Exam 1 Flashcards
Give a one sentence summary of Gestalt psychology
“The whole is greater than the sum of its parts.”
What does “Gestalt” mean?
In German, “form” or “whole”
Major way structuralism differs from Gestalt psychology
structuralism emphasizes the basic elements of perception = reductionist model
What are Gestalt grouping rules? What do they explain?
- A set of rules describing which elements in an image will appear to group together.
- They explain what it is that causes the brain to see something as one whole.
Gestalt ideas that fall under rules for contours
good continuation & occlusion
Define good continuation
two elements will tend to group together if they lie in a similar orientation, or along a similar plane (on the same contour)
Explain good continuation in simpler terms
When something looks like a continuing line, we see it as a whole. The further it gets from being a straight line, the less likely that it’s part of a whole.
Book rewording of good continuation
the tendency of lines of similar orientation to be seen as part of the same contour. Such lines “support” each other = two visual bits of an edge will make it easier to perceive a third colinear segment that lies between them, even if its not visible.
What happens if visible bits/lines form a closed shape? Give ex
The little segments support eeach other even more strongly. Ex/ multiple lines in the shape of a circle
Explain the question behind occlusion.
When we see an unexpected edge pop up in an image, our visual system tries to figure out why
Explain occlusion
When we see an unexpected edge in an image, our visual system assumes it is being occluded by another object, causing us to “create” an image where none may actually exist.
Define texture segmentation
carving an image into regions of common texture properties
Give ex of texture segmentation
in green an blue picture on page 84, the portion of the image with a coarser texture is separated from the rest of the image
List gestalt grouping principles
similarity, proximity, parallelism, symmetry, common region, connectedness
Explain similarity simple way
figures that are similar go together
explain similarity book way
image chunks that are familiar group together
features that similarity can be based off of
color, size, orientaton, form
What doesn’t work well with similarity? Give ex
Combinations of features (color, size, orientation, form)
ex/ trying to make distinctions based on color and shape - left side is orange diamonds/green squares and right side is green diamonds/orange squares
Explain proximity
Items near each other (close in space) are likely to group together
Try to figure out proximity in example 4.10 (a) on p. 84
figure out
Explain parallelism
Parallel contours are likely to belong to the same figure
Explain symmetry
Symmetrical regions are more likely to be seen as a figue
look at examples of occlusion
p. 82 & 83
Explain the principle of common region
Two features will go together if they appear to be part of the same larger region
Explain the principle of common region in simpler terms
Even if??
If there’s some other contour putting two features in the same space, they go together. Common region = enclosed in something
Even if they’re not proximal
Explain connectedness
two items will tend to group together if they’re connected
Explain middle vision
A stage of visual processing that comes AFTER basic features have been extracted from the image and BEFORE object recognition and scene understanding (before you’ve figured out what it is)
middle vision is when what starts. Explain
when the object recognition process starts - we’re unaware that this processing time is happening.
middle vision determines what?
which regions of an image should be grouped together to form objects
a big part of middle vision/detecting objects involves doing what?
finding edges
One of the first things we need to do in order to detect objects is what? Explain what these are
find contours = the outlines that tell us where the edge of an object is
Explain the arrow picture (p. 81) and how it relates to our detection of contours
In some places, the object is darker than the background. In others, it’s lighter. If the changes are continuous, there must be places where the edge of this shape disappears. The occasional lack of an edge doesn’t bother our visual system.
Are computers good at finding edges? explain
Not as good as our human system. Sometime computers leave out edges or detect way too many (p.81)
define illusory contour
a contour that is perceived even though nothing changes from one side of the contour to the other in the image
Explain the gestalt rule of an illusory contour
p. 82 example
Visual system detects an edge even tho there’s none there - visual system superimposes an edge.
Give 2 committee rules for detecting objects
1) honor physics
2) avoid accidents
define ambiguous figure. Example?
one that generates two or more plausible interpretations. Example = necker cube
Ambiguous figures are _______ to the rule. Explain
exceptions - in theory, every image is ambiguous, but the perceptual committees almost always agree on a SINGLE interpretation
Explain the accidental viewpoint
a viewing position that produces some regularity in the visual image that is not present in the real world
the perceptual committees do not ____ on accidental viewpoints. Explain
Bet. Figure 4.7 on p. 81 = very unlikely that that’s 4 worms meeting
Explain the honoring physics aspect of perceptual committees. Give example
figure with arrow - we infer the arrow-shaped object because of our IMPLICIT understanding that solid objects block light. Not rules we have to verbalize - we just know
Define figure-ground assignment
The process of determining that some regions of an object belong to a foreground object (figure) and other regions are part of the background (ground)
List the principles that work in figure-ground assignment
surroundedness, size, symmetry, parallelism, extremal edges, relative motion
Explain surroundedness in F/G assignment
if one region is entirely surrounded by another, then the surrounded region is the figure.
Figure 4.18 on page 88 - Explain one possible explanation
surroundedness explains why the green region is seen as the figure
Explain size in F/G assignment. Give example
the smaller region is likely to be the figure.
Ex/ the cow is smaller than field in which she stands, letters smaller than the page
Explain symmetry in F/G assignment
a symmetrical region is more likely to be seen as figure
Explain parallelism in F/G assignment
regions with parallel contours tend to be seen as figure
Figure 4.20 on page 89 - Explain one possible explanation
parallelism causes the purple regions to appear figure
Explain extremal edges from the powerpoint
If edges of an object are shaded such that they seem to recede in the distance, they tend to be seen as figure
Explain figure 4.21 upper left figure
the edge of the white disk has shading suggesting that the edge is curving away from the viewer and toward the gray texture = white disk must be closer so it’s figure
Explain figure 4.21 upper right figure
white donut shape has extremal edge cue, so it becomes the figure
Explain relative motion in F/G assignment
If one region moves in front of another, then the closer region is figure
define relatability
the degree to which two line segments appear to be part of the same contour
Explain relatability in detail
lines that can be related by a simple curve (like an elbow or a bend in the road) are more likely to be seen as relatable than something that requires something complex like an S curve
look at relatabiility example
p. 90
define heuristic
mental shortcut
Wordy explanation of global superiority effect example on p. 90 (Fig 4.25)
the big letters interfere with the naming of the little letters more than the little letters interfere with recognition of the big letters
define global superiority effect
the properties of the whole object take precedence over the properties of parts of the object
implicit assumption that the global superiority effect follows
the first goal is to carve the retinal image of large-scale objects
5 goals of middle vision
- bring together what should be brought together
- split asunder that which should be split asunder
- use what you know
- avoid accidents
- seek consensus and avoid ambiguity
