Object perception - VS Flashcards
which area in the brain of a monkey is all purpose simple form & all object processors found
V4/IOT (inferior occipital temporal cortex)
which area in the human brain is all purpose simple form & all object processors found
LOC (lateral occipital cortex) area
in both monkey and human brain, where are more specialised object processors found for recognising faces and buildings
IT (inferior temporal), fusiform gyrus, several areas
in both monkey and human brain, where are more specialised object processors found for spatial scenes/places and navigating through them
IT (inferior temporal)/ParaHippocampal gyrus
which serial processing is used to rebuild the object image
from: V1/V2 - V4 - IOT - IT (where real objects are encoded)
what are the form processing cells in area v1 & v2 and what do they respond to
orientation selective
respond to lines & edges
where are the form processing cells located in area v1 & v2
in interblob zones of layers 2-4A & 5-6 (in v1)
what are the form processing RF sizes in area v1/v2
relatively small 0.1 (fovea) - 2 (bit eccentric) degrees, depending on visual eccentricity
what are the form processing RFs organisation like in area v1/v2
- simple ~15%: linear discrete ON & OFF sub-regions
- complex ~75%: non-linear, overlapping ON/OFF regions
- hypercomplex ~10%: end-stopped inhibition, so also size selective and well as orientation selective
what do RF organisations of form processing cells in area V1/V2 interested in
only like lines/edges of a particular orientation and hate ones that are at 90 degrees
how many % off cells in area v1 are orientation selective and what do they respond best to
75%
respond best to contours (lines, edges) over a narrow range of possible orientations but not to other orientations of the same stimulus
what are the orientation selectivity cells in area v1 responses independent of, and except which population
independent of the size of the oriented stimulus
except among a small population 5-10% of ‘end-stopped’ or hyper-complex cells
what two sub-divisions is monkey area V4 believed to contain
- one primarily concerned with colour vision
2. other mainly concerned with form vision
what are the RFs of the sub division of monkey area v4 concerned with form vision size, compared to area v1 at equivalent visual field locations
2-5x larger
what component is 2-5x compared to another component in a monkey brain
the RFs of the sub division of monkey area v4 concerned with form vision size, compared to area v1 at equivalent visual field locations
what does monkey area v4 sub-division concerned with form vision, contain none of, and only of which RF organisations
no simple cells (beyond v1 & v2)
all RF organisations are complex or many are hyper-complex (end-stopped)
in monkey area v4 sub-division concerned with form vision what are ~50% of cells
orientation selective
but they ALL prefer stimuli of a particular size (width verses length)
what do the ~50% of orientation selective cells in monkey area v4 sub-division concerned with form vision all prefer
stimuli of a particular size (width verses length)
how is RF sizes in functional specialisation for object & motion perception areas V4, IOT & IT compared to V1 cortex
increased RF sizes at all eccentricities (neurons analysing a larger region of space)
how is orientation selectivity in functional specialisation for object & motion perception areas V4, IOT & IT compared to V1 cortex
reduced orientation selectivity and absent in IT
replaced by size-shape or object category-selectivity
what are orientation selectivity in functional specialisation for object & motion perception areas in V4, IOT & IT replaced by
size-shape or object category-selectivity
how are the specific form attribute in functional specialisation for object & motion perception areas V4, IOT & IT compared to V1 cortex
columnar organisation for that specific form attribute that the area is responsible for and specialised to encode
how are the representation of central vision where perception mainly occurs in functional specialisation for object & motion perception areas V4, IOT & IT compared to V1 cortex
biased representation of central vision, where object perception mainly occurs, in object processing pathway areas only concerned with central vision, central 20 degrees where we identify & recognise objects and peripheral VF representations disappear the higher up the pathway you go
how is the retinotopic order in functional specialisation for object & motion perception areas V4, IOT & IT compared to V1 cortex
loss of retinotopic order
what does RF sizes increase with for V1, V2, V4 & IOT & IT cells
eccentricity, by a linear correlation
what does the increase in size of RF with eccentricity increase gradually from
V1 to V2 to V4 & becomes really large in IT & IOT cortex
what is the increase in RF size in V1 V2 V4 & IT & IOT mirrored by
increase in RF complexity
what are V1 and V2 RFs mainly concerned with
fragments of objects (lines and edges)
what are V4 and IT & IOT RFs mainly concerned with
simple geometric shapes & real objects and complex shapes (components of objects)
what are most of the ~50% of v4 cells which are orientation selective
end stop tuned
to the size of the stimulus (like hyper column v1 cells)
where e.g. in a box RF, a line placed inside it must be confined to the RF size, if extended the RF will shut off
what do the other of the ~50% of v4 orientation selective cells which are not end stop tuned respond to
respond selectively to simple geometric shapes (squares, rectangles) of a particular size e.g. can like a small rectangle placed in its RF, or a square placed in its RF or a rectangle placed in its RF..
what do monkey IOT & IT cortex RF properties of single cells correlate with
object perception
what are the monkey IOT & IT RFs like in relation to V4 RFs
much larger than V4 RFs & always involve the fovea & central (10-20 degrees) visual field
what do the monkey IOT & IT RFs always involve
the fovea & central (10-20 degrees) visual field
what do the monkey IOT & IT RFs cells respond to what type of stimulus
a specific form stimulus anywhere in their RF & regardless of its size
what do IOT cells respond to
general shapes like squares or triangles, or to particular patterns
what do IT cells respond to
only category-specific objects, like faces, hands or types of fruit
what are the names of the dots/plots on a graph which shows a share cell in IOT response to particular shapes
rastor plots, the more there are, the more the IOT likes that shape i.e. stimulus fires lots of action potentials
why are there rows of dots on the graphs which shows the amount of rastor plots of a shape cell in the IOT cortex e.g. a row of 8 dots
indicates that the stimulus was placed into the cell’s RF on 8 separate occasions to look at the cell’s response to make sure the activity produced by one stimulus presentation was reliable and consistent from one to the next
what does an accumulative histogram show of a shape cell in IOT cortex
the number of action potentials that were fired shortly after the stimulus was presented, so a big mountain = a big response and no mountain = no response
what does a histogram of different types of squares of a shape cell in IOT cortex results investigate
if its components of a square which fires action potentials and shows e.g. the cell may not like a square with a hole in the middle or patterned stripe squares, but may like the solid square
what type or organisations are there for neurons in IOT cortex
columnar
what do the vertical columns represent for neurons in IOT cortex
response of cells are the same to those types of stimulus can be, simple geometric and complex shapes, hands or patterns
what do horizontally across the cortex cells represent in IOT cortex
types of cells you find are different preferences
what does a recording from a monkey face cell in IT cortex show
that it can only respond to one face of interest where it gives a big response (out of 4 faces which was placed in its RF, the responses were completely different and depended on the face that was placed there) regardless of the viewing angle or facial expression
in the ventral, what? processing stream, where is area v4 located
on the lingual gyrus
what does area v4 in the ventral what? processing stream represent
form & colour
in the ventral what? processing stream, where is the IOT/LOC located
inferior temporal sulcus and gyrus
what does the IOT/LOC in the ventral, what? processing stream represent
all objects, general object processor
in the ventral, what? processing stream, what does the IT/fusiform gyrus represent
specific objects (faces)
when images of a banana vs scrambled version of banana (stimulus) was presented to a person and the activity of the brain was monitored, what was the difference in cortical action with the LOC, compared to area V1
LOC only activated with the real object - banana
V1 was not activated when viewing banana vs scrambled version as cells in V1 like oriented lines in the scrambled version and the blocks and edges too
what type of objects is the LOC activated by
all types of objects
what area is only activated by faces
face area on fusiform gyrus (not interested in objects)
what usually happens in a general form agnosia (damage to LOC)
- conscious visual sensation of the components of objects is preserved (because area v1 & v2 is intact)
- but px cannot recognise whole objects that they see or understand their meaning (what they are and what they’re there for)
- although memory of what they look like & their purpose are preserved
what usually happens in category-specific agnosia (damage to fusiform gyrus)
- is restricted to particular classes of object
- most commonly faces = prosopagnosia, but can be just animals or buildings etc which can be damage to areas right near the fusiform gyrus
what can be seen in form agnosia, damage to LOC
lines and contours and other elements due to v1 being intact
what cannot be understood in form agnosia, damage to LOC
the over all picture, what it is
what can carbon monoxide poisoning effect
inferior occipital & temporal parts of the brain and can cause a bilateral LOC lesion
what remains intact from a LOC lesion
- visual fields
- VAs, stereo
- spatial acuity
- colours
- motion
what cannot be matched with a person with LOC lesion
simple shapes or simple objects
nor can they copy-draw familiar objects but can reproduce-draw them from memory
what can people with prosopagnosia: damage to fusiform gyrus not recognise
faces, including their own
what can people with prosopagnosia: damage to fusiform gyrus identify
familiar people from other visual cues e.g. hair colour, spectacles & from their voice
can also interpret emotions in facial expressions, even though unable to identify to whom the face belongs to, as this involves another part of the brain called the amygdala
what are the physical properties of real objects e.g. your own face
invariant, they are immune to changes n ambient illumination, viewing angle, size, location
doesn’t change from moment to moment, regardless of other things that do change
what is the fundamental challenge of the perceptual ventral stream
is to create object-based representations of the physical constancies, while ignoring these other random factors e.g. when light levels drop
extra striate cortex contains neurons that represent…
object constancies e.g. faces, regardless of their viewing angle, location, size or expression
but what do the perceptual constancies for object forms, colours & motion seem to be represented in
different, widely separated (located in different parts) extra striate cortical areas
what are objects with similar features (faces or volkswagons) represented by
many single cells as a ‘population code’ e.g. face cells in IT cortex respond to several faces with similar features, not just to one (such as cells in a monkey brain tend to encode particular categories of people, e.g. some cells like people with narrow thin noses with eyes close together but hate, people with round flat noses with eyes wider apart and vise versa) so has category specific rather than groups of faces in the IT cortex
in the extra striate cortex, only damage to the fusiform gyrus can cause…
prosopagnosia
in the extra striate cortex, none of which cells are orientation selective
LOC cells
in the extra striae cortex, damage to area V5/MT in the left cortex would result in…
right hemi-akinetopsia
