Perceptual Organization LOs Flashcards
What is the basis for the Gestalth approach to perception
Groups of stimuli take on a pattern-like quality, the whole is different than the sum of its part
Our experience is holistic not a myriad of separate pieces
Perception is organized/dynamic process
Minimum principle: physical systems tend to settle into equilibria involving minimum energy or surfaces
Psychophysical isomorphism: correlation between psychological experiences and physiological events in the CNS
What are some Gestalt laws of Grouping
Law of proximity: things that are near to each other tend to be grouped together
Law of similarity: similar things tend to be grouped together
Law of good continuation: points that if connected would result in either straight or smoothly curving lines are seen as belonging together or lines tend to be seen in such a way as to follow the smoothest path
Law of closure: space enclosed by a contour tends to appear as a figure
Law of common fate: things moving in same direction tend to be grouped together
Law of meaningfulness (familiarity): things tend to form groups if items appear meaningful or familiar
Law of Pragnanz (good figure, simplicity) : every stimulus pattern tends to be seen in such a way that the resulting structure is as simple as possible
Common region: elements tend ot be grouped together if they are located within same closed region
Element connectedness: elements tend to be grouped together if they are connected by other elements
Describe Pros and Cons of gestalt approach
PROS: Identified important phenomena
CONS: often circular, poor definitions (say tend to), needs explanations not just post hoc descriptions, perception doesn’t always obey minimum principle
What are some artificial intelligence approaches to organization
Blocks World
SEE program
OBSCENE progam
Blocks World
analyzed lines in images produced by edges
Issues: real world not comprised of straight lines
Objects have texture
Based on general viewpoint assumption: small shifts in position of viewer do not affect configuration of line drawing, rules of possibility of accidental alignment of image features into an accidental junction
SEE program
junctions are places where lines in an image meet, including L,K,peak,fork,X,muti arrow and T junctions. T junction (3 concurrent lines, 2 of them collinear) denotes object segregation. Arrow junction (3 lines meeting at a point with one of angles greater than 180 degrees) denotes edges of same object.
Set of positive and negative cues to connectivity of objects. Positive (solid) links suggest that regions in question correspond to faces of same object. Negative (dotted) links suggest that the regions belong to diff objects
Algorithm based on link inhibition: two regions are parts of the same object triggered by positive cues form a junction, only in the absence of negative evidence from the junction at the other end
PRO: intuitively attractive
CON: limited (only junctions considered), incomplete (fails to find a possible interpretation for some objects) and attributes impossible interpretations to some objects
OBSCENE program
richer info about each junction is necessary, adding info makes the problem easier to solve. Categorized edges by line labeling: interior edges (convex (+), or concave (-), boundary or occluding edges (->) (to its right is the body for which the arrow line provides an edge or its left is space)
Each junction has a limited number of possible interpretations (constraints). Arrow junctions have three interpretations, Ls have 6, Ts have 4, Ys have 5.
PRO: could reject impossible objects
Edge consistency constraint: an edge must be given the same line label at both ends (eg. convex edge cannot become concave at another junction)
PRO: could handle cracks and scenes having shadows
CON: required a good deal of typing to explicitly code every possible edge and junction
How is full primal sketch obtained
Place tokens: neighboring components are assigned same location
Aggregation: adjacent place tokens are clustered/grouped according to texture or curvilinearly according to the orientation of elements
Theta aggregation: differs from intrinsic orientation of features.
Done for many diff levels of detail: representations created from small features to more global properties
Natural constraints applied: in general things that are adjacent to each other,
and/or similar to each other tend to belong together
Full primal sketch is a representation of whats on the retina (2D) not of the 3D world
How does Marr’s approach relate to the Gestalt approach
Structural Description Approach: instead of processing artefacts found in the proximal image describe the structure of the distal object
Blocks world is too oversimplified
Marr applied program to everyday object
Place tokens like gestalt law of proximity
Aggregation similar to similarity and good continuation
Theta aggregation similar to closure
