Attention & Scene Perception Flashcards
(43 cards)
1
Q
What is attention?
A
- Ability to preferentially process some parts of a stimulus (more clear than others)
- At the expense of not processing other parts of the stimulus
2
Q
Why is attention needed?
A
-
Perceptual system has a limited capacity
- Can’t process everything in visual scene simultaneously
- Helps in not being overwhelmed
3
Q
What is the difference between overt and covert attention?
A
- Overt → Looking directly at an object
- Covert → Looking at one object but attending to another
4
Q
Describe two scenarios where covert attention would be applied
A
-
Calming down a child
- Not looking in order to pretend you’re not paying attention
- However still making sure child is okay
-
Basketball passing
- Not looking in direction of teammate whom you are passing the ball to in order to not signal to opponent
- But paying attention so ball is successfully passed
5
Q
What are saccades?
A
- Ballistic eye movements between fixations
-
Jump from point to point instead of smoothly
- Generally always looking at object that you are attending to → Can tell where someone is attending by tracking eye movements
6
Q
What are fixations?
A
- Rests inbetween visual jumps (saccades)
- Eyes stay looking directly at one part of the scene
7
Q
What directs attention?
A
- Initial involuntary
- Attentional capture → First presented, fixations captured by salient parts of scene
- Subsequent voluntary
- Directing to goals
8
Q
Name 4 types of contrast
Attentional Capture (Salience)
A
- Colour/ luminance
- Size
- Orientation
- Motion/ flicker
9
Q
What are fixations determined by?
A
- Cognitive Factors
- Goals
-
Expectations
- If object is unexpected → Fixate on it for longer + More often
10
Q
Distinguish between semantically and syntactically inconsistent objects
Expectations
A
-
Semantically inconsistent
- Wrong object
-
Syntactically inconsistent
- Right object but doing something ‘weird’
11
Q
What are the effects of attention?
A
-
Change speed of response
- Attention on a region of space = Faster
- Influence appearance
- Same contrast
- Where attention is = Contrast seems higher
- Influence physiological responding
- Neurons in brain respond more strongly to attended stimuli
12
Q
What is the binding problem?
A
- Issue of how an object’s individual features are combined to create a coherent percept
-
Different aspects of a stimulus are processed independently (separate brain areas)
- Dorsal stream = Motion
- Ventral stream = Form and colour
-
Different aspects of a stimulus are processed independently (separate brain areas)
- More difficult with multiple objects
13
Q
What is the Feature Integration Theory (FIT)?
A
- Suggests that binding problem is solved by attending to only one location at a time
-
Only features associated with location are processed → Only they are bound together
- Avoids binding features from different objects
14
Q
What are illusory conjunctions
A
- If attention is inhibited, features from different objects will be incorrectly bound together
- Prediction of FIT
- Experiment example
- Coloured letters and numbers flashed
- Followed by noise mask to remove after image
15
Q
Describe the case of RM
A
-
Parietal lobe damage → Balint’s syndrome
- Multiple objects present = Difficulty focusing attention on a single object
-
Prone to experiencing illusory conjunctions because he cannot focus attention on a single object
- Shown two letters with different colours
- Reported the wrong letter colour combinations on 23% of the trials → Even when allowed to view for 10s
16
Q
What is a conjunction search?
A
- Target different from distractors only by its particular conjunction of features
- Specific combination
- Attention needs to be applied to each object in turn (FIT) to find target
- Therefore are slow
17
Q
What is a feature search?
A
- Without solving binding problem
- Target has feature that distractor doesn’t have
- Because binding isn’t required, attention does not need to be applied to each item in turn (FIT)
- Searches predicted to be fast
18
Q
What is change blindness?
A
- Things not attended to are forgotten
-
Change needs to be missed when attention is not drawn to the location of the change
- But change must then be obvious when attention is drawn to the change
- Can’t make small changes
- Need to be large but also missed
19
Q
Why doesn’t change blindness occur all the time?
A
-
Changes usually generate motion transients that draw attention to the location change
- Easy to spot change
- If there is a blank screen right after first image, motion transients are for the entire scene
- Harder to spot change
20
Q
What is the problem of object and scene perception?
A
- Perception seems effortless but is much harder
- Trying to get a computer to perceive → Is worse than humans
21
Q
Why is there concern behind computers misclassifying stimuli?
A
- Use of automatised vehicles
- Risks safety of drivers
- If they haven’t been trained on images associated with accidents (have stimuli in different state) = Flaws arise
22
Q
Why is perception difficult?
A
- Stimulus on retina is ambiguous
- Objects can be hidden or blurred
- Objects look different from different viewpoints and in different poses
23
Q
Describe stimuli being ambiguous on the retina
Difficulties of Perception
A
-
All lines form the same retinal image
- 1D retinal image → Ambiguous
-
2D → Also ambiguous because multiple stimuli can give rise to the same 2D image
- Pole A shorter but closer, Pole B longer but further away - Yet form same retinal image
24
Q
Describe objects being partically hidden or blurred
Difficulties of Perception
A
- Objects can be covered by things like shadows
- Machine would probably have difficulty detecting object
25
Describe objects looking different
## Footnote
Difficulties of Perception
* *Machines* find it **hard to recognise objects** when they...
* Appear in **unexpected poses**
* Are viewed from **unexpected angles**
26
What is structuralism?
* *Distinguishes between sensations and perceptions*
* **Sensations** → **Elementary processes** occur **in response to stimulation**
* **Perceptions** → **Conscious awareness** of objects and scenes
* Claims that **sensations combine to form perceptions**
* *Conscious awareness **merely** sum of sensations*
* **Contain nothing that isn't present**
27
What is gestalism?
* Directly **contradicts structuralism**
* *Conscious awarenress is **more than** sum of sensations*
* Can have **characteristics not present in sensations**
* Is **adaptive**
28
What is the evidence for gestaltism?
* **Apparent Motion**
* 2 **stationary dots flash in succession**
* Although both stationary **motion is perceived** → Constructed AKA **wasn't actually present**
* **Illusory Contours**
* Locations with **no physical contours**
* **Awareness of contour constructed**
29
How is perceptual organisation achieved?
## Footnote
Gestaltism
* **Grouping**
* **Parts** of an image are **perceptually bound together** to form a **perceptual whole**
* Elements of an object brought together
* **Segregation**
* **Parts** of a scene are **perceptually separated** to form **separate wholes**
* Separate objects
* *Together allow scene to be perceptually organised into objects*
* **Make sense** of the scene
30
Are you able to make sense of a scene without grouping or segregation?
* **No**
* **Would not** be able to **perceptually organise scenes** into **meaningful units**
31
Name the gestalt principles of grouping (5+2)
* 5 principles
* **More applied** = **More likely** components will be **grouped to form perceptual object**
* Original
* Good continuation
* Pragnanz
* Similarity
* Proximity
* Common fate
* Added later
* Common region
* Uniform connectedness
32
Describe good continuation
## Footnote
Gestalt - Grouping
* Helps the difficulty of when objects are hidden
* **Aligned contours grouped together** to **form single object**
33
Describe Prägnanz
## Footnote
Gestalt - Grouping
* Principle of good figure/ simplicity
* Groupings occur to make **resultant figure as simple as possible**
* Grouping into one entity
* Panda > Bunch of blobs
## Footnote
German for "Good figure"
34
Describe similarity
## Footnote
Gestalt - Grouping
* **More similar** = More likely to be grouped
35
Describe proximity
## Footnote
Gestalt - Grouping
* **Closer** = More likely to be grouped
36
Describe common fate
## Footnote
Gestalt - Grouping
* **Moving in same way** = Grouped
37
Describe common region
## Footnote
Gestalt - Grouping
* Elements within **same region of space** = Grouped
## Footnote
Was added later
38
Describe uniform connectedness
## Footnote
Gestalt - Grouping
* **Connected regions with same visual characteristics** = Grouped
## Footnote
Was added later
39
Why is segregation needed?
## Footnote
Perceptual Organisation
* **Not enough to only group** → Also need to **segregate**...
* Objects **from each other**
* Objects **from background**
* **Without** = **Perceive** image as **single object**
* Confusing
40
Describe figure-ground segregation
* **Objects = Figures**
* **Identifying figures** → Would typically be able to **identify objects**
* **Background = Ground**
41
What are figural properties?
## Footnote
Figure-Ground Segregation
* In front of rest of image
* At bottom of image
* Convex
* Recognisable
## Footnote
Depth ordering changes figure perception
42
Describe using experience to segregate objects
* **Prior knowledge aids segregation**
* Seeing something → Cannot unsee
43
Describe gist perception
* **Overall impression** of a scene
* When scene is being **flashed rapidly**
* Can be done **rapidly but not very detailed**
* **27ms**
* **Longer observed = More detail**
* **250ms → Very accurate**
