PAA 4

retinotopy

neighbour relationships in retina are retained in the cortical map  

organisation of V1

Neurons all responding to the same type of orientation stacked on top of each other

-> orientation columns  

 

All orientations covered within about 0.5mm 

Ocular dominance bands  

hypercolumn

a part of V1

neurons analyse the same part of visual space and respond to all different types of orientations (orientation column), and neurons that respond mainly to the right eye and mainly to the left eye  

Essentially a processing device that extracts what orientations are present in visual space and what shape and size 

Sinewave or Sinusodial gratings – look at luminous profile 

Different waves of luminsocity from black to white -> a smooth gradual change  

Can manipulate orientation and width of bars (spatial frequency) - how often does the stimulus go from black –white-black again 

Spatial frequency = 10 cycles/image e.g. 

 

Contrast = changing the depth of the lines  

Phase = starting with black, white or in between  

visual angle

The visual angle tells us the size of the retinal image (given a certain object size at a certain distance). 

larger angle = larger visual image

comparing spatial frequency with contrast

Jean-Baptiste Joseph Fourier

• Every image can be broken down into sinewave components – Fourier analysis 

-> by studying sinewave components can help us learn about how the visual system works  

 
• The visual system conducts the equivalent of a local Fourier analysis 

Fourier analysis

allows you to break down visual image into the sinewave gratings its composed of (and put it back together to create image again) 

smallest units of the image

early vision as local Fourier analysis

Hypercolumns contain neurons tuned to different orientations and spatial frequencies  

All of these neurons analyse the same patch of visual space 

-> essentially conduct a local Fourier analysis of the visual space as together they extract spatial frequencies and orientations contained in their local patch  

e.g. Top left neuron reacts, middle neuron reacts and bottom right neuron reacts to this patch of stimulus 

adaption

• Both a method and a process in the visual system 
• Method: Stare at the same stimulus for a long time 
• Process: As a consequence of long exposure, those neurons that are tuned to the stimulus property decrease their sensitivity 

V1 receptive fields

population code of tilt aftereffect

The more neuron responds, the more neuron fatigues 

The neuron that best responds to vertical lines, is not the one that responds the most as fatigues 

The lines around it has less adapted, so responds more, biases the orientations to the left when is in fact vertical  

contrast sensitivity and spatial scale

whether you are equally sensitivie to different spatial scales

We are mostly sensitive to midrange spatial frequencies  

Sensitivity of our visual system differs for different visual spatial scales  

contrast sensitivity function (CSF)

like a window of visibility of spatial scales

peak sensitivity in the midrange

what produces the CSF?

populations of neurons tuned to similar spatial frequency forms a spatial frequency channel

these spatial frequency channels have different sensitivities

close by see as Albert Einstein - presented in high spatial frequency channels

the distinct channels of neurons that process different spatial frequencies

pasted together different frequencies of the two images

far away Marilyn Monroe - presented in low spatial frequency channels