-
vision
constructive process, take in light and infer what must be present in the world
- we see what we expect
- unaware to things we aren't conscious to
- we see what we need to know
- hard to see without expectations
-
light to vision
sensory transduction: process by which a signal from the world is transformed into electrochemical signals in the nervous system
photoreceptors in the retina transduce a narrow band of the electromagnetic spectrum (“visible light”)
-
Anatomy of the Eye
light passes through cornea, iris controls amount of light that enters, goes through pupil and lense and focuses on retina
-
retina
photoreceptors
rods and cons capture light in electrochemical activity through phototransduction
-
retina
phototransduction
- light strikes pigment molecule in the
- photoreceptor, causing the pigment molecule to break into pieces, which then act on proteins in the cell to change its resting membrane potential
-
Retina information flow
photoreceptors --> bipolar cells ---> ganglion --> optic nerve
amacrine and horizontal cells communicate with other parts of the retina
-
rods
- LIGHT
- sensitive to light but do not respondselectively to any particular wavelength
- detects degrees of light, good for dim areas
concentrated in periphery
-
cones
- COLOUR
- less sensitive to light, but the 3 cone types are maximally responsive to different wavelengths
- bright environments
concentrated in fovea (central vision)
-
cones vs rods spatial resolution
- cones > rods b/c
- 1. more light can access photoreceptors directly, smaller ganglion
- 2. each cone connects to its own bipolar cell and then its own ganglion
-
ganglion cells
- each retinal ganglion cell only responds to a specific location of the visual scene (receptive field)
- info of where light originated maintained
- optimised to detect edges via differences in light levels from one to the next
-
centre-surround structure
- on center vs off center
- think circle with circle inside
-
off centre
- work best when centre on and surround off
- weak when all on
-
on centre
- best when centre off and surround on
- weak when all on
-
eyes to brain
L and R visual fields, temporal semiretina signals stay ipsilateral, nasal semiretina move contralaterally, go to visual cortex
-
retinal ganglion cell types
- parvocellular: input from cones. Fine details, form, colour
- magnocellular: input from rods. Depth, brightness, movement
-
eyes to thalmus
- retinal ganglion cells are sent to the lateral
- geniculate nucleus (LGN) in the thalamus
-
thalamus to visual cortex
axons travel from the LGN to V1 via the optic radiation
-
visual cortex
v1
- V1 contains a retinotopic map: each neuron responds to a particular part of the visual field
- v1 has new neurons
-
primary visual cortex
- simple cells: preferred orientation at particular place
- complex cells: preferred orientaiton at any place
-
binocular disparity
- each eye gets slightly different view of the world, how we see in 3D
- stereopsis: using difference to infer depth
-
V2
- direct connections from V1, also retinotopically (topographically) organized but with larger receptive fields
- illusiory contours
-
V3, V4, V5
- cells respond to more abstract stimuli in their receptive fields (e.g., shapes, textures, movement)
- V5: specialized for motion detection, sends signals to the parietal lobe
-
inferotemporal cortex
- representation of abstract and complex stimuli (e.g., faces, scenes, objects)
- no longer organized retinotopically
-
processing streams
ventral and dorsal
-
ventral stream
- WHAT pathway
- from parvocellular cells in LGN, through V1, V2, V4 and inferotemporal cortex
- tools, animals faces
- position and size invariance
-
dorsal stream
- WHERE pathway
- from magnocellular cells in LGN, through V1, V2, and V5
-
fusiform face area
recognises everything as faces
- prosopagnosia: inability to recognize face
- visual agnosia: an impairment in the ability to recognize objects by sight
-
sparse coding
small number of clustered neurons that become active in response to a specific visual stimulus
for stimuli that we are expert at
-
population coding
- many neurons distributed over cortical regions are involved, showing varying levels of activity
- the pattern is important
-
motion after effect
- perception of an opposite direction of motion after prolonged viewing of a moving stimulus
- adaptation to reduce response over time
neurons will be firing below baseline in favour of the opposite direction of motion
-
motion blindness
- happens after damage to motion detecting areas in the dorsal stream (e.g., V5)
- can't detect motions, just looks like a bunch of still photos getting closer
-
spatial attention
- directing attention to particular locations in space
- dorsal stream is critical for guiding and adjusting the “spotlight” of attention
-
disorders of attention
- hemineglect: ignores half of the world; caused by damage to the right parietal lobe
- balints syndrome: cannot comprehend the big picture of a visual scene though they can see objects individually
- simultanagnosia: inability to recognize multiple elements in a scene
-
binocular rivalry
if different images are presented to each eye, perception shifts back and forth between them
-
unconscious inference:
- the brain infers what is likely
- to be out there given incoming signals
|
|