M3, face recognition and face processing problems Flashcards
Prosopagnosia
- Not unitary disorder -> a general term describing an inability to recognise familiar faces and the inability to learn new faces (can be acquired or developmental)
- termed coined by Bodamer (1947), possibly first report of distinct face recognition impairment
- Charcot (1883) and Wilbrand (1882) reported patients with face recognition problems but patients also had widespread perceptual problems and memory difficulties
Cognitive evidence for a specialist face processing module (system)
Visual Illusions with Faces
- Face Inversion effect (Yin, 1969)
- Thompson or Thatcher Illusion (Thompson, 1980) (inverted features)
- Composite Face Effect (Young, Hellawell & Hay, 1987)
Face inversion effect
- Face recognition more difficult for inverted compared to upright faces
- Evidence – upright faces = holistic analysis
- Inverted faces = analysis by parts (cannot use face system)
- Evidence consistent with prediction of impaired face recognition only upright faces
- > Normals – correct 94% upright, 82% inverted
- > LH (prosopagnosic patient) – correct 58% upright, 72% inverted
Patient (cognitive) evidence for specialist face processing system
Farah (1990), Acquired alexia
- Reading problems, read letter by letter
- Visual confusions between words e.g., ball doll
- Left hemisphere lesion – angular gyrus posterior region parietal lobe
- Temporal lobe – face processing
- Specific type visual agnosia – patients can comprehend, speak and write -> Deficit – within category (words)
- Often occurs with impairments in object recognition
- Object recognition – decompose stimuli into parts
- Face recognition – overall configuration (holistic)
Neural evidence for specialist face processing system
- Often co-occurrence of object and face recognition impairments
- Impairments face recognition – associated with multiple lesions
- Bilateral lesions – multiple strokes, head trauma, encephalitis, poisoning
Neural evidence for specialist face processing system - single cell recordings in primates
Stimuli presented: monkey faces, human faces, stimuli with face features/characteristics
Found cells that selectively responded to
-> frontal monkey profile
-> others all facial stimuli
But cannot conclude cells purely for face processing
Infant development of facial discernment
Normal infants
- Can recognise differences between face like patterns and patterns with same features but randomly organised as soon as a few hours after birth
- Newborn babies can distinguish between their mother’s face and the faces of other women at 2 days old
- Babies a few months old can recognise familiar faces
- The rapid development of face recognition suggests that this skill might be modular (Fodor, 1983)
Evidence for genetic basis for prosopagnosia
Developmental prosopagnosia – includes individuals with congenital prosopagnosia and individuals who have sustained brain damage either before birth or in early childhood
Congenital prosopagnosia - impairment in face processing that is present from birth, in absence of brain damage; normal intellect and sensory processes
One idea - Developmental Prosopagnosia is caused by a generalised deficit in configural processing
Evidence suggests that Developmental (Congenital) Prosopagnosia runs in families
Bruce and Young (1986) model of face recognition
- Hierarchical and parallel model of face recognition
- Uses stages of processing
- Allows an understanding of the types of different information used when processing/recognising faces
- look up image
Structural encoding stage - Bruce and Young (1986) model of face recognition
- Form a face percept via parallel extraction of different types of feature based information from faces
- Viewer Centred Descriptions – viewpoint specific face representation of the face
- > Provides information relevant to recognition of facial expression, understanding speech, and processing similarities/differences between unfamiliar people (gender/age etc.) by strategic attending to visual appearance of a face
- Expression Independent Descriptions - encode different view points of the face, including the configural layout of the face and specific features
Bruce and Young 1986 model of face recognition - analysis and processing stages
Expression Analysis – determine facial expression independent of face
Facial Speech Analysis – information from lip and tongue movements of speaker
Directed Visual Processing – attention to face characteristics, learning new faces
Face recognition units - Bruce and Young 1986 model of face recognition
- Input from Expression Independent Descriptions
- FRU contains stored structural (visual) description of familiar faces
- FRU activates when there is a strong match between the face encoded and a stored structural description
- Separate FRU for each face we know
- FRU can also be activated via semantic information in the person identity nodes
- FRU responds to a known face, any given angle, but not to other person specific information (e.g., voice)
Nodes and name - Bruce and Young 1986 model of face recognition
Person Identity Nodes
Post FRU activation, then semantic (bibliographic) information about the person is activated (e.g., occupation, person characteristics etc.).
Name Generation Stage
Activation of the PIN and subsequently semantic information allows you to be able to name the face
Cognitive system - Bruce and Young 1986 model of face recognition
- All other processes that may be involved in or relevant for face recognition
- episodic memories, attention and decisional processes
Prosopagnosia and the Bruce and Young 1986 model of face recognition
Prosopagnosia Types - two distinct groups
Impaired ability to perceive faces
-> Defect affects Structural Encoding – model
Impaired face recognition
- > Intact perceptual abilities but cannot recognise or process faces satisfactorily
- > Problem with face recognition units
Burton, Bruce and Johnston (1990) face recognition model
- An interactive activation and competition network
- Three primary pools of units - Pools of units containing face recognition units, person identity nodes, and semantic information units
- Within each pool, every unit is connected to every other unit with an inhibitory link
- Also excitatory links between the pools, connecting particular units (related).
- Model is an implementation and development of Bruce and Young’s (1986) functional account of face recognition.
Haxby et al (2000) model of face recognition
*look up image
Evidence of face recognition without awareness
- Overt vs. Covert face recognition and processing
- Automatic vs. controlled processes
- Does prosopagnosia mean no face processing?
- > Overview of physiological studies
- > Overview of behavioural studies
Covert face recognition physiological evidence
- Patient LF (Bauer, 1984)
- Task to select the correct name from five alternatives to match a photograph of a familiar face
- LF performed at chance to identify the faces
- Skin conductance responses (SCRs) were greater and more frequent to the correct name compared to distractors
Covert face recognition behavioural evidence
De Haan et al. (1987a) Patient PH
- Cannot recognise familiar faces
- Performs at chance on a forced choice test requiring to choose the familiar face from two items
- PH shows evidence of covert face recognition on a
- > Face matching task
- > Interference paradigm
- > Associative priming task
Face matching task - Patient PH - covert face recognition behavioural evidence
Face Matching Task – are 2 simultaneously presented faces the same person?
- Controls are faster to indicate a match for familiar compared to unfamiliar faces
- PH also shows that effect
Interference experiment - Patient PH - covert face recognition behavioural evidence
Interference Experiment/Task – faces are shown with a printed name
- Name is that of the person, a person with a related occupation, or an entirely unrelated person
- Participants indicate if the name is that of a Politician
- Controls take longer if the name is unrelated to the face
- PH also shows this effect
Associative priming task - Patient PH - covert face recognition behavioural evidence
- A prime face is presented before a written name target, decide if target name is familiar or not
- Controls faster responses when the prime is associated with a target compared to when there is no association between the prime and target or the prime is the face of an unfamiliar person
- PH shows the same effect
- Not all Prosopagnosia patients show convert effects, some show affects only to a subset of known people
Anomia - people’s names
- Patient GBL - McKenna & Warrington (1980)
- Patient anomic only for people’s names
- Could only name 3/20 photos of famous people correctly
- But could accurately describe 18/20 correct who they were
- Correct 16/20 European towns and 12/12 English towns
Face anosognosia
Young, De Hann & Newcombe (1990)
- Woman suffered a severe right hemisphere stroke
- Poor at recognising familiar faces
- She was completely unaware of this impairment
- Complete lack of acceptance of this impairment
- Able to recognise other cognitive limitations
- Possibly explained by the automaticity of face recognition
Metamorphopsia - perceptual distortion
Bodamer (1947) Two patients saw distorted faces
- One patient described faces as having noses turned sideways by several degrees, one eyebrow higher than the other, mouth askew, hair shifted like an ill-fitting cap
- Third patient metamorphic but could recognise faces to some extent
Face processing - patient JS - Heutink et al (2012)
- Patient JS had a stroke, reported problems in recognising family members visiting her in hospital
- > JS did not recognise the daughter she sees regularly, recognised the daughter she had not seen for 8 yrs
- > Would not allow grand kids to sit on her lap because they looked repulsive
- Impaired perception of facial expression
- Reported the pictures of family members had distorted facial proportions and she could not tell which way the proportions were distorted
Patient JS - SCR and diagnosis
- Compared to controls JS was slower and less accurate at identifying celebrity faces and even more so for family members
- SCR indicated lower psychological arousal to faces than controls but her SCR for family was higher than celebrities
- > Impaired recognition of basic emotional expressions
- > Poor recognition of familiar faces in general but this is worse for family members
- > Face of family members appear distorted
- JS does not have Prosopagnosia, JS does not have Capgras Delusion, might have mild Metamorphopsia (Prosometamorphopsia)
- Could be due to poor integration between identity and emotion related processing of faces
Super recognisers of faces - Russell, Duchaine and Nakayama (2009)
Tested face recognition abilities of 4 participants who never forget a face (Super-recognisers)
Tests included
Before They Were Famous test, Cambridge Face Memory test, Cambridge Face Perception Test
- Super recognisers out performed controls on before they were famous and memory task
- Super recognisers had better face perception than controls and were about as good as prosopagnosics were bad
-> Continuum of face recognition skills?
Super-recognisers of faces - Bobak, Hancock and Bate (2016) - experiment 1
Compared the face processing skills of super-recognisers and typical perceivers (Experiment 1)
- Face Matching Task
- Super-recognisers were more accurate, better sensitivity scores and higher response confidence than typical perceivers
- CFMT performance correlated with face matching task performance
Super-recognisers of faces - Bobak, Hancock and Bate (2016) - experiment 2
Compared the face processing skills of super-recognisers and typical perceivers (Experiment 2)
- Face Matching Task using photos for study and video clip for test
- Super-recognisers were more accurate and better sensitivity scores than typical perceivers
- No significant difference between super-recognisers and typical perceivers for response confidence
- CFMT performance correlated with face matching task (photo/video) performance
Police force super-recognisers of faces - Robertson et al (2016)
- Metropolitan Police Force London Super-Recognisers
- Do Police Force Super-recognisers outperform Controls (Police Trainees or University Students) on Face Processing/Recognition Tasks?
- > Glasgow Face Matching Task (GFMT) [Police Super-recognisers, Police Trainees]
- > Models of Face Matching Test (MFMT) [Police Super-recognisers, University student controls]
- > Pixelated Lookalikes Test (PLT) [Police Super-recognisers, University student controls]
