Face Perception LOs Flashcards
Why study face perception?
Provides important information critical to social function (identity, gender, age, facial expression, gaze direction, health, speech)
Provides insight into intraclass discrimination (not addresses by most theories of object perception which look at interclass)
May inform on the nature of hemispheric processing
LH: decomposition / featural processing
RH: composition / holistic processing; summation of first order arrangement and second order relational information
Interesting example of object perception in the ventral stream
Holistic Evidence to face processing
Holistic View: Individual features are only important in that they comprise part of the face context and help make up a whole face, face perception primarily based on processing holistic info, proposes face patches (specialized mechanisms for face perception, brain circuits that are not used in the perception of other objects)
Holistic Evidence: newborn infants: gaze longer at pictures of mother than of a stranger, look longer at attractive faces, can discriminate and imitate facial expressions at 36 hours after birth
Tanaka/ Farah: part vs whole (1993): training phase was participants asked about faces and houses. Isolated part condition: given choice of two object parts pick which one had been presented before (eg. Which of these is Larry’s nose, which of these is Bills door). Whole-object condition: given choice of two whole objects pick out the one they had seen earlier (eg. which of these is Larry’s face, Which of these is Bill’s house). Participants equally good at recognizing parts of houses and whole houses. For faces participants were not as good at recognizing face parts as they were at recognizing whole faces. Face recognition seems to be holistic
Inversion effect: upside-down faces are more difficult to identify; perception of features is unaffected. Faces are represented as undifferentiated wholes, inversion interferes with this processing necessitating feature-by-feature analysis (slower/less accurate)
Scrambled faces difficult to identify: upright faces cause strong N200 brainwave in right fusiform gyrus but not scrambled or inverted faces, cars or butterflies
Neuro Imaging Studies: FFA located in right lateral fusiform gyrus, fMRI found face-specific attention activated by FFA more strongly than objects of similar complexity. PET scans found right fusiform gyrus activity greatest during face memorization tasks; activity correlates with performance. Face-selective cells found in amygdala, frontal lobes and other regions
Brain damage/Neuropsych: prosopagnosia: failure in visual processing of faces which is not due to a general intellectual impairment, sensory impairment or language disorder, caused by damage to FFA
Double dissociation of face/object perception:
case study of W.J. : severe prosopagnosia due to RH damage, normal visual acuity and LH functioning, was much worse at recognizing famous human faces than individual sheep from a flock he has acquired after getting prosopanosia. Recognition memory performance with unfamiliar sheep surpassed that of normal controls, matched for profession and experience with sheep. W.J. had face-specific deficit not generalizable to other classes of complex stimuli.
Case study of C.K. : visual object agnosia (cannot identify things visually) with intact face perception. Can identify famous faces and match unfamiliar faces but cannot distinguish a feather duster and a fart. When shown paintings of fruit/faces could not see the fruit for the faces.
Featural Approach to face processing
Featural Approach: face processing is just an example of intraclass discrimination; there is nothing special about it, its just a result of experience. Face perception primarily based on processing facial features (or a combo of features), no specialized brain mechanisms for face perception (no different than differentiating between cars)
Featural Evidence: RH bias for faces not found until age 12: may develop with experience
Inversion effect: inversion interferes with processing because it changes the context in which we normally see faces. Europeans show greater inversion effects for European than Asian faces. Viewing dogs elicited an inversion effect comparable to faces, but only for experts (judges with 10 years of experience and only for their specific breed of expertise)
Neuroimaging: in addition to having face selective areas” ventral temporal lobe has areas that prefer other stimuli like hands, letter string and gratings. Occipital face area located in inferior occipital gyrus activated by individual facial features. Face selective region in STS sensitive to face parts, not to correct facial configuration
Visual Experience: task: classify imaginary, faceless plantlike creatures called greebles according to sex and family. Participants displayed elevated fusiform gyrus activity as they identified pairs of matching greebles, only found for greeble experts (10 hours of training), only for upright greebles, right anterior fusiform, another face area, not activated by greebles at all (may be specialized for faces)
Prosopagnosia: those with prosopagnosia show deficits in intraclass discrimination (greebles, snowflakes) as well as face perception deficits. C.K. (visual object agnosia) could not discriminate greebles, this does not support featural view
What are some face perception disorders that can help explain face processing in the brain
apperceptive prosopagnosia
association prosopagnosia
capgras syndrome
fregoli delusion
Apperceptive Prosopagnosia
Apperceptive Prosopagnosia: impairment in basic face perception, patient cannot “see” faces normally, cannot determine they are looking at a face and cannot identify it. Visual association areas within the right occipital and temporal regions implicated.
Case study of Unteroffizier S: patient unable to identify previously familiar faces (famous faces, friends, family and own face)
damaged featural processing system
Association prosopagnosia
can determine they are looking at a face but associated information (eg. name) can no longer be retrieved, preventing identification. Right anterior temporal regions implicated.
Case study of Mr. W: 54 year old Belgian farmer could copy drawings of faces, match faces, determine gender through faces. Poor at identifying famous faces and at familiarity judgement of personally known faces
damaged holistic processing system
Capgras syndrome
delusion that a close relative or friend has been replaced by an exact double (an imposter), imposter is a key figure for patient at time of onset of symptoms, patient can also see themself as their own double, may believe that inanimate objects have been replaced by exact doubles, not obtained when talking to people on the phone. Associated with brain injury, history of delusions, psychoses, may be failure for recognition to elicit appropriate emotional output.
Hirstein/Ramachandran (1997) syndrome due to disconnection between face processing areas in temporal lobes and amygdala. Hypothesized path: fusiform gyrus -> superior temporal sulcus (STS) -> amygdala. Connection aids in perceiving facial expressions of emotion and feeling emotions in response. It bypasses areas for semantic interpretation of faces, language areas. Unaccustomed lack of emotional response to close relatives interpreted as reaction to imposter. Those with prosopagnosia have normal GSR to familiar faces but those with Capgras do not.
Double-dissociation of face perception and emotion: people with Capgras can identify faces but seem to lack an emotional response to them. People with prosopagnosia cannot identify faces but may be able to identify emotional facial expressions
Fregoli delusion
patient has delusional belief that different people are one single familiar person who changes appearance or is in disguise. Patients usually believe this familiar person to be a persecutor who is following them. Right anterior fusiform gyrus damage is implicated as well as temporal lobe locations involved in face recognition.
Ramachandran/Blakeslee (1998): in Capgras syndrome problem may be underactivation of normal autonomic arousal, in Fregoli Delusion problem may be excess inappropriate arousal to viewing unfamiliar faces in temporo-limbic connections
What makes faces attractive
Averageness: Face-Space theory: proposes faces are coded as points in hypothetical multidimensional space, average face is located at the centre, faces are normally distributed, those closer to the average are more prevalent, easier to process
Symmetry: our genes predispose us to develop symmetrically however disease and infections during development produce asymmetries, symmetry = healthier
Sexual dimorphism (masculinity in female faces and vice-versa)
Youthfulness
Good grooming
Pleasant expression
Liking (for familiar faces)
What are some effects of facial attractiveness?
Hamermesh and Abrevaya: happiness, analyzed data sets from Canada, Germany, UK and USA. Attractiveness is correlated with happiness and life satisfaction,(for every 1 SD change in beauty, there is a +0.08 (men) or +0.07 (women) SD change in life satisfaction/happiness effect is halved by taking into consideration covariates of educational, marital and labour-market outcomes, monetary outcomes also have a small effect
Mocan and Tekin: criminal behavior. 15 000 high schoolers were interviewed in 1994, 1996 anf 2002. Interviewers rated the physical appearance of the student on a 5-point scale from very attractive to very unattractive. Unattractive individuals commit more crime in comparison to average looking ones. Very attractive individuals commit less crime compared to average looking ones
Todorov et al. Election Results: participants viewed black and white photos of actual US political candidates. Inferences based only on facial appearance predicted outcomes of elections better than chane. Candidate viewed as more competent won: 72% of senate races, 67% of house races. Same inferences drawn when faces presented for only 1 second. Swiss children asked to choose the captain of their boat, chose the winner of French elections 71% of the time
How does facial attractiveness affect face perception
Cross et al (1971): other’s faces. Participants looked at photos of students and teachers taken from high school yearbooks, were later given a surprise recognition test. Faces judged to be attractive were more likely to be recognized (54% vs 38%)
Bower and Karlin (1974) Deeper Traits. When faces were judged on deeper emotional traits (honesty, likeability) and not on surface details like attractiveness facial recognition improved
Epley and Whitchurch (2008) your own face, 1) participants face was computer enhanced to be 20% more (or less) attractive, attractively morphed face produced faster recognition than actual face (176 ms faster) or unattractive morph (240ms). 2) when own face was unchanged, uglier or prettier, participants were more likely to choose the more attractive face to be their real face we see our own face as being more attractive than it actually is
