Face recognition (within category judgement)
Familiar / unfamiliar face rec. is different
Familiar Faces (Paul McCartney)
Viewpoint independent
Time independent
Change independent
Unfamiliar Faces
Generally Worse
Kemp: Credit card image matching 36%correct
Bruce: Video Line Up 80% correct
Influenced by pose, lighting and viewpoint
Neuropsychological Evidence for indepent Modules by: Young: ex-servicemen
Concious (overt) vs. covert recog. points to seperate neural pathways
Covert Dorsal
Overt Ventral
IAC explains this: FRU to SIU-Not enough activation for covert/ enogh for overt
Face perception Organisation (Stages)
Starts where perception ends=structural classification
Bruce and Johnstons Speed Measuring: Recogntion happens in Stages
1. Perceptual classification, 2. Sematic classification, 3. Naming
Highly difficult task: High leval of detail
We have expertise through attentional bias from birth
Bruce and Youngs General Model of Face Recog
Cognitive System : Overrule decisions, if unsure
Yin: Immediate recognition is good - probably image recognitino, not face
Face Recognition
Face Recognition units (Familarity)
Person Identiy Nodes (Semantic knowledge)
Name Gerneration
Facial Expression Analysis
Facial Speech Analysis (Lip reading)
McGurkEffect (ba, ba, ga)
Burton & Bruce IAC (Interactive activation and competition)
Visual Recog
FRU's (Face Rec Units)= sensory id
SIU's (It's a singer and his name is Michael J.)
PINS (Personal Id nodes) One node matches a person
Say: its M. J.
Name Recog
WRUs (general word recgo units)
NRU's (Name Recog Untits)
Is it different from Object Recog?
Neuropsychological Evidence
Double Dissotiation
Prosopagnosia
Inahbility to recognize familiar faces
Capgras Delusion
Agnosia
Inability to rec objects ("outpouchings")= stop at percept (2,5 D or 3D) but no recogn)
Damage at earlier stages visual procc: A brush: "A wooden stick and a black object" (Humprheys)
Different Locations (fMri) in the brain from object recog
Is it just different because of within/ between category classification (unfair comparison)
Same processing for face vs. object recognition
Location specific: Evidence for a "Granmother" cell
Innate/ but Kick Start Johnson & Morton: Newborns show attentional bias to faces
Holistic/ configural processing
Diamond & Carey: Dog experts vs. non Experts worse at inversed dogs
First order: eyes above nose
Evidence: Searcy & Bartlett: Inversed spatial relationships were less grotesque
Second order: Differences in those relations
Quoalitatively Different: Yin: Inversion effect not for objects ?
Issues with explanation of face recognition
Same three stages as object recognition: Percept.recog , sematic knowlege, naming
Evidence Youngs diary sutdes
Faster Rts of famil vs. semantic
Face Experts: innate bias toward faces:
Its a within categ. comp - with same cat. objects
Facees all look similar, objects dont
We need to account for facial emotions and facial speech
Difference familiar / unfamiliar faces
Matching rate only 80%
More holisitc
Yins Inversion effect is more destructive for faces comp to objets
But dog experts have more problems with inversed dogs
Grotesque
1st order: eyes, nose, mouth
Inverted impaired more 2nd order features: Epert tuning lost
Objects and Faces in different areas
Double dissotiations
fMRI
Neurological issues
Prosppagnosia
Two Routes: Dorsal / Ventral
Youngs ex-serivcemen
Facial Expression Analysis may still work
Familiar / Unfamilar reogntion may be affected
Bauers Covert Recognition
GSR response
Points to damaged Ventral (what)
IAC expalins this as weekend FRU to PIN connections
PIN to SIU ok (Therapy)
Capgras delusion both for face and objects
Damaged ventral
Bruce & Youngs ("Square" model)
How?
1.Structural encoding (its a face?)
Viewpoint (2D) dependent
Viewpint indendent (3D) descriptions
2A. Face Recogntion units (FRU: i know him)
2 B. Person identity nodes PIN (a musician)
2 C. Naming (its Mick Jagger)
3A Facial Expression (emotional) analysis
3B. Facial Speech analyis
4. Directed visual Processing: Other semantic information (singers are attractive- check out his eyes
5.Cognintve System : Evaluate and feed back on what we see
Evidence
Naming=different
Young, Hay Allice: Diary Study: we can't name unless semtic info
Object / Face diference
Double dissitiation
Face recogntion
Propagnosia: some dont recognize faces
Object recognition
Agnosia: "Outpouchings" (Oliver Sacks)
3 Stages: Bruce Rt's
a) Familiarty is faster than sematnic (occupation)
b) x is a politician faster than naming
Two Routes?
Familiar
Structural: its face
Face Recognition (i know him)
Identity nodes (its a singer)
Name generation (its mick jagger))
Unfamiliar
Structural
Expression
Speech
Directed visual processing (difference to Familiar to check specific facial detail)
Evidence
Familiar / Unfamiliar=different or just harder
Malone: Doulbe dissistiation with 2 Patients (famous statesmen)
_Burton & Bruce IAC
connectionist model