perception - Visual architecture 1

Question 1

sensors in the-retina

what is the optic nerve
blind spot
macula fovea

Answer 1

-optic nerve (nerve that takes information from the eyes to the brain)

-blind spot-place where the optic nerve starts , no rods or cones here, cant see anything if image is cast here

macula fovea - most sensitive part of eye, whenever light hits , it hits there

Question 2

human retina
-how many photoreceptors/megapixels in eyes
-how many cones in the fovea?

Answer 2

-126 mp (megapixels) in each eye (more than any current camera)
- 6 (million) mp are cones (which is actually very poor)
- 12- mp are rods

-fovea has 200,000 cones (most sensitive part of the eye where the light is focused) (so only 0.2 mp)

eyes are good but not the sole reason why we see so well

Question 3

why is human vision so good?

Answer 3

-because of our brain

-enormous processing power dedicated to vision /perception
-15-20% of cortex dedicated just to vision
-and about 40% is involved in vision
-a quarter of what you eat fuels the visual cortex in the brain

Question 4

visual pathway from diagram explained

Answer 4

left visual brain
-light from the left visual field goes to both eyes, so to the inner part of the left eye (the nasal part of the retina) and outer parts of the right eye , which is temporal part of the right retina.
- so light goes to both of those places and then it travels to the brain in different pathways

-following the left eye, info enters the left eye and crosses over at the optic chiasm (intersection) and goes to the right half of the brain

-following the information that comes into the right eye, it goes to the optic chiasm , doesn’t cross over and stays on the right part of the brain

-any info present to your left is being is being processed by the right brain

Question 5

• any information present to your left is being is being processed by the _____ brain
  -any info present in right is going to the ____ brain, ____visual cortex

Answer 5

right

left

Question 6

what happens when light reaches the cortex

Answer 6

-light goes into eye, then is converted into neural signals and it goes to the lateral genocide nucleus (LGN), and then in comes to the cortex

-when light reaches the cortex it is being processed by the visual cortex (occipital cortex) but it is processed in very systematic ,different steps

Question 7

steps after light reaches cortex

Answer 7

firstlight it is processed by the region called V1 because its the first visual region
-from v1 goes to v2,to v3, to v4 and other regions

this is all at the back of the brain in the occipital cortex

Question 8

V1- primary visual cortex (also called striate cortex)
-where is it?
-where does it receive input from

Answer 8

-sits around the calcarine sulcus (back of brain)
-receives input from LGN -lateral genocide nucleus
-first part of the cortex to get visual input
(eyes get info that goes to lgn (subcortical region, and then comes to v1)

Question 9

how many layers does the cortex have

Answer 9

six layers (all other parts of cortex have 6 layers each apart from hippocampus)
-each layer have diff cells, processing , different kinds of connections, inputs and outputs

Question 10

why is the primary visual cortex sometimes called striate cortex?

Answer 10

if you stain and look at the different layers if v1 cortex, some parts are darker than others

-the darker part is due to striations which are a sort of lines, because they have extra fibres coming in from the LGN to give information to the V1

-v2,v3,v4 called extra striate because they are outside the striated part of the brain

Question 11

macroscopic level diagram of visual field

Answer 11

-look at diagram in slides around 23-24 mins
retinotopic layout
-diagram shows your visual field split up into numbers and then how the cortex processes it (number by number shown)

-(not only is visual field and the processing part (left field processed by right brain eg) swapped left to right its also swapped top to bottom)

middle of map (centre numbers) small parts of the field processed by manyyy neurons, whereas outer larger parts of visual field (outer numbers) processed by small amount of neurons

Question 12

cortical magnification

Answer 12

Cortical magnification refers to the fact that the number of neurons in the visual cortex responsible for processing the visual stimulus of a given size varies as a function of the location of the stimulus in the visual field.

Stimuli occurring in the centre of the visual field that have been detected in the fovea of retina are processed by a very large number of neurons in the primary visual cortex of the occipital lobe, though these neurons handle only a very small region of the central visual field.

Conversely, stimuli detected in the peripheral visual field tend to be processed by a much smaller number of neurons in the primary visual cortex.

Question 13

primary visual cortex function

Answer 13

-v1 processes basic building blocks of your perceptual world, known as features

Question 14

v1 processes building blocks (features) what are the features?

Answer 14

-orientation, edge
-colour
-spatial frequency
-depth
-motion
(there are more but these are some of the main ones)

Question 15

receptive field

Answer 15

a part of the visual world that elicits a response in that neuron
(any given neuron in the visual cortex doesn’t see the entire world-only you do)
-so the part of the world any neuron sees is its receptive field

Question 16

how does v1 process orientation?
experiment that discovered this

Answer 16

-using electrodes can see what neurons are responding to (put electrode next to neuron and they would present something in its receptive field and then see how the neuron responds to it.

-if the stimulus is angled a certain way the response is more (more neurons firing) eg horizontal stimulus had not many neurons firing
(every time a neuron fires its called a spike)
-if you count up all the spikes you can then plot it giving a graph (so shows different orientations have different cell responses

Question 17

tuning curve

Answer 17

look at slides
the curve shows how neurons prefer vertical lines and fires the most when the stimuli orientation is vertical

Question 18

‘simple’ cells

Answer 18

-simple cells have receptive field structures which have both stimulatory and inhibitory regions
-many diff kinds of simple cells
-if light falls in the positive region it activates the cell etc, if falls in negative region it inhibits the cell

Question 19

What happens if you
- present some stimulus outside the receptive field of a ‘simple cell’
-present stimulus on the stimulus or inhibitory part
-present stimulus of different orientation

Answer 19

-neuron doesn’t know bout the receptive field
_but if you put the bar in the stimulatory part of the cell it will many spikes
-if its in the inhibitory part it doesn’t fire (supresses it)
-when you take the stimulus out, (after supressing then taking it out) it’ll fire again then go back to baseline
-if you present stimulus in different orientation then it doesn’t fire

Question 20

simple cell - why so important

Answer 20

-this is how v1 actually knows objects of different spatial frequencies

Question 21

simple cell is constructed from …

Answer 21

-simple cell is constructed from the outputs of different LGN neurons
-there are LGN neurons which have different circular receptive fields in the retina
-in the lgn u have circular receptor fields, and if you add them all together you get simple cells like this

Question 22

what is a complex cell

Answer 22

complex cells : receptive fields without inhibitory regions , only have excitatory regions

Question 23

complex cell
what happens when you present something outside its receptive field

Answer 23

-when you present something outside the receptive field, it doesn’t fire
-if you present something within it, just like a simple cell, it fires
-but you can present this in any location in the cell and itll still fire unlike the simple cell
-

Question 24

why are complex cells called complex

Answer 24

because its position invariant
-the position of the thing being presented doesn’t matter, doesn’t care about the precise location of the object

-another reason its called a complex cells is because to build a complex cell you need multiple simple cells like this (add outputs of many simple cells)

Question 25

hyper complex cells

Answer 25

like a complex cell receptive fields but with inhibitory regions at the ends (end stopped cells)

Question 26

what happens when you present something outside the receptive field of a hyper complex cell

Answer 26

-outside the field it doesn’t fire
-inside it fires as long as its smaller than the excitatory region
-if you present a bigger bar it wont fire because now you have positive and negative regions which cancel each other out

Question 27

how is v1 organised

Answer 27

-v1 has 6 layers
-orientation sensitive cells are organised systematically
-columns in v1: adjacent neurons have slightly different preferences in their sensitivity. so the kind of stimulu that they preferentially respond to keep changing in a systematic way throughout that region
-other cells in the same column so from going from layer one to all the way to layer six, almost all cells in that region,will prefer similar orientation of the objects that they are seeing.
-all the cells here will more or less fire to the same kind of orientation in these regions
-the columns are called orientation column
-multiple orientation columns,within an orientation column, the cells or the neurons prefer a particular orientation

Question 28

why is the visual system organised this way

Answer 28

-it allows it to be a powerful tool in trying to figure out what the world out there
allows for population coding

Question 29

population coding

Answer 29

Population coding is a method to represent stimuli by using the joint activities of a number of neurons

-v1 structure uses its organisation to try to figure out what is there in the world
-it can do this because we have something known as tuning curves
-eg there is a single neuron which has this particular tuning curve showing it prefers eg the vertical ones and does not prefer the horizontal ones
-therefore each cell will have their own tuning curve (each column has its own tuning curve)

-advantage of the system: lets say in the external world there is a particular orientated line like that , and you want to know what its orientation is , because in order to construct an object you need to know where the edges and lines are
-system allows you to precisely know the orientation of the line (diff cells are all seeing the same bar and each of them will fire to a diff extent.