Object recognition - week 3 (Chris) Flashcards
(16 cards)
Tanaka & Farah (1993)
Faces are processed holistically.
Face task: Learn to associate faces with names
House task: Learn to associate houses with names.
After learning, subjects asked to identify a) individual features or b) whole faces/houses.
Equally good in both conditions for houses.
For faces, aren’t as good at recognising individual parts than whole faces.
Recognise faces in a different way.
Rossion (2013)
The composite face illusion.
Eyes are perceived differently depending on the facial expression they sit on.
Are actually the same eyes.
Can see that when face is misaligned.
Prosopagnosia
Acquired deficit in face recognition after brain damage
Patients lose the ability to recognise friends and relatives
Also lose the ability to learn identity of new acquaintances
Patients can still recognise people by their voices
Cognitive skills and other visual abilities often remain intact
Rossion (2014)
Prosopagnosia can result from lesion in any region of right ventral occipitotemporal cortex
Right hemisphere dominance
Difficult to pinpoint a specific region that is always damaged in prosopagnosia
The Fusiform Face Area (FFA): a face processing ‘module’? Evidence from fMRI
Kanwisher et al. (1997)
Used a region of interest (ROI) approach
Functional localiser scan to identify face-selective voxels
Subsequent scans to test the selectivity of voxels to other stimuli and rule out confounds
Additional fMRI Evidence for other category selective processing modules in ventral visual cortex
The parahippocampal place area (PPA)
A region in ventral visual cortex that activates selectively to scenes
(Epstein et al., 1999)
The extrastriate body area (EBA)
A region in ventral visual cortex that activates selectively to pictures of human bodies
(Downing et al., 2001)
No other category of objects shows a selective pattern of activation in a circumscribed cortical region
Only biologically important stimuli seem to have dedicated processing modules
Challenges to the FFA module hypothesis
Expertise-related activation in FFA
Activation in other brain regions to faces
Developmental prosopagnosia
Distributed patterns of activation to different object categories in ventral visual cortex
Challenges to the FFA module hypothesis- Expertise-related activation in FFA
Gauthier et al. (1999) trained subjects to recognise novel objects (‘Greebles’) and found activation in FFA in greeble experts but not in greeble novices
Gauthier et al. (2000)
Showed bird experts and car experts pictures of birds, cars and faces
Stronger FFA activation to birds in bird experts and to cars in car experts
Evaluating the expertise hypothesis
Evidence for increased FFA activation for ‘expertise’ is weak and inconsistent – increases are small and several studies have failed to replicate findings
Greeble experiment confounded by similarity of stimuli to faces
Prosopagnosics can become experts at identifying other objects – e.g. case of a prosopagnosic sheep farmer who could recognise individual sheep
Part/whole behavioural effects are observed for faces but not for other ‘expertise’ objects e.g. dog experts
Challenges to the FFA module hypothesis
Developmental prosopagnosia
Impairment in face recognition that is not the result of a brain injury
Impairment is present from birth
Affects ~2.5% of population
Neural basis of developmental prosopagnosia still a matter of debate
Clearly no obvious pathology e.g. lesion in FFA
Functional imaging evidence inconclusive
Some studies have shown differences in activation/connectivity between developmental prosopagnosics and controls, whilst others have not.
Challenging the modularity of the ventral visual object recognition system – evidence from multivoxel pattern analysis (MVPA)
Univariate fMRI looks for ‘peaks’ of activation
MVPA fMRI looks for patterns of activation
Crucial distinction as MVPA can ask what information is represented in patterns of activation across a brain region
What can MVPA tell us that univariate analysis cannot?
MVPA can pick up on differences in the information represented in populations of neurons that show little sensitivity to such differences in univariate analysis
Haxby, J. V., Gobbini, M. I., Furey, M. L., Ishai, A., Schouten, J. L., & Pietrini, P. (2001)
Presented pictures of different categories (faces, cats, houses, chairs, scissors, shoes, and bottles) in the scanner
Subjects were scanned in 12 ‘runs’
Preprocessing of fMRI data included no spatial smoothing
For each run and for each category, measured activation in each voxel in the ventral visual cortex to each category of object
For each pair of categories, compared the within-category correlation and the between category correlation r values across pairs of runs.
Within-category correlations consistently higher than between category correlations for all voxels across the ventral visual cortex
This pattern of results remained even when they removed voxels that showed higher activation to each category, e.g. FFA
Suggests that patterns of activation across the whole of the ventral visual cortex contain information about the category of object someone is looking at.
In other words: We can predict, from patterns of activation across ventral visual cortex, what category of object someone is looking at.
Results suggest that different categories of objects (inc. faces, scenes etc) are represented in patterns of activity that are distributed across the ventral visual cortex, not in isolated modules such as the FFA
Pitcher, D., Charles, L., Devlin, J. T., Walsh, V., & Duchaine, B. (2009)
fMRI used to identify 3 cortical regions that respond selectively to faces, objects and bodies
TMS during a discrimination task involving faces, objects or bodies
Facephenes
Neurosurgical patient implanted with electrodes along fusiform gyri
Electrocorticographic responses showed selectivity to faces
Electrical stimulation in region of FFA produced illusory experience of seeing a face (“facephene”)
Causal evidence that FFA is involved in face perception
