Perception

Front 1

naive realism
for and against

Back 1

“perception directly reflects reality”

In favor of naïve realism- Perception feels effortless, immediate, and direct
• Against naïve realism-
o computer scientists in the mid 20th century thought it would be easy to build a visual system, still working on it today.
o “easy problems are hard”- 4 year olds can say sentences, insanely difficult for a computer to be programmed.
o “hard problems that are easy”- hard for humans, easy for computers- information crunching
o reality and perception sometimes don’t match- illusions

Front 2

o subjective idealism -
for and against

Back 2

“reality is subjectively or socially constructed” there is no objected external reality that is independent from our visual experiences
o in favor of subjective idealism
• perception can be wrong
• people’s perceptions can differ
o Against subjective idealism
• Just because we can be wrong doesn’t mean reality doesn’t exist. We are imperfect observers.
• Differing perceptions aren’t random- we can usually track down the source. It is systematic and understandable. Example: one person is color blind, or perceiving in different lighting conditions.
Reality is what can kill you, even if you don’t believe in it

Front 3

o the "in between" view (representative realism)-

Back 3

“perception is intelligent guess-work”
• sensory information is “gappy” and “noisy”. Because of this, your brain does intelligent guess work
• the world contains regularities, and our brains take advantage of those and fill in the gaps of fuzzy and noisy information we perceive.

Front 4

perception as a reconstructive process (intelligent guesswork)

Back 4

the job of our perceptual systems is to track reality. Have evolved to do the best possible guess work the brain can

Front 5

• distal stimulus-

Back 5

what’s out there

The tree that you see
The dog whose barking you hear
The baking cookies you smell

Front 6

• proximal stimulus-

Back 6

what your receptors receive

Light on your retina
Pressure waves against your eardrums
Molecules in your nasal cavity

Front 7

the inverse projection problem-

Back 7

we are getting the distal stimulus figured out by the proximal stimulus
o How to reconstruct the distal stimulus from the proximal stimulus.

Front 8

panda inverse projection problem example

Back 8

Distal and proximal stim. And invsere projection problem
Information goes to the back of the brain where the primary visual cortex is. The part of the brain asks “what is this?”. Goes somewhere else in the brain and you end up with a not perfect but a pretty accurate perception of the information.
Information (example: panda) distal stimulus
Proximal stimulus- the eye attaining the information of the light
Inverse projection problem – what the hell is it?
Inverse projection problem solved- the mental representation of a panda (not an actual panda)

Front 9

Transduction-

Back 9

physical contact with the world turns into electrical impulses in your nervous system. Physical attraction from the outside world to the nervous system

Front 10

top-down-

Back 10

prior knowledge, expectations, ect., influence how a stimulus is processed. Can literally change how we see something

Front 11

Bottom-up

Back 11

connecting with senses with no prior knowledge and expectations

Front 12

behavioral methods

Back 12

o Phenomenology/psychophysics

accuracy, and reaction time

Front 13

• Phenomenology-

Back 13

what a stimulus “seems like” to an observer. Subjective to the observer.

Front 14

Psychophysics:

Back 14

study of relationships between physical properties and phenomenology

Front 15

• Absolute Threshold-

Back 15

the smallest amount of stimulation that you can detect.

Front 16

o Threshold-

Back 16

the point where a system goes from not-responding to responding. A threshold and JND is not strictly a point. There is always some slop in the system. Detect it some of the time then most of the time then all of the time.

Front 17

• Just Noticeable Difference (JND)-

Back 17

100g vs. 100g+5g. can tell that the right one is heavier. 5g is the JND.

Front 18

• Weber’s Law:

Back 18

JND is proportional to the magnitidue of the stimulus. The bigger the stimulus the bigger the chunk of the weight of the JND.

Front 19

• Accuracy

Back 19

The word Superiority Effect- Easier to identify a letter in the context of a word than by itself. (top down processesing) (notice that the word “shouldn’t” make it easier)

Front 20

• Reaction time-

Back 20

measuring milliseconds. Shows how the brain does its processing.

Front 21

physiological methods
•

Back 21

Single cell recording
Lesions
Brain imaging, PET, fMRI, ERP, TMS

Front 22

single cell recording

Back 22

observe changes in voltage or current in a neuron.
electrode measures the changes in charge as the neuron reaches its action potential.has the highest resolution of all brain imaging techniques.

Front 23

lesions

Back 23

lesion is any abnormal tissue found on or in an organism, usually damaged by disease or trauma

Front 24

PET

Back 24

o PET (positron emission tomography) radioactive tracer that goes into the bloodstream. More blood goes to the parts of the brain used the most in a particular activity.
• Okay spatial resolution
• Poor temporal resolution

Front 25

fMRI

Back 25

o fMRI (functional magnetic resonance imaging)
• seeing images in “slices”. Either Horizontal, Coronal (square looking at the face), Sagittal (square looking at the profile)
• excellent spatial resolution
• good temporal resolution- resolution in an order of a couple of seconds. See differences in a couple of seconds in the brain.

Front 26

ERP (Event Related Potential)

Back 26

• ERP (Event Related Potential) take EEG and relate it to a particular stimulus, and average it. record a bunch of places on the brain.
• Poor spatial resolution (but getting better) going through the skull, mutes waves
• Excellent temporal resolution

Front 27

TMS (transcranial magnetic stimulation)

Back 27

• TMS (transcranial magnetic stimulation)
o Magnetic pulse temporarily disrupts processing in a particular area. Also used as a therapeutic process with depression using multiple pulses.

Front 28

coronal slice-

Back 28

(square looking at the face),

Front 29

Sagittal

Back 29

(square looking at the profile)

Front 30

why is the cortex wrinkly? also, what does it consist of

Back 30

so it has more surface area. also it has 6 layers and white and grey matter. Some places might be thick or thin. This can be because of different type of cells or different thicknesses of layers

Front 31

grey matter

Back 31

cell bodies at the surface of the cortex

Front 32

white matter

Back 32

underneath the grey matter of the cortex. consists of all the connections

Front 33

Brodmann's areas-

Back 33

patches of cortex defined by organization of the layers. Went through the whole cortex and found out where they change and differ.

Front 34

Gyrus-

Back 34

lumpy bulge

Front 35

Sulcus

Back 35

grooves

Front 36

left vs. right hemisphere. what connects them?

Back 36

Left hemisphere: language processing
o Right hemisphere: spatial processing
o Connected by the corpus callosum

Front 37

where are the frontal, parietal, occipital, and temporal lobes located?

Back 37

o Frontal – in the front
o Parietal- on top
o Occipital- back
o Temporal- near the temples

Front 38

o Thalamus-

Back 38

first place sensory information goes. Receives and processes information

Front 39

• Cerebellum-

Back 39

has layers and hemispheres, and involved in motor coordination among other things. Outputs information

Front 40

spinal cord-

Back 40

Sensory and motor nerves
o Enter/exit brain through spinal chord

Front 41

dendrites-

Back 41

receiving imput from other cells

Front 42

axon-

Back 42

usually covered in myelin sheet which speeds up the process

Front 43

terminal buttons-

Back 43

Terminal endings with terminal buttons leads to the synapse which is a space between two neurons

Front 44

synapse with different functions-

Back 44

A synapse can be excitatory or inhibitory. A certain neurotransmitter can make a neuron excited or shut down, depending on the neurotransmitter and neuron. Always the same, not random.

Front 45

why a neuron fires

Back 45

o A neuron “sums” the inputs, if the sum exceeds the neuron’s threshold, the neuron fires.

Front 46

how a neuron fires

Back 46

• Reaches activation threshold
• Action potential is sent down the axon. Once it starts, cannot be stopped.
• Over time, excitatory and inhibitory inputs affect rate of firing. The more inhibitory, over time, the slower the rate of firing becomes.

Front 47

action potential

Back 47

occurs when a neuron sends information down an axon, away from the cell body. The action potential is an explosion of electrical activity that is created by a depolarizing current. This means that some event (a stimulus) causes the resting potential to move toward 0 mV.

Front 48

Retina

Back 48

two dimensional surface on the back of the eye. covered in photorecepotrs rods and cones

Front 49

rods

Back 49

black and white, dim light, movement

Front 50

cones

Back 50

color, detail

Front 51

Photopigment-
dark and light adaptation

Back 51

Photoreceptors contain photopigments
• This molecule changes shape when light hits them, and eventually they change back
• Light adaptation: lots of light- less photopigment available. Bright light looks less bright then previously. Get used to it
• Dark adaptation: little light- more photopigment available. Get used to the dark, can see more.

Front 52

purkinje shift

Back 52

• Rods and cones dark-adapt differently
• Creates the “purkinje shift” cross over to use mostly cones to mostly rods. (dark adaptation) when at the cone threshold, rods start to give more information. Seeing with rods gives you a big difference of how you perceive.

Front 53

Fovea-

Back 53

small patch on the retina that has the sharpest vision. We have the impression that we see a whole scene clearly because our eyes jump around- saccades. Our eyes jump around about 5 to 6 times a second.
o The fovea is all cones and no rods
o Other layers of cells are pushed aside so the light goes straight to the cones.

Front 54

blind spot (optic disc) and why it's there-

Back 54

Area where all the axons converge. All cells all over the retina comes together to send information to the brain. No room for photoreceptors.
o No vision at all
o Our brains “fill in” the blind spot. They make a best guess of whatever else is out in the visual field

Front 55

convergence of photoreceptors onto ganglion cells-

Back 55

Photoreceptors “funnel” to fewer ganglion cells (ganglion cells connect to the optic nerve). More photoreceptors funnel down to fewer bipolar and then to even fewer ganglion cells.
o This is because “sums” or “pools” the light input, allows you to detect very tiny amounts of light.
o More convergence for rods than cone. Cones: small receptive field, good detail vision. Rods: large receptive field details is “smeared”. Can see color a lot better when you are focusing on it. can see dim light better when you are not focusing on it.

Front 56

receptive field-

Back 56

region of the retina from which a cell gets input

Front 57

lateral inhibition-

Back 57

cells that converge may compete with each other. This creates a “center-surround” receptive. When a cell is excited, it shuts off its neighbors.
More on Lateral Inhibition
• Photoreceptors excite “their” ganglion cell
• They inhibit neighboring ganglion cells
• Every little patch of retina is in competition with its neighbors
o If all are receiving light, nobody wins, nobody loses
o If one is receiving more light, it can inhibit its neighbors
• Results: strong signals get stronger, weak signals get weaker

Front 58

Center-surround receptive field-

Back 58

light in the center of the RF turns the cell on. Light in the surround area turns the cell off.
Why lateral inhibitions? Because it emphasizes differences, and sharpens edges.

Front 59

Mach bands-

Back 59

area right between the lighter and darker square “line”. It’s a general strip that is right next to the borderd that it tends to look even brighter and darker. Enhances contrast

Front 60

• Magnocellular

Back 60

Distributed across the retina
More input from rods than cones
Large receptive fields
Respond to movement

Front 61

• Parvocellular

Back 61

o Mostly at or near the fovea
o More input from cones than rods
o Smaller receptive fields
o Respond to color and detail

Front 62

Pathway to the Cortex

Back 62

o Retina
o Optice nerve
o Optic chiasm
o Lateral geniculate nucleus (LGN) in thalamus
o Primary visual cortex

Front 63

left and right eye, visual field and LGN

Back 63

left LGN gets information from right eye and left eye, right visual field, left side of the retina
right LGN gets information from right and left eye, left visual field, right side of retina

Front 64

electromagnetic radiation-

Back 64

color stimulus. a form of energy exhibiting wave-like behavior as it travels through space.

Front 65

visible spectrum

Back 65

- light

Front 66

wavelengths and spectral colors

Back 66

o Different wavelengths correspond to different (perceived) colors
o Maximal sensitivity in the middle, which is yellow to us. Yellow is our peak, that is why it looks so bright to us. That is also why we cannot see a dark color

Front 67

non-spectral colors

Back 67

• Most colors have more than one wavelength

Front 68

reflectance curve

Back 68

o All colors have a “reflectance curve” most colors are not just some of this wavelength, and some of that one. They are usually a whole continuous curve of color wavelength.

Front 69

trichromatic theory Theory:

Back 69

three basic color codes in the brain
Early evidence: only 3 colors needed to “match” (perceptually) any color. Usually colors are a continuous wavelength of many colors, but we can perceive any color with only three wavelengths.

Front 70

3 cone types in the retina-

Back 70

there are in fact three types of cones, that maximal respond to short, med, or long wavelength.
Approx. “blue” “green” and “red” blue=short green=med red=long

Front 71

metamers (a consequence of having a trichromatic system)-

Back 71

Different spectral curves can look the same to a human
o Not true that the two different color spectrums can be metamers of each other always. Two colors can be metamers in one light and not in another!
• Perceived color depends on reflectance curve of the object and spectrum of the light source.

Front 72

opponent process theory

Back 72

- our basic color codes in the brain
• pairs of opposites (red/green, yellow/blue) (I’m seeing red, and I’m really not seeing green)

Front 73

colorblindness as evidence for the opponent process

Back 73

- Color blindness goes in pairs (red/green, yellow/blue)

Front 74

opponent cells in LGN-

Back 74

The LGN has “opponent cells”
• Each cell has a spontaneous firing rate (reminder: all cells have a resting rate of firing)
• Each cell increases firing to one color, decreases firing to the opposite color. ( a reduction in firing is not absense of information, it is important for your brain to know)
• Four types of cells: b+y-, y+b-, r+g-, g+r-
• Plus, dark/light opponent cells

Front 75

how we get from 3 colors in the retina to 4 colors in the LGN

Back 75

o In terms of wiring, red+green=yellow
o R+G- cells:
• Excited by “red” cones, inhibited by “green” cones
o Y+B- cells:
• Excited by red and green, inhibited by blue
• Retina has photoreceptors that are sensitive to 3 different wavelengths. Lgn is interested in 4 colors. Retina is red green blue. Lgn has yellow too. Red and green become yellow.

Front 76

• Monochromats-

Back 76

only have short-wavelength cones or lack cones altogether (rods only). Everything looks shades of grey. Get shading

Front 77

• color constancy

top down or bottom up?
.

Back 77

An object appears to keep its color under different lighting conditions
o We perceive (reconstruct) the object’s color, not the actual reflected wavelength

(top down)

Front 78

• lightness constancy-

top down or bottom up?

Back 78

(top down)
o An object appears to keep its amount of lightness or darkness under different amounts of light
o we perceive the objects lightness, not the actual amount of reflected light.
o Luminance edge- difference of color because light falls on the different sides differently
o Reflectance edge- brain will conclude that the two shades are actually different colors, not because of light.

Front 79

neon color spreading

Back 79

o The brain fills in gaps in color
• Pale “neon” colors spread into white areas
• Spreading isn’t random
• Brain perceives transparent “overplay”

Front 80

color assimilation-

Back 80

The brain “averages” color over small regions

Front 81

o basic color terms-

Back 81

Basic color terms example: red, green, blue, orange, yellow, purple

Front 82

picking the "best" focal color in a category
o learning new words for new color categories-

Back 82

New color terms are easiest to learn for focal colors. Easier to learn the difference between yellow, orange, red then three type of green, even when they are all have the difference same wavelengths

Front 83

layers in V1, visual cortex-

Back 83

Parvo and magno pathways
o Layer 4 is the input layer from the LGN

Front 84

Magnocellular (shapes) and parvocellular (detail and color) input
retinotopic map- Cortical magnification:

Back 84

in V1, fovia gets a lot of space

Front 85

simple cells

Back 85

edges and bars. in the striate cortex

Front 86

how to build a line detector or a bar detector

Back 86

• Edges and bars (simple cells)
o Specific orientation
• Fire response of a cell will fire a lot if the line likes it, to its preferred orientation. If it is way off, it doesn’t fire, if it is a little off, it’ll fire some.
• Lateral inhibition creates a contrast, when seeing different tilted and vertical lines simultaneously.
o Specific location
o Specific color
o Specific thickness (spatial frequency)

Front 87

• Columnar organization

Back 87

o Columns of the cortex prefer the same optimal orientation. Same goes for special frequency and ocular dominance (right vs left eye)

Front 88

• Hypercolumns

Back 88

o Each hypercolumn is for one retinal location
o They each contain lots of smaller columns

Front 89

tuning curve

Back 89

a cell doesn’t respond only to it’s preferred stimulus

Front 90

tilt after-effect-

Back 90

after looking at lines that are straight for a while and look at one that is horitzonal, itll look like its going the other way

Front 91

blobs

Back 91

color sensitive located in hypercolumns

Front 92

ventral stream

Back 92

"what" pathway,

mostly parvocellular

–“Ventral” DOWN

–V1, V2, V4, temporal lobe

–Areas dedicated to:
•color
•objects and shapes
•places (e.g. rooms, houses, outdoor scenes)
•faces
•human figures and body movements

Front 93

Distributed coding (why it’s more likely)

Back 93

each person is a mixture of cells. If cell is destroyed then everything goes down a little, not just one all the way

Front 94

dorsal stream

Back 94

"where" or "how" pathway,

magnocellular
–“Dorsal” means UP

–V1, V2, V3, MT, parietal lobe
–The “where” system (or “how” system)
–Areas dedicated to:
•motion
•influencing eye movements
•shifting attention to different locations
•preparing to reach, preparing to grasp

Front 95

the “put the card in the slot” evidence

Back 95

woman had damaged in her ventral, what. She could put it in the slot but couldn’t line it up.

Front 96

where is the ventral, dorsal and v1 located?

Back 96

Temporal is part of the ventral
Periatal lobe is where the dorsal stream goes
V1 is in the back occipital lobe

Front 97

simple and complex cells are located...

Back 97

in v1, primary visual cortex

Front 98

•1880’s - 1920’s: Structuralism

Back 98

–Perception is built up from little elements
–bottom-up

Front 99

apparent motion-

Back 99

railroad lights. See it as a light going back and forth not just two lights one going on and off

Front 100

illusory contours

Back 100

- filling the lines in. almost see the lines that aren’t there. Box example

Front 101

amodal completion-

Back 101

square in front, circle in back. We see a square in front of a circle, not a half circle on top of a square.

Front 102

good continuation-

Back 102

perceive two lines that cross each other, not a line that bends right at the middle and make angles. Visual system does not believe in coincidences

Front 103

Marr’s proposal-

Back 103

how do I get a computer to see? We see edges and illumenance edges very easily but hard for a computer. Filtered out a lot of the hard stuff, texture, shading etc. and took out all the high and low frequency and left all the edges and stuff. Computer could find edges
computational modeling
Mostly bottom up

Front 104

shape constancy

Back 104

we continue to see an objects shape even when the shape is distorted on our retinas (door opening). Keep a constant view of what that shape is even when it changes and moves around. Shape constancy make lines receding in depth look longer than lines going sideways

Front 105

illusory contours

Back 105

the visual system “fills in” lines that aren’t there (but are strongly implied by good continuation)

Front 106

Oculomotor Cues

Back 106

o Convergence, accomodation

Front 107

o Convergence:

Back 107

what you do when you cross your eyes. But we do it often all the time
• The closer an object is, the more the eyes point inward (towards our nose). Brain knows when we are doing this and we can tell how far something is.

Front 108

o Accommodation
•

Back 108

The closer an object is, the more your lenses contracts.
• Think lens-object is near
• Thin lens- object is far

Front 109

Monocular Cues

Back 109

o Pictorial Cues, • Motion cues, • Binocular Disparity

Front 110

o Pictorial Cues:
•

Back 110

look at a painting or photograph but it looks like its in depth

Front 111

• Occlusion-

Back 111

come stuff is in front of other stuff

Front 112

• Relative height-

Back 112

things look farther away if they are higher in the scene or closer to the horizon

Front 113

• Linear perspective:

Back 113

lines meet. Hallway.
• Parallel lines appear to converge to a point at the horizon.

Front 114

o Motion parallax

Back 114

As we move, near things “move” quickly
• As we move, far things “move” slowly

Front 115

o Deletion and accretion

Back 115

• Occlusion changes as things move. As things move, parts you can see from one angle you cant see anymore from another angle. Some stuff disappears while other stuff reappears.

Front 116

o Horopter-

Back 116

everything that will look the same distance from .fixation in both eyes. Out in visual field where everything is equal distance from your visual field in both eyes in your fovea. You don’t get a double image from objects on that plane. Everything is in focus that is your horopter.