Brainscape's Knowledge GenomeTM

Browse over 1 million classes created by top students, professors, publishers, and experts.


See full index

Cognition & Emotion II > Object Recognition II: Faces > Flashcards

Object Recognition II: Faces Flashcards

1
Q

TANAKA & FARAH (1993)

A
  • faces are processed holistically
  • when split horizontally in half and mixed w/each other, identical top faces were perceived as being dif when aligned with dit bottom halves; misaligning the bottom halves broke the illusion
  • this powerful visual indicates that the visual system automatically glues 2 halves of a face into an integrated configuration aka. holistic face perception
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

PROSOPAGNOSIA

A
  • acquired deficit in facial recognition post brain damage; patients lose ability to recognise friends/relatives/to learn new identities
  • can still recognise people by their voices
  • remote memories of known people remain intact
  • cognitive skills/visual abilities oft remain intact
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

PROSOPAGNOSIA: ROSSION (2014)

A
  • prosopagnosia can result from lesions in any region of right ventral occipitotemporal cortex (esp. lingual gyrus/parahippocampal gyrus/fusiform gyrus aka. ventral cortical surface)
  • right hemisphere dominance
  • difficult to pinpoint specific region always damaged in prosopagnosia
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

FUSIFORM FACE AREA (FFA): KANWISHER ET AL. (2014)

A
  • used region of interest (ROI) approach
  • functional localiser scan to identify face-selective voxels
  • subsequent scans to test selectivity of voxels to other stimuli/rule out confounds
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

VENTRAL VISUAL CORTEX SELECTIVE PROCESSING: ADDITIONAL FMRI EVIDENCE

A
  • no others show selective pattern of activation in a circumscribed cortical region
  • only biologically important stimuli seem to have dedicated processing modules
    EPSTEIN ET AL. (1999)
  • the parahippocampal place area (PPA) = region in ventral visual cortex; activates selectively to scenes
    DOWNING ET AL. (2001)
  • extrastriate body area (EBA) = region in ventral visual cortex; activates selectively to pictures of human bodies
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

FFA MODULES HYPOTHESIS: CHALLENGES

A

1) expertise-related activation in FFA
2) activation in other brain regions to faces
3) developmental prosopagnosia
4) distributed activation patterns to dif object categories in ventral visual cortex aka. multivariate fMRI evidence

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

FFA CHALLENGES: EXPERTISE

A

GAUTHIER ET AL. (1999)
- trained pps to recognise novel objects (“Greebles”)
- found activation in FFA in “Greeble” experts BUT not novices

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

FFA CHALLENGES: FACES VS EXPERTISE

A

GAUTHIER ET AL. (2000)
- showed bird/car experts pics of birds/cars/faces
- stronger FFA activation to birds in bird expertise (vice versa)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

EXPERTISE HYPOTHESIS: EVALUATION

A
  • evidence for ^ FFA activation for “expertise” = weak/inconsistent; increases = small; several studies failed to replicate findings
  • “Greeble” experiment confounded by similarity of stimuli to faces
  • prosopagnosics can become experts at identifying other objects ie. prosopagnosic sheep farmer could recognise individual sheep
  • part/whole beh effects = observed for faces BUT not for other “expertise” objects ie. dog experts
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

FFA CHALLENGES: MULTIPLE FACE-SELECTIVE CORTICAL REGIONS

A

KANWISHER ET AL. (2017)
- brain regions never work alone, regardless of specificity; they all need inputs (to provide info to process) & outputs (to inform other regions what they’ve learned)
- much confusion sowed by referring to similarly selective regions spaced far apart as “distributed cortical system” BUT multiplicity/spatial separation of such regions doesn’t argue against functional specificity

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

FFA MODULE HYPOTHESIS: CHALLENGES

A

DEVELOPMENTAL PROSOPAGNOSIA
- facial recognition impairment NOT as a result of brain injury; present from birth
- affects 2-5% population
- neural basis still in debate; clearly no obvious pathology (ie. FFA lesions)
- inconclusive functional imaging evidence
- some studies showed difs in activation/connectivity between developmental prosopagnisics/controls; others haven’t

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

FFA CHALLENGES: MULTIVARIATE FMRI

A
  • univariate fMRI looks for activation “peaks”
  • multivariate fMRI looks for activation “patterns”
  • crucial distinction; multivariate fMRI can ask what info is represented in activation patterns across brain region
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

MULTIVOXEL PATTERN ANALYSIS (MVPA)

A
  • picks up on dif in info represented in neuron pops that show little sensitivity to such difs in univariate analysis, which doesn’t show this
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

MVPA: HAXBY ET AL. (2001) PROCEDURE

A
  • presented pics of faces/houses/other objects in scanner; pps performed 1-back task to ensure attention to stimuli
  • subjects scanned in 12 “runs”
    DATA ANALYSIS
  • no spatial smoothing (interested in single voxel responses); measured activation in each vowel to each object category for each run/category
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

MVPA: HAXBY ET AL. (2001) RESULTS

A
  • within-category correlations = consistently ^ > between category correlations
  • even when they removed voxels that showed ^ activation to each category (ie. FFA)
  • suggests ^ distributed architecture of ventral visual cortex
  • technique became known as multivoxel pattern analysis (MVPA)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

DECODING ALGORITHMS

A
  • uses machine learning
  • trains algorithm to distinguish 2 object categories on training set
  • test success of algorithm on distinguishing 2 object categories on test set
  • if algorithm succeeds > chance = brain region encodes info about object category
17
Q

KANWISHER: RESPONSE

A
  • even if brain region only represented dif between 2 faces, it might still make a dif response pattern to cars VS chairs
  • MVPA shows only info available via response given by brain region, not info that rest of brain reads
  • we want to understand neither mean response of region/info content of neural response; INSTEAD we want the causal role of neural response in beh
  • the only way to determine causal role of brain region in beh is to intervene on it
18
Q

PITCHER ET AL. (2009)

A
  • fMRI used to identify 3 cortical regions that respond selectively to faces/objects/bodies
  • TMS used during discrimination task involving 3 stimuli
19
Q

SCHALIK ET AL. (2010)

A
  • neurosurgical patient implanted w/electrodes along FFA
  • electrocorticographic responses showed selectivity to faces
  • electrical stimulation in FFA region produced illusory experience of seeing face (“facephene”
  • causal evidence that FFA = involved in face perception
20
Q

SUMMARY: PRO DISTINCT PERCEPTION MODULES

A

BEHAVIOURAL EVIDENCE
- holistic processing of faces
NEUROPSYCHOLOGICAL EVIDENCE
- prosopagnosia
FMRI EVIDENCE
- Fusiform Face Area (FFA)

21
Q

SUMMARY: ANTI DISTINCT PERCEPTION MODULES

A
  • expertise-related activation in FFA
  • activation in other brain regions to faces
  • developmental prosopagnosia
  • distributed patterns of activation to dif object categories in ventral visual cortex

Cognition & Emotion II (15 decks)

Brainscape helps you reach your goals faster, through stronger study habits.
© 2024 Bold Learning Solutions. Terms and Conditions