Perceiving and Recognizing Objects

1
Perceiving and Recognizing Objects
2
Perceiving and Recognizing Objects

• Finding edges
• Grouping and texture segmentation
• Figureground assignment
• Edges, parts, and wholes
• Object recognition
• Objects in the brain
• Development of object perception

3
What Do You See?
4
Object Ambiguity
5
More Ambiguity Accidental Views
6
More Ambiguity Accidental Views (contd)
7
Every Image Is Ambiguous
8
Pandemonium

• Oliver Selfridge (1959)
• Simple model of letter recognition
• Demons loosely represent neurons each level
  represents a different brain area
• Perception by committee
• Pandemonium Model Online

9
Finding Edges

• Computer algorithms dont do a very good job

10
Finding Edges (contd)
11
Creating Edges
12
Rules for Linking Contours

• Good continuation group elements to form
  smoothly continuing lines

13
Gestalt Psychology

• The whole is greater than the sum of its parts
• Wertheimer, Köhler, Koffka (1920s1950s) Palmer
  and Rock (1990s)
• Reaction to earlier structuralist school of
  psychology

14
Texture Segmentation

• Linked to 2 Gestalt grouping principles
• 1. Group elements that are similar together

15
Texture Segmentation (contd)

• 2. Group elements that are proximal (close to
  each other) together

16
Parallelism/Symmetry

• Somewhat weaker grouping principlesgroup
  parallel and symmetric elements together

17
Dynamic Grouping Principles

• Common fate group elements moving in the same
  direction together

18
Dynamic Grouping Principles (contd)
19
Dynamic Grouping Principles (contd)

• Synchrony group elements changing at the same
  time together

20
Dynamic Grouping Principles (contd)
21
Modern Gestaltism

• Common region Elements perceived to be part of a
  larger region group together
• Connectedness Elements that are connected to
  each other group together

22
FigureGround Segmentation

• What is the to-be-recognized object and what is
  the background?
• Gestalt figureground assignment principles
  surroundedness, size, symmetry, parallelism

23
Meaning and FigureGround Assignment

• Is brown or white the figure here?
• Object recognition starts before figureground
  assignment finishes

24
A Complicated Process...

• Which circles are figures and which are holes?

25
The Relatability Principle

• Q When should we complete edges behind
  occluders?
• A When the edges are relatable by an elbow
  curve (Kellman and Shipley, 1991)

26
The Relatability Principle (contd)
27
Meaning in the Edges

• Non-accidental features provide clues to object
  structure

28
Is the Whole Seen Before the Parts?

• Global superiority effect (Navon, 1977)

29
Parts and Wholes

• Pairs of concavities indicate part boundaries
  (Hoffman and Richards, 1984)
• What happens when cuts can go either way?

30
Parts and Wholes (contd)
31
High-Level Vision

• Processes in object recognition
• Determine features present in image (Low-level
  vision)
• Group features into objects (Middle vision)
• Match perceived to encoded representations
  (High-level vision)

32
What Is a Representation?

• A thing that stands for another thing, for
  example

33
Naive Template Theory

• Lock-and-key representations
• Problem You would need too many templates!

34
Structural Description Theory

• Represent the structure of an object, not what it
  looks like from one view
• Recognition-By-Components (RBC Biederman, 1987)

35
The Effect of Viewpoint

• RBC predicts viewpoint invariance
• Many empirical studies have found viewpoint
  dependence
• Viewpoint Effects Activity Online

36
Levels of Object Categorization

• What are these objects?
• Entry level birds
• Subordinate level sparrow/ostrich
• Superordinate level animals

37
The Entry Level

• Which label comes to mind first?
• The entry level term may be determined by a
  perceptual committee

38
Face Recognition

• Whats wrong with this picture?

39
Face Recognition (contd)
40
Face Recognition (contd)

• Special processes may be involved in
  identifying individual faces
• Prosopagnosia Selective inability to recognize
  faces

41
Objects in the Brain

• What system object identification (IT cortex)
• Where/how system object localization/
  manipulation (parietal cortex)

42
Grandmother Cells

• Could a single neuron be responsible for
  recognizing your grandmother?