3. Object Recognition

Question 1

Why is object recognition difficult (2 reasons)

Answer 1

• objects overlap but form coherent perceptual experience

- retinal image varies but size and shape of object doesn’t

Question 2

Challenges of object recognition: variability on the retinal image

Answer 2

translation, rotation, size, colour

Question 3

Challenges of object recognition: Variability of a visual scene

Answer 3

occlusion and presence of other objects

Question 4

challenges of object recognition: intra-class variation

Answer 4

an object is still identified as such despite it being a different shape, colour, design (features). Eg diff types of chair. Can also be recognised when partially hidden.

Question 5

Challenges of object recognition: viewpoint variation

Answer 5

can identify objects regardless of what view were seeing them from

Question 6

2D pattern matching theory: template theories

Answer 6

• template in LTM of all known patterns. Can have multiple templates and compare for greatest overlap
• eg barcodes and fingerprints

Question 7

Issues with template theory

Answer 7

imperfect matches and not accounting for flexibility of pattern recognition system (need to be identical orientation, size, position etc)

Question 8

2D pattern matching: prototype theories

Answer 8

• more flexible templates
• process the average of characteristics
• no perfect match

Question 9

Evidence for prototype theory pattern matching

Answer 9

Franks and Bransford 1971: presented object based on a prototype, Ps were confident they’d seen the prototype before seeing the images that made the average prototype

Question 10

2D pattern matching: feature theories: what is it and a limitation

Answer 10

• pattern has set of features/attributes
• must know the relationship between the features
• too many objects for this to be the only mechanisms

Question 11

2D pattern matching: structural descriptions

Answer 11

• a breakdown of the pattern and relationship between it’s features

Question 12

3D object recognition

Answer 12

comprised of early image processing: objects separated from background. segregated from each other and then matched to object description in memory

Question 13

Marr and Nishihara 1978: 3D object recognition theory

Answer 13

• the relationship between objects made up of cylinders is it’s structural description
• not possible for all objects, good for biological things

Question 14

3D object recognition: Biederman’s recognition by components theory

Answer 14

• objects composed of geons (geometrical ions)
• 36 diff volumetric shapes
• concave parts of object can help segment image into separate geons

Question 15

This theory is viewpoint invariant

Answer 15

Recognition by Components theory (Biedermans geons)

Marr and nishiharas cylinders

Question 16

Properties to decide what geons make up an image (5)

Answer 16

Curvature, parallel, co-termination, symmetry, collinearity

Question 17

Evidence for recognition by components theory: concavity (Biederman 1987)

Answer 17

taking away the points of concavity made it harder to recognise as the geons couldn’t be identified

Question 18

Evidence for recognition by components theory: cortical neurons in monkeys (vogels et al 2001)

Answer 18

some neurons in inferior temporal cortex responded more to changes to geons than object size changes

Question 19

Disadvantages/evaluation of recognition by components theory

Answer 19

• Edge information/structure not always key: texture and colour also important
• ignores viewpoint dependence: easier to recognise some objects from some views
• ignores within category discrimination (all chairs dont look the same)

Question 20

Viewpoint invariant theories are better for easier between-category discrimination. Viewpoint dependent theories are better for…

Answer 20

harder between-category discrimination and within category discrimination

Question 21

Define viewpoint dependence:

Answer 21

Some objects are faster and more accurately defined from certain viewpoints

Question 22

The binding problem asks:

Answer 22

how do we integrate different kinds of information when identifying objects?

Question 23

What are the stages of object recognition according to Humphreys et al 1988

Answer 23

• separation of object from background/other objects
• structural description
• semantic representation: ie how is it used
• name representation

Question 24

Agnosics struggle with which part of object recognition?

Answer 24

structural description

Question 25

Which type of agnosia provides support for Humphreys 1988 object recognition stages?

Answer 25

Associative agnosia: know the object but cannot identify or use it

Question 26

The Cascade Model of object recognition (Humphreys et al 1968)

Answer 26

• the structural, semantic and name stages all interact
• there are within and between category stages (info flow)
• a problem at one stage has knock on effects on other stages, but dont get stuck on stage

Question 27

This test is often used to diagnose agnosia

Answer 27

Birmingham object recognition battery test (BORB)

Question 28

Apperceptive agnosia

Answer 28

• cant recognise, draw, or copied in the visual modality
• often affects unusual viewpoints of objects
• patients may have issues drawing an object which has missing contours

Question 29

Associative agnosia

Answer 29

• can draw, copy, and know what they are seeing but cannot recall the name or details of the object or feel familiarity
• can usually describe an object but cannot name it
• visual representations intact

Question 30

This type of agnosia affects later processing/recognition

Answer 30

Associative agnosia

Question 31

This type of agnosia affects earlier processing/shape extraction

Answer 31

Apperceptive agnosia

Question 32

A type of agnosia in which animate objects harder to identify than inanimate objects

Answer 32

category-specific agnosia

Question 33

Different knowledge domains about the object that may be affected in agnosia

Answer 33

• its functional significance

- sensory properties

Question 34

These types of agnosia can occur together or separately

Answer 34

agnosia - objects

prosopagnosia - faces

Question 35

involves between category discrimination

Answer 35

Object recgonition

Question 36

involves within category discrimination

Answer 36

Face recognition