Lecture 7&8 - Perceiving faces

Question 1

Why is face perception interesting?

Answer 1

• Faces are uniquely rich in information: identity, familiarity, age, race, gender, gaze direction, attractiveness, mood, communication
• There are practical applications to this research: CCTV, passports, face ID

Question 2

What sort of objects is a face?

Answer 2

• One with extremely similar distractors: it is withing-class, not between-class recognition
  A very changeable one: rigid transformations (head movement, viewpoint), non-rigid (expressions, speech), shape and texture (aging), colour (emotion, health, temperature, tan

Question 3

We are very good at facial recognition

Answer 3

• Even when extremely distorted, we can still recognise faces
• We even see faces where there are none (i.e. in clouds, soapy water)
• This is known as pareidolia
• Studies suggest all you need is a symmetrical noise pattern with a natural distribution of spatial frequencies

Question 4

What is BOLD signal and MRI

Answer 4

• BOLD signal (Blood Oxygenations Level Dependent) is used in fMRI (functional Magnetic Resonance Imaging) to see brain activity
• When neurons in a brain area become active, they need more energy, which comes from oxygen in our blood
  1. Oxygen-rich blood flows to active areas
  2. MRI detects the differences between blood that has a lot of oxygen and blood that has less. It turns these differences into images showing which brain areas are most active
  3. fMRI results are always a difference between conditions to remove background or unrelated brain activity and highlight areas that are specifically involved in the task

Question 5

Where are faces processed?

Answer 5

• The Fusiform Face Area (FFA) is a face-selective region, as shown by contrast studies. They compared BOLD signal for faces vs other objects; highlighted regions had higher signals for faces
• Evidence also comes from physiology
• Neural signalling in monkey FFA was highest for faces (of the same species). When scrambled or partially obscure, the response went down
• Multiple areas are involved in face processing

Question 6

Perspectives on face recognition: 2 hypothesis

Answer 6

• There are two main ideas on facial perception:
  1. The domain specificity hypothesis (faces are special) – we are born with dedicated mechanisms for facial recognition, which operate differently to those that serve typical object recognition)
  2. The expertise hypothesis (faces are not special) – face perception simply shows us how general object recognition mechanisms work for objects we are extremely well practiced at observing)

Question 7

Domain Specificity Hypothesis: Evidence 1: neonatal face discrimination

Answer 7

• Is there an innate ability to recognise faces? Newborn babies seem to prefer to look at face-like patterns more than non-face-like patterns
• However, this might be a broader preference for top-heavy patterns (i.e. More patterns at the top of the image than bottom)
• But babies as young as 1-4 days old seems to be able to tell their mother’s face from that of a stranger
• Uses habituation (e.g. show baby picture of mother and stranger and times how long the baby stares at both images)

Question 8

Domain Specificity Hypothesis: Evidence 2: prosopagnosia

Answer 8

• Some people cannot recognise faces, they often have different gaze patterns. Acquired through damage to occipito-temporal regions, although very rarely isolated completely to faces
• Can be developed hereditary (2% population have developmental prosopagnosia) – show slower performance on object recognition tasks
• However, some people are also very good at distinguishing between faces (called super-recognisers) and can be explained by genetic factors

Question 9

Domain Specificity Hypothesis: Evidence 3: the inversion effect

Answer 9

a) Bistable ambigram face drawings (different face when presented upside down)
b) Pareidolia is orientation specific (cannot see the face when image is presented upside down)
c) The thatcher effect (we didn’t notice anything was wrong with the upside-down face until it was presented normal)

Question 10

Domain Specificity Hypothesis: Evidence 4: sensitivity to facial configuration

Answer 10

• The inversion effect disrupts configural information more than featural
• Spacing between features have been changed and the faces look differently
• Results: inversion disrupts configural more than featural information
• This is evidence of holistic processing: the inability to attend to one part of the face
• Change one part (e.g. the mouth) and the whole face looks different

Question 11

Domain Specificity Hypothesis: Evidence 5: part-whole effect

Answer 11

• Sub-parts of faces are not independently recognisable
• Training phase: pp’s were given a face to remember, either whole or scrambled
• Testing phase: the pp’s were given a distinguishing task, where one thing (i.e. nose) had been changed
• Results: pp’s trained on the whole face were better at identifying the whole face. Pp’s trained on scrambled faces were better at identifying individual parts
• This is evidence that when given the whole face to learn, it was processed holistically
• When trained on the scrambled face, face-specific mechanisms were not activated, and the component parts were processed individually

Question 12

Domain Specificity Hypothesis: Evidence 6: composite effect

Answer 12

• We can’t help but see the whole face
• Measured RT for identifying the top and bottom faces, either aligned or misaligned, upright or inverted
• The composite effect slows the RT for aligned faces, but only when they are upright
• More evidence of compulsory holistic processing for upright faces

Question 13

Advantages of domain specificity hypothesis: and what evidence do we have?

Answer 13

1. Information on specific features of faces can be misleading because different faces might share similar features (e.g. colour or shape of eyes) or because face features can change
2. Integration could rely on parallel processing and ensure a faster recognition of a highly relevant stimulus
   What evidence do we have?
3. Part-whole effect – we are better at recognising specific facial features when those features are presented within the context of the whole face
4. Composite face effect – when two halves of different faces are combined, we perceive the combination as a face as a single, unified whole
5. Inversion effect – we struggle to process up-side-sown faces

Question 14

Expertise hypothesis: Evidence 1: the effect of (un)familiarity

Answer 14

• We are so much better at identifying people we have already seen
• So facial recognition is heavily dependent on familiarity and might involve specialised processes
• Disclaimer! This evidence could be used against the expert hypothesis

Question 15

Expertise hypothesis: Evidence 2: the “other race” effect

Answer 15

• Sensitivity to differences between faces seems to require specific experience
• People are better at remembering, more accurate at matching, and can make finer discriminations amongst faces of their own race rather than another

Question 16

Expertise hypothesis: Evidence 3: object inversion in experts

Answer 16

• Orientation is more critical in situations where the pp has extensive practice in making subtle object discrimination
• Tested experts in dog breeds on subtle differences in pictures of dogs, upright and inverted
• Results: both dog experts and non-experts were worse at recognizing faces when they were upside down
• But only dog experts were worse at recognising dogs when they were inverted

Question 17

Expertise hypothesis: Evidence 4: the part-whole effect in objects

Answer 17

• Parts are often recognised better in their original context, not just faces
• Trained pp’s to recognise a stimulus
• Showed them a target stimulus and told them which part of it to attend to (e.g. ears), then showed that part either in isolation or in its trained configuration
• Results: accuracy was much better when the parts were still in situ
• So, the part-whole effect exists for objects too

Question 18

Expertise hypothesis: Evidence 5: FFA activation in car experts

Answer 18

• The FFA may simply be an area responding to expertise
• FFA activation as a response to faces, animals, cars and planes
• Most voxels preferred faces
• But the amount of these voxels was activated by cars were correlated with how expert the person was with cars
• So, the FFA may aid the perception of images in which we are experts

Question 19

Expertise hypothesis: Evidence 6: prosopagnosia

Answer 19

• According to expertise hypothesis, brain damage in prosopagnosia might affect areas associated with expertise in general rather than faces
• There is evidence that some patients with acquired prosopagnosia have problems recognising objects who have expertise in
• This suggests they might have general expertise-related deficits
• However, individuals with developmental prosopagnosia can become experts, showing an independent mechanism for faces and expertise in objects

Question 20

Summary of The Domain Specificity Hypothesis

Answer 20

• facial recognition appears to be innate
• some people are exclusively bad at recognising faces
• inverting a face makes recognition more difficult
• faces may have compulsory holistic processing - its hard to selectively attend to specific parts

Question 21

Summary of The Expertise Hypothesis

Answer 21

• familiarity with faces is important, for individuals and races
• experts experience objects inversion effects
• parts of an object are also recognised better in their original environment
• FFA may support general expertise, not just faces