Object Perception Theories

Question 1

Limitations of Gestalt Theories

Answer 1

• qualitative/descriptive: not based on normative systems, provides descriptions of how the perceptual system functions, but does not say WHY or HOW
• is not quantitative

Question 2

Theory of Relatability

Answer 2

more quanitative

Question 3

Theory of Non-accidental features

Answer 3

junctions

Question 4

Theory of Geons

Answer 4

???

Question 5

The visual system connects spatially separated visible

areas using what two processes:

Answer 5

contour interpolation

surface interpolation

Question 6

Contour interpolation

Answer 6

if there is a missing edge, the brain must fill it in

Question 7

Surface interpolation

Answer 7

if there is a part of an object missing due to occlusion, the brain must fill in the surface to perceive a complete object

Question 8

What are the steps to the contour interpolation process?

Answer 8

1) Begins with the locating of contour junctions
2) Interpolated edges begin and end at these junctions
3) Contour interpolation follows a smoothness constraint, known as contour relatability

Question 9

Theory of relatability

Answer 9

What extent of curviness or angle is tolerated by the visual system in the process of interpolation

(same object vs different objects)

Too large an angle –> un-relatable
Acceptable angle –> relatable

NOTE: relatability is related to the gestalt idea of “good
continuation”

Question 10

What are the steps to surface interpolation?

Answer 10

1) The contour interpolation process depends on oriented edges leading into junctions
2) The surface process complements contour interpolation
3) Surface properties “spread” under occlusion within real and interpolated boundaries
4) This process depends crucially on matches of color, lightness, and texture

Question 11

Whether we form a unit or not depends on…

Answer 11

surface properties AND contour properties

*example: color differences violates relatability, sharp angle violates relatability

Question 12

Shape perception is invariant to…

Answer 12

Scale invariance (size)
Orientation invariance

Question 13

Template representation of shapes

Answer 13

*any pattern that would fit the template ‘x’
would be recognized as ‘x’.
* template is specific to orientation in
plane, orientation in depth, size, location,
etc.

Question 14

Problem with template representation

Answer 14

you would need billions of templates for each object, because it would always require the same activation of the same neurons for each template (x-invariant)

example: could not recognize the letter A in different positions, scales, orientations, viewpoints, etc

Question 15

Volumetric/Structural representation of shapes

Answer 15

• ideal way of representing objects, because it is 3D and thus takes care of the invariances

Question 16

Examples of volumetric theories

Answer 16

Polyhedra
Superquadratics
Generalized cylinders
Geons

Question 17

Recognition by Components (RBC) is better known as

Answer 17

Geon Theory

Question 18

Geon Theory

Answer 18

the visual system represent objects by their geometric icons (geons) and the spatial relationships between them, metric changes are not important (width/height, etc)

• strengths: provides invariance, by combining geons you could create thousands of shapes
• visual system finds non-accidental features, then extracts structural (or geon) description

Question 19

How many geons are there in the brain?

Answer 19

About 36

Question 20

What is another word for a geon?

Answer 20

volumetric primitive

Question 21

Evidence for geon theory

Answer 21

If the non-accidental features (e.g., corners) are available, geons can be recovered and the object can be recognized.

If the same amount of contour is available but the non-accidental features are missing, the geons cannot be recovered and object recognition will be more difficult.

Question 22

Invariance in geon theory

Answer 22

• Scale (size) invariant
• Position invariant
• Rotation (view-point) invariant

Question 23

Problems with geon theory

Answer 23

• not good for describing natural/non-manmade objects
• not good for within-category recognition
• ## human representation is view-based, and not invariant

Question 24

What is another more simple adaptation of a structural or geon theory?

Answer 24

concavities

Question 25

Examples of View-point Object Representation Theories

Answer 25

• Templates
• Graphs
• Feature sets

Question 26

Elastic Graph Representation

Answer 26

each node represents a local feature (hypercolumn in v1), edges are connections between the hypercolumns (synaptic)
local features combined with spatial relationships between them considered a graph

Question 27

Invariance in Elastic Graph

Answer 27

• position invariant
• somewhat orientation invariant
• somewhat scale invariant
• viewpoint robust

Question 28

Feature-set Representation

Answer 28

• selective neurons are filters sensitive to specific features
• the retinal image gets processed in parallel
• if the right set of signals gets activated, signals an object classifier and we recognize it in different views

Question 29

Strengths and weaknesses of feature-set representation

Answer 29

strength: very fast recognition
weakness: ignores spatial relationships, so system could be easily fooled to falsely recognize something