6. Depth Perception

Question 1

Describe the Inverse Problem

Answer 1

• Any retinal image is consistent (the same) with infinitely many possible configurations of the world
• The visual system works out which is the correct configuration of the world

Question 2

What are the Multiple sources of 3D information?

Answer 2

○ Binocular
○ Motion
○ Pictorial
○ Oculomotor

Question 3

Why is Depth Perception Important?

Answer 3

3D perception is vital for interacting with the world and recognising objects

Question 4

The prior assumption that lines in the world tend to be parallel or perpendicular to each other is used to disambiguate which depth cue?

Answer 4

Perspective

Question 5

In what situation are we most likely to experience an illusion?

Answer 5

When our assumptions are invalid

Question 6

When two cues are integrated by compromise the final percept will be what?

Answer 6

When two cues are integrated by compromise the final percept will be Biased towards the most reliable cue

Question 7

Describe binocular cues.

Answer 7

○ It arises because we have 2 eyes that are in different positions in our heads. So they both have a slightly different view of the world = the retinal images will be different.

○ The visual system can match up the left and right eyes images and find the differences in the images (binocular disparities)
It uses these to work out the 3 dimension of the world must look like.

Question 8

What is the most important Binocular Cue?

Answer 8

Binocular Disparity (Stero Vision)

Question 9

What are the 2 motion cues?

Answer 9

1. Motion parallax - We move around

2. Kinetic Depth - Objects move around

Question 10

Describe Motion Parallax.

Answer 10

• Cause motion on the retina
• Things that are close to us move more and things further away will move less.
• The visual system can use that different elements of an image are moving at different speeds to work out how fast they are

Question 11

Describe Kinetic Depth (Kinetic Depth Effect/ KDE).

Answer 11

• When objects move it creates motion in the retinal image.

- Because of the 3D shape of an object different parts of that object will appear to move at different speeds.

Question 12

What do Pictorial cues tell us?

Answer 12

• Can tell us how close or far away something is.
• Perspective cue:
  ○ Lines converge as things move further away from us.

Question 13

What are all the 6 Pictorial Cues?

Answer 13

1. Texture
2. Elevation
3. Relative Size
4. Perspective
5. Shading
6. Occlusion

Question 14

Describe Oculomotor Cues.

Answer 14

• Information that is not present in the retinal image. It is information we get due to the position of the eyes.

Question 15

Describe the 2 types of the Oculomotor Cues.

Answer 15

1. Convergence
   ○ Eyes converge when looking at a closer object
2. Accommodation
   ○ The lens changes shape to focus an image of an object onto the retinal
   § Thinner retina = far away object

Question 16

If we have so many different cues to depth what is the problem?

Answer 16

1. Many cues are ambiguous – 2D image compatible with infinite 3D worlds
2. With multiple cues available – how do we perceive a single unified world?

Question 17

Which type of depth cues are thought to be most ambiguous.

Answer 17

Pictorial Cues.

Question 18

What depth cues can create ambiguity?

Answer 18

1. Perspective
2. Shading
3. Texture
4. Elevation

Question 19

How does the visual system overcome ambiguity?

Answer 19

• We can overcome this ambiguity by using prior knowledge, or prior assumptions, to interpret the image.
• These assumptions are gained through our unconscious knowledge and experience of the physical properties of the world.
• This is a type of top-down processing and supports the constructivist approach to vision

Question 20

What assumptions do we use about perspective?

Answer 20

• make the assumption that lines in the world tend to be parallel.
• We assume that since lines are parallel and the lines in the photo converges then they must be moving further away. It tells us about depths in a scene.

Question 21

What assumptions do we use about shading?

Answer 21

use an assumption that light comes from above.

Question 22

What assumptions do we use about texture?

Answer 22

• we assume surface textures are isotropic (unbiased orientation) and homogeneous (uniform density)
• Making this assumption means any changes in image texture orientation or density are attributed to changes in 3D surface orientation

Question 23

What assumptions do we use about Elevation?

Answer 23

we assume that objects rest on a ground plane (surface parallel to the ground)

Question 24

What is the problem with Depth Perception Assumptions?

Answer 24

• Assumptions can lead to errors
• Assumptions will be valid in most situations, but in some cases they will not be valid, which will lead to perceptual errors (illusions)

Question 25

What does the Ames Room Illusion show?

Answer 25

- When we view the Ames room we assume lines are parallel or at right-angles, however this is not the case. - The perspective assumption of parallel lines is invalid which causes an erroneous (incorrect) percept. - Only works with one eye so there isn't binocular disparity information.

Question 26

Describe the types of multi-cue integration.

Answer 26

1. Compromise – take average, but take into account how reliable each cue is. 2. Dominance – only trust one and ignore the other. 3. Interaction – information incomplete or ambiguous, need one source to disambiguate the other

Question 27

Why is Multi-Cue Perception important?

Answer 27

- Real-world scenes have multiple cues present - Cues must be integrated to achieve a single unified percept - Integration helps to overcome problems of: ○ Reliability ○ Ambiguity ○ Conflict

Question 28

What happens when two sources of depth information are conflicting?

Answer 28

-The brain will try to find a compromise between the two. ○ However the brain will take into account how reliable each source is. - The perceived shape is biased towards motion cue

Question 29

What did Young et al (1993) study show?

Answer 29

- Looks at how Ps combine two sources of 3D information. - Participants viewed computer generated cylinder defined by 2 cues: ○ Texture & Motion

Question 30

what is dominance?

Answer 30

- When two cues define very different shapes or depths, the brain may choose to ignore one in preference for the other. - e.g. Elevations vs Relative size

Question 31

what happens in Richard Gregory’s Ames window illusion?

Answer 31

- Motion (KDE) in conflict with perspective, again perspective wins.

Question 32

How does interaction help perception?

Answer 32

- Some cues are ambiguous, but other cues can disambiguate them ○ e.g. binocular disparity can disambiguate ambiguous texture cues.

Question 33

what did Adams and Mamassian (2004) do?

Answer 33

- introduced conflict between the texture and binocular disparity cues. - Texture disambiguated by binocular disparity (interaction) AND - Then 2 cues averaged (compromise)

Question 34

what did Adams and Mamassian (2004) conclude about the order of integration?

Answer 34

- 2 stages 1. interaction stage - where cues disambiguate each other 2. compromise stage - where you average together disambiguated cues.

Question 35

What problems can cue integration overcome?

Answer 35

○ Unifying our percept of the world ○ Differences in cue reliability (compromise) ○ Conflicting information (compromise & dominance) ○ Cue ambiguities (interaction)

Question 36

What problems arise when there is conflict between binocular and oculomotor cue?

Answer 36

headaches eye-strain nausea

Question 37

What do cues tell us when we are looking at art & photography?

Answer 37

binocular, motion and oculomotor cues will all tell you these are 2D surfaces and yet we perceive vivid three-dimensionality due to pictorial information

Question 38

What do cues tell us when we are looking at Movies?

Answer 38

Pictorial and motion cues specify a 3D scene, but binocular cues and oculomotor cues define a flat surface

Question 39

What do cues tell us when we are looking at 3D Movies?

Answer 39

pictorial, motion and binocular cues all specify 3D scene, only oculomotor cues specify flat surface. - problems can arise.