Visual Perception Flashcards
Why is the eye not a camera?
One image can lead to many percepts
- small, medium, or large item
- stationary object vs. moving object
- if a camera, the image couldn’t lead to different percepts.
What is the explanatory gap?
How we can account for subjective awareness based on brain activity. How do you get from sensation of light on the retina to a phenomenal awareness of something?
Examples of multiple percepts from one retinal image:
- Gibson Cylinders: look the same size when different physical sizes, but look like they are different sizes when they are the same physical size.
- Ambiguous figures: old woman/young woman
- Figure dround illusions: face/vase illusion
- Moon illusion: Moon in the sky looks smaller than the moon on the horizon because there’s nothing of real size to compare it to.
What is the inverse problem?
How do we explain the transformation of sensory input into perception? Specifically, how do we go from a 3D object in nature to a 2D retinal image back to a 3D percept?
All distal stimuli can be associated with the same retinal image, so how do we know which distal stimulus responds to the correct stimulus? How do you get an unambiguous percept from ambiguous stimuli?
proximal stimulus
the retinal image
distal stimulus
object in the environment that emits light onto the eye, the image of which is cast onto the retina
percept
phenomenal experience derived from perception/interpretation of the proximal stimulus plus other non-retinal information
Nativist view of perception
everything is innate, and we are born with an enormous amount of visual capacity
empiricist view of visual perception
you have to learn everything and nothing is innate. (This can’t be true or how would you know you need to learn to see?)
Structuralist/Atomist view of visual perception
perception can be reduced to atomic units that make up the basic units of perception (like an alphabet, called “perceptual primitives”)
Holist view of perception
Perception is based on interrelationships between elements and images on the retina
Constructionist view of perception
(Arien’s view) perception is constructed by the visual system using both top-down and bottom-up processes
What is a stimulus error?
Including what we know from experience as basic units
Structuralists (Wundt, Titchener) believed perceptions were based on sensations, called qualia. We can’t find the basic elements of perception by looking at things because we have to peel away all the associations until we get to raw, pure sensations: quality, intensity, and feeling-tone.
What is an experience error?
The mistake of assuming we have direct access to the distal stimulus. Because we do not have access to the distal stimulus, the image on the retina supports an infinite number of possible percepts.
Who was concerned about stimulus error?
Gestaltists: Koffka, Kohler, Wertheimer
They believed the whole is based on elements of the whole AND the relationship between the elements that make up the shape. The shape is the emergent property. You can change the key of a song, but as long as the relationship between the notes remains the same, the whole is preserved.
Pragnanz
simplicity. Perceptual processes are an attempt to seek the simplest possible percept given the input.
The rules of grouping preserve pragnanz (Gestaltists).
emergent property of the whole
As an M and a W merge, a diamond pattern emerges and is perceived, tather than the two individual letters. Follows gestalt principles/experience error/pragnanz.
The whole is different from the sum of the parts
Direct theory of perception
Gibson
Our perceptions are explicable in terms of the array on the retina. We don’t need to disambiguate the retinal image, memory doesn’t play a role in perception. Everything you need to form an accurate percept is present on the retina.
affordances
Follows direct theory of perception, all “action possibilities” latent in the environment, objectively measurable and independent of the individual’s ability to recognize them, but always in relation to agents and therefore dependent on their capabilities. For instance, a set of steps which rises four feet high does not afford the act of climbing if the actor is a crawling infant. Gibson’s is the prevalent definition in cognitive psychology.
Indirect theory of perception
Helmholtz, Rock, Arien
Percepts are constructed, not abstracted.
- unconscious processing makes a difference in perception
- there is a disjunction between physical aspects of the world and our experience of them
Grouping
Proposed by gestaltists as a method of perceptual organization, perhaps an attempt to partially solve the inverse problem. Thought to be parallel and automatic.
- Similarity
- Proximity
- Closure/continuity
- Good continuation
- Common Fate
- Symmetry/parallelism
similarity grouping
(like test for color blindness, based on similarity of color grouping. Also rows vs columns, etc.) features that look similar are associated.
Similarity of color, size, or orientation
proximity grouping
If they’re close together, they probably belong together
Closure grouping
If lines are far apart, they will seem like they don’t belong together, but if you make them appear to close together, they will appear grouped (closed rather than open figures). Stronger than proximity.
Good continuation grouping
contours based on smooth continuity are preferred to abrupt changes of direction
Common Fate Grouping
The grouping together of objects that have the same trend of motion and are therefore on the same path. For example, if there are an array of dots and half the dots are moving upward while the other half are moving downward, we would perceive the upward moving dots and the downward moving dots as two distinct units.
veridical
true to the distal stimulus, non-illusory percept
Parallel processing
gestalt view, occurring on the retinal all at the same time, pop-out
serial processing
must search among similar objects (like Ts and Ls) looking one by one
Palmer Grouping Rules
- Common Region
- Element Connectedness
- Synchrony
Common Region
Palmer grouping, elements within a single region are grouped together
Element Connectedness
Grouping rule: When all else is equal, images connected by an added element (like a line) are grouped together. Stronger than proximity.
Region Segmentation
Areas of UC of luminance, texture color, motion, etc., enable this mapping. Maximizes similarities within regions and differences between them.
Palmer Perceptual Flow Chart
Modal/Amodal completion
Amodal completion: Occluded surfaces are completed, full perception of an object even though it is only partially perceived
Illusory contours added (seeing contours when none exist in distal stimulus)
Modal completion: one shape is perceived to be occluding other shapes, even when the shape itself is not drawn (like Kanizsa triangle)
Figural scission
- modal completion
- no contours, but completion of both occluded and occluding figure
- single homogenous region splits into two perceived regions
- stimulus is ambiguous, so either figure can be seen in front
Types of scission
- visual completion
- illusory contours
- figural scission
- transparency
Orientation transformances that do not affect shape equivalence
(orientation can, if there is no canonical top or bottom)
Frames of reference hypothesis:
Object and subject are relative. Shape can be defined relative to an object, like when a rectangle is around either a square or a diamond, we can change our perception of whether it is a square or a diamond, though ambiguous.
Organization of shapes
- how do we know what objects are for?
- Affordances (Gibson, direct): opportunities for action or interaction inherent in objects (knee-high, flat surface, must be for sitting, so it’s a chair)
- Categorization (Palmer, indirect): function follows a category. Identify as a chair, then identify function (for sitting)
