PSYCH 169: Lecture 5, Veridical perception (visual system's representation of the external world), change blindness, inattentional blindness, theories of vision

visual system's representation of the world

-visual system generates an internal model (representation) of the world

-Two questions:

1. How accurately is the world represented?

2. How detailed is this representation?

Perceptual Illusions

illusion: something that it is not what it appears to be

-cafe wall illusion

-Hermann Grid Illusion

retinal ganglion cells

-have a center-surround receptive field

Is visual perception veridical?

-veridical perception: direct perception of stimuli as they exist

-visual world is not always accurately mapped (internally represented)

other evidence that visual perception is not veridical:

visual "filling in" of missing information

Blind spot

-where the optic nerve leaves the retina

-6 degree visual angle

-What do you see at the blind spot?

-not a blank space

-if background is boring gray, you see gray

-if background is blue/yellow stripes, you see blue/yellow stripes at blind spot

What if it's a complex scene: population of people/faces

-pebbles on the beach

-it is filled in with the appropriate stimulus (people/pebbles)

-note: blind spot is 15 degrees away from fovea

the brain paints in properties at this location (fovea)

active cognitive processes:

-completion

-interpretation

-filling in

How does filling in work?

-isomorphic

-symbolic

isomorphic filling

low level visual processes fill in the details

symbolic filling

filling in occurs higher up in the visual system and is more conceptual

Oliver Sacks

-tumor destroyed much of the retina of his right eye

-had a scotoma

-if he looked at the sky he would see a grey dot

scotoma

large blind spot

Oliver Sacks: filling in

-plain colors fill in quickly

-patterns on carpet in his office: 10 to 15 seconds

-size of pattern mattered

-2 feet from brick wall: scotoma filled with the brick's red color, but no details

-20 feet from brick wall: perfectly formed bricks

Oliver Sacks: foot

-his foot would look like a stump then one minute later becomes a phantom foot

V. S. Ramachandran (UCSD)

-blind spot experiments

-two vertical lines with a gap

-gap falls on blind spot

-subjects report continuous line

V. S. Ramachandran (UCSD): Bicycle wheel

-spokes are completed and no gap is perceived

V. S. Ramachandran (UCSD): Yellow doughnuts

-if blind spot falls on a hole, doughnut is perceived as a solid disk

-"pop out" effect

Josh (patient)

-right primary visual cortex (V1) injured by steel rod in industrial accident

-Large scotoma in left visual field

-position scotoma on a solid vertical line

-First, I see a gap

-After a few seconds, gap closes and I see a continuous line

X experiment

-filling in did not work for column of large Xs

-worked for column of small x

-Different levels of processing:

-large "X"s activated object recognition area (ventral/temporal)

-column of small "x" activated lower level texture coding

fMRI studies on filling in

studies show that early visual areas (V1, V2) are not involved in "filling in"

-areas V3 and V4 are involved

change blindness

-a demonstration of the sparseness of visual representation

eye saccades

cause massive blue

-trans-saccadic memory is very poor

Change blindness: early vision theories:

-successive pictures (from multiple saccades) must be integrated into one big representation

-requires massive computations

-Claim: we have a detailed view of the world

-change blindness challenged this assumption

Change blindness: we do not have a detailed view of the world

-massive integration of successive views is not necessary

-we overestimate our ability to detect change

Inattentional blindness

-if you are not directly paying attention to something, will you see it?

-easy to miss something obvious (does not enter your consciousness)

-the most important factor affecting inattentional blindness is a person's own attentional goals

-hard to see something that is truly unanticipated

Inattentional blindness: Mack & Rock (1988)

-circles and cross

-which is longer, horizontal or vertical?

implications for theories of vision

-vision is not a process of building up detailed inner representation that allow comparison from one moment to next

-we do not hold on to nearly as much information as we seem to think

Simons and Levine (1997)

-initial rich visual experience

-extract meaning ("gist")

-if eyes move (saccade) we get a new experience

-if "gist" is the same, you assume details are the same

-will not notice change

a second theory, Resink (2000)

-you never form a complete representation of the world

-not even during fixations

-object representations are build one at a time

-focused attention takes a few basic objects from low-level processing to a "coherence field"

-objects that make it there will persist for a few seconds

-when attention is released, the object loses coherence and "dissolves"

a third theory, Kevin O'Regan (1992)

-no need to store large amounts of information

-the outside world acts as an external memory storage

-no internal models

-visual world is not something we build up, it's something we do

-sensorimotor theory of vision

-what you see are those aspects of the scene that you are currently "visually manipulating"

-counter intuitive because seeing does not feel like manipulating

-what about "pop out" effects