Visual Perception
focusing on visual perception
same principles apply to all our senses
allow us to understand and act upon the world
3 important components
stimulus in the world, conversion, then our action and behaviour
allow us to get info about what's out there in the world into our minds
we don't have direct access to that info
there needs to be some kind of representation
physical thing isn't in mind, there is some representation of it
neural representation
get energy converted out there in the world to a neural signal that we can process
stimulus and our processing of it helps to guide behaviour
there is a lot of processing that happens with stimuli in the world and our
we have to have receptors that will be sensitive to light energy
sensors receptor to limited range of electromagnetic energy, any electromagnetic energy that is outside of that spectrum, we can't perceive it
colour perception
certain neurons that respond to diff wavelengths of light
different organisms have different receptors that respond to different range of wavelengths
e.g. same flower, but end perceptions depends on what energy our cells respond to
"invisible" gorilla
attention not focused anywhere other than ball
not focused on gorilla in background
without intention, it fails to reach conscious awareness
sensation vs perception
sensation is first two steps, physical stimulus in the world
signal is sent to cortex, cortex detects that is physical stimulus
perception adds third step, understanding, what is stimulus
e.g. is it familiar stimulus, what do i need to do about it
need everything to work together in order to perceive
the eye
main structures that allow us to see
pupil - hole that allows light to enter
iris - group of muscles, ring around pupil, can change size of pupil according to light
cornea - outside of eye, protective surface, helps to refract (bend) light
cornea is static, refracts light the same amount every time
want light to come to single point, to see light
e.g. when u zoom in on picture, things get blurry
lens - change size (structure and shape), combination
stretches to focus on things far away, refracts light less
becomes round and thick to see things that are near
receptors
optic disc - place where there are no photoreceptors, part of that retina that connects to optic nerve
optic nerve - carries out optic info to brain
not aware of blind spot bc eyes are constantly moving, so fill in blind spot with previous info (sensory memory)
fovea - exact centre of eye
retina
when light gets to retina, finds photoreceptors at very back
photoreceptors are almost final layer of retina
at back, pigmented epithelium
absorbs any light not
don't want any light reflected
nocturnal animals - don't have pigmented epithelium, want light reflected cuz less light at night
middle layer - two main important cells for vision
bipolar cells - same direction as photoreceptors
horizontal cells
first layer is ganglion cells -
photoreceptors: rods and cones
outer segment and inner segment
outside of retina, or closer to inside
outer segment of photoreceptor contains parts that respond to light
inner segment contains functions that keep cells alive
rods
rods only contain one photopigment, not sensitive to colour
is there light out in env
mostly located in retinal periphery (none in fovea)
sensitive to changes in light, focusing on one item, but want to know about movement in env
movement indicated by change in light
as moving, blocking light in diff locations
movement in periphery
even if focusing in one place, need some of the visual input for periphery and surroundings
that's why rods located mostly in periphery
only need to know about changes, don't need details
cones
visual acuity - fine details of things we are looking at
photopic system
used for vision in bright light
mostly located in the fovea
looking straight at thing, in bright light, want detailed info about thing we are seeing
ganglion cells and receptive fields
bipolar cells influenced by horizontal cells
horizontal cells act to inhibit response btwn neighbouring cells
that is important for signal that is sent past bipolar cells
signal starts at photoreceptors, signal is altered in middle layer by horizontal and bipolar cells, and altered signal sent to ganglion cells
every time light signal is passed to different type of cell, it is converged
larger signal getting converged to eventually reach single ganglion cell
amount of convergence differs btwn rods and cones
rods sending light to smaller # of bipolar cells, 2 bipolar cells send signal to single ganglion cell
cones have less convergence
1 ganglion cell will respond to area of back of retina
that area is ganglion cell's receptive field
receptive field of ganglion cell that receives signal from rods is larger than receptive field that receives signal from cones (small)
P cells
little convergence, small receptive fields leads to high acuity
connect to cones
M cells
large receptive fields, lots of convergence, poor acuity but high sensitivity
sensitive to wide area (large receptive field), so sensitive to things going on but not with good detail or focus
ganglion cells
on cells respond preferentially to small point of light that corresponds to centre of receptive field
off cells respond to ring of light, darkness in centre, ring of light around
response of those ganglion cells depends on how well
"on"
A activated
bar of light covered "on", will respond quite strongly
not as strong as for just a point of light
visual pathways
light from ganglion cell leaves retina through optic nerve
each eye can see the left and right half of space
left and right retina
visual fields -
both eyes receive info from left and right visual field
info being projected onto right is upside down and backwards
anything on left of space will project to right side of eye on both eyes
upper half of visual field, will project onto lower half of retina
two optic nerves meet up at optic chiasm
at this point, some of the axons continue ipsilaterally (same side of brain), so switch sides and continue contra-laterally
each retina receives info from both visual fields
after optic chiasm, left only from left
temporal retina - closest to outside of head
nasal retina - half of retina closest to nose
crossover at nose
info from right visual field when it travels to eye, will project to temporal retina of ....
temporal retina of one eye, nasal retina of other eye ...
superior colliculus (in tectum, midbrain)
we don't have conscious awareness of signals travelling to midbrain or thalamus
geniculostriate pathway
named after synapses
striate cortex, another name for primary visual cortex
optic nerve to optic chiasm to optic tract
no change in synapse, change in name
stop in thalamus
LGN lateral geniculate nucleus, organized in six layers
organized in same way as retina
project to adjacent cells in LGN from retina
each layer receives info from M cells or P cells
layers 3 to 6 receive input from p cells
parvocellular layers
magnocellular layers
each layer receives input from one eye
info travels via optic radiations to V1 ventrally
V1 (primary visual cortex, striate cortex) corresponds to Brodmann area 17
one running more dorsally
carries info from bottom of visual field
Meyer loop
carries info from top of visual field
Page 24 - damage to area, and corresponding vision loss
top diagram
red - left
purple - right
bottom diagram
red - vision
purple - no vision
if there is damage before optic chiasm, will lose vision from one eye or other (?)
damage to left optic tract
if damage is after optic chiasm, will not have full loss of vision to one eye or other
will have loss of vision in one visual field
will lose half of vision in visual field, will be in right visual field of both eyes
right visual field bc damage to left optic tract
damage to optic chiasm
info can't cross over
any info that travels ipsilaterally is fine
loss of info from temporal retina
V1
as bar of light approaches cell's preferred signal, more and more response
one single cell has preferred orientation, response drops off as u move away from preferred orientation
location column (hyper-column)
each ocular dominance column, contains orientation column
one orientation column contains cells that prefer particular orientation
colour sensitive blobs don't respond to orientation, only wavelengths of light
colour is processed separately from orientation and contrast
colour is processed through V1 and beyond
V3, V4, V5 (MT/MST)
V5, motion processing
MT/MST, research done in monkeys, analog to that is area V5
when start moving, rectangle looks more like cylinder
visual processing streams
"where" - dorsal stream
double dissociation
difficult to guide action towards objects
"what" - ventral stream
deficit
no difficulty reaching, grabbing or using object
but can't recognize what object is
visual agnosias are visual perception deficits
no damage in either case to eye or early visual pathway
no deficit in visual sensation
deficit is in perception
visual agnosia - deficit in identifying what object it
apperceptive agnosia
no identification through vision
sometimes can draw objects from memory
issue is connecting visual input w/ info stored in memory
inability to copy shapes
things are pixelated in early processing
lines and edges are processed separately from motion and colour
separate visual features, can't be combined into a whole
associative agnosia
can put visual features together
can't name
can't match form to memory
can describe what they're looking at but can't name what it is
ventral pathway: specialized cells
"cell hat responds only to grandmother"
might have single cell in visual processing areas that respond only to one thing
there is a debate
can't look at single cell, but looking at relatively small region in brain
1000s of cells
temporal lobe, hippocampus
cells in hippocampus that responded preferentially only to images of Halle Berry and not other celebrity
responded equally to character in movie that Halle Berry played and written name
most likely memory cells
memory of who Halle Berry is as a concept
and different cells to Jennifer Aniston
not typical surrounding activity around hippocampus or brain regions
thought to be evidence of sparse coding
very small group of cells, work tgt, to code for identity of one person
ventral pathway: specialized brain areas
fusiform face area
recognizing that it is a face
cells in FFA respond preferentially to faces
ventral surface of temporal lobe
parahippocampal place area
responded to images of houses
doesn't only respond preferentially to houses
scenes (more of cities than natural landscapes), buildings, locations
temporal lobe
extrastriate body area
disembodied body parts
very close to V1, just outside striate cortex
fusiform body area
silhouettes of entire bodies
LOC
human made objects
patient DF had damage to this area
apperceptive agnosia
STS
biological motion
social cognition
see living thing moving, and based on movement, see identity of living thing
see joints moving, but can't see outline of body, only outline of movement (shining light on joints and moving parts)
lifting smtg, walking, etc.
info from motion of biological being
e.g. bird, dog
these are all considered perceptual areas
not language, memory, etc.
locations of these not important for tests
superior temporal sulcus (more dorsal) important
btwn
STS processing stream
right through centre of ventral and dorsal streams
carrying info from area V5 (important for motion)
posterior temporal sulcus, anterior
generally accepted
STS pathway
integrate input from ventral and dorsal streams
involved in processing biological motion
direction of eye gaze
when u look at someone and their eye gaze is moving
facial expressions
audiovisual speech integration
combining auditory signal with mouth movements
face processing
if u have damage to specialized face processing area, will lose ability to recognize faces
prosopagnosia
"who does this face belong to"
different from object agnosias
double dissociated w/ object agnosia
need to know individual identity for face
there are individual differences btwn ppl in face recognizing ability
some ppl seem unable to recognize faces despite no brain injury
acquired prosopagnosia
recognize ppl due to other features, hints and tricks
e.g. may not recognize person out of context
face and house study
showed participants isolated parts of previously shown faces and houses
are ppl processing individual parts of these things?
for houses, participants good at recognizing parts of houses
for faces, performance was a lot worse for recognizing parts of faces than whole faces
worse for recognizing parts of faces than parts of houses
when seeing objects, we are processing the different parts separately
for faces, we don't seem to process separate parts, faces are processed holistically or con-figuratively
inversion effect
when face-like shape is in upright configuration, can recognize as face
but not when flipped
faster at processing
if want to know if other things are processed like faces, can look for inversion effect
why are objects processed analytically and faces con-figuratively?
is it bc FFA is specialized for processing faces or is it specialized for holistic processing?
bc faces are rly important
expretise, holistic processing as a result of processing
we don't have an answer, still ongoing
configural processing
we learn through experience (expertise) to use this type of processing (holistic)
we learn to move from step by step process to looking at relationship of the whole
since we see a lot of faces
e.g. u don't need to know if the person has a nose, u need to know relationship btwn eyebrows, nose, mouth, etc.
light from right visual field projects to nasal retina of left eye
false
temporal retina continues ipsilaterally
light from nasal retina crosses over
right visual field to left eye is already crossed over
bottom two layers of LGN are important for motion perception
M ganglion cells receive input from rods (signals of motion, sensitive to light)
deficit from lesions in
area V4 is part of which stream?
ventral, "what"
support for independent dorsal and ventral streams comes from double dissociations between
optic ataxia and apperceptive agnosia
apperceptive and associative
result from damage to