the left and right retinas
82
monkey experiment showing the physiology of depth perception
monkeys viewed textured gradients that were 2D picutres and 3D sterograms
recordings from a neuron in the parietal lobe showed
cells responded to pictorial cues
cell also responded to binocular disparity
83
cat experiment on binocular disparity and depth perception
cats reared by alternating vision between two eyes (otherone covered by an eyepatch)
results showed they had
-few binocular neurons
-were unable to use binocular disparity to perceive depth
this happens in around 10% of adults = stereoblind. but it is ok as stereopsis isn't helpful for distances beyond about 6ft anyway
84
what is the visual angle
proportion of the visual field occupied by an object
think of eye christmas tree drawing with angle labelled
see slide40 lecture 9 if cannot remember
85
how can we infer the size of an object
| so if written in an equation...
relationship between the visual angles and distance
perceived size= retinal image size x perceived distance
P=RD
86
what is emmert's law
objects that generate retinal images of the same size will look different in physical size if they appear to be located at different distances
87
what is an optic array
structure created by the surfaces, textures and contours in the environment
88
what happens to the appearances of objects as the observer moves past them
optic flow
| self-prodiced information
89
what is the gradient of flow in optic flow
difference in flow as a function of distance from the observer
90
what is focus of expansion in optice flow
th point in the distance where there is no flow
the focus of expansion is always centered wherever you are heading and provides invariant information that reain constant while the observer is moving
91
circle explaining optic flow
movement creates flow which provides inromation for more movement = back to movement at the top of the circle again
92
swinging room experiment
walls of a room moving = child will fall over
93
name the three brain areas involved in movement, action and perception
premotor (mirror area(
posterior parietal
medial superior temporal area MST
94
the physiology of navigation
| studies to back up
optic flow neurons - neuron in the MST of mokeys respond to optic flow patterns
maybe go over end of lecture 9 for clarification of end exmaple
95
what does affordances mean
made up of information of what an object is udes for
| they indicate potential for action as part of our perception
96
evidence for affordance of objects in humans
people with certain types of brain damage show that even though they may not be able to name objects, they can still describe how they are used ot can pick them up and use them
97
evidence for affordance of objects in monkeys
PRR - parietal reach area
neurons in the posterior parietal cortex respond before monkeys grasp an object thus signal the intention to grasp
neurons from this region send signlas directly to the premotor area that are immediately anterior
