Re-cognition Flashcards
Explain the Global Precedence Effect using Navon (1977) stimuli experiment:Pattern recognition
-lots of s letters making up a global/big H
-Global features are typically identified faster (H) than local features (s)
-Global precedence applies more to conscious perception rather than initial visual encoding
Perceptual organisation: Define Perceptual segregation
–Separating visual input/scene into individual objects where the thought to occur before object recognition
What does Gestalt psychology include?
The law of Prägnanz (important but vague and hard to test):
“Of several possible geometrical organisations of one that will actually occur, possesses the best, simplest and most stable shape”
What 4 things does the law of Prägnanz state?
1.Law of proximity:objects that are close together are perceived to belong in the same group
2.Law of similarity:similar objects are perceived to belong in the same group
3.Law of good continuation
4.Law of closure
What’s Figure-ground segmentation?
Figure=object of central interest
Ground=background
What’s Figure-ground segmentation using Faces-goblet illusion?
■Ambiguous drawing that can be seen as two faces or as a goblet
■More attention is paid to the figure than the ground
■Several factors (e.g., colour, luminance) influence figure-ground segmentation
What 2 principles are used in Gestaltism in the real world?
1.Adjacent segments of any contour have very similar orientations
2.Segments of any contours that are further apart have slightly different orientations
■We use knowledge of real objects when making decisions about contours (what we pay attention/focus on)
What do Gestaltists assume about
figure-ground segregation?
-It’s innate BUT memory (i.e., object familiarity) is important
–Healthy controls identify regions including familiar objects as figure whereas amnesic patients do not (showing relevance of memory in FGS)(From Barense et al., 2012, Elsevier)
Give 3 advantages of the Gestalt approach
■It focused on key issues
■Most organisational principles
have stood the test of time (grouping laws are valid)
■Law of Prägnanz has proven
useful as observers strive for
simplicity in visual perception
Give 5 limitations to the Gestalt approach
■Underestimated role of
knowledge (in recognition/paying attention to objects)
■Descriptive rather than explanatory
■Largely based on 2-D drawings
■Little emphasis on complex
interactions among grouping laws (e.g. law of proximity/similarity)
■Too inflexible
How have theories on recognition advanced?: Explain Marr’s theory (it’s a progressive process to how we recognise/perceive things)
Primal sketch:
–2D description of light-intensity (in retina main info perceiving things) changes:Edges, contours and blobs
–Observer-centred(representation perceiving world from person perceiving it)
2½-D sketch:
–Incorporates depth and orientation of surfaces:Shading, texture, motion, binocular disparity, etc.
–Observer-centred
3-D model representation:
– 3-D object shape
– Viewpoint-independent/invariant (from knowing where things are from me and from objects next to it)
How have theories on recognition advanced?: Explain Biederman’s recognition-by-components theory
-Objects consist of combinations of geons (Geometric icons/36 basic shapes)
-Object recognition is viewpoint-invariant (can be identified from any angle) (Biederman & Gerhardstein, 1993)
–Bottom-up processes
Give aspects of the recognition by components theory
1.Aspects of an object invariant across viewing angles
2.Edge straight or curved
3.Contour concave or convex
4.Non-accidental properties can be detected even when part of the target object is concealed by other objects
5.Geons are constructed from non-accidental properties
Recognition-by-components theory: Give 3 strengths
■Good evidence for geons being
important in object recognition
■Evidence the identification of
concavities and edges are also of major importance
■Non-accidental properties can
facilitate viewpoint-invariant
object recognition
Recognition-by-components theory: Give 3 limitations
■De-emphasises top-down influences (e.g., context; expectations; knowledge) even though it influences perception/what you choose to pay attention to
■Fails to account for most within-category discriminations (e.g.,
“Is that a poodle?”)
■Some objects lack invariant geons but are easy to recognise (e.g., clouds)
What’s the influence of viewpoint on object recognition?
■Categorisation of objects (e.g., “Is that a dog?) doesn’t depend on viewpoint
■Identification of objects (e.g., “Is that a bulldog?”) does depend on viewpoint
■With learning, object recognition often becomes increasingly viewpoint-invariant
■With learning, viewpoint dependent and viewpoint-
invariant representations often
both formed in LTM (Tarr & Hayward, 2017)
Explain the Interactive-iterative framework (Baruch et al., 2018)
*There are complex interactions between top-down processes
and bottom-up processes in object recognition
*Hypotheses top-down influences
attention which then influences
bottom-up processes
*Top-down processes more
important with degraded (e.g.,
blurred) stimuli as level of attention required differs
Define Holistic processing
Faces are processed in an integrated manner; it is claimed this is less true of objects
Define the Face inversion effect
Inverted faces are disproportionately harder to recognise than upright faces
relative to objects (Bruyer, 2011)
Define the Part–whole effect
Recognition of a face part more accurate when presented within the whole face (Farah, 1994)
Define the Composite effect
Perceiving half a face is more difficult when it appears against a different complementary half
(Richler et al., 2011)
Define Prosopagnosia and the 2 types
Known as “face blindness”
■Acquired prosopagnosia:caused by brain damage
■Developmental prosopagnosia: no obvious damage (often there’s also impaired object recognition)
What did Busigny et al., 2010 find in Selective “face blindness”
■Patient with acquired prosopagnosia (GG) had intact object recognition for birds, boats, cars etc. but not faces (shows specific brain area which processes faces)
■Thus, GG had a face-specific impairment rather than general
impairment of object recognition
If face recognition involves different processes to object recognition, what do double dissocation do we expect to see?
(1) Some patients with intact face recognition but impaired object recognition
(2) Some patients with intact object recognition but impaired face recognition
■few individuals in category (1)
GG (Busigny et al., 2010); AW (Germine et al., 2010)
■More individuals in category (2)
Geskin and Behrmann (2018): 20% of developmental prosopagnosics had intact object recognition
■Findings suggest face recognition harder than object recognition because it involves finer
discrimination
What is the Fusiform face area (FFA)?
–In the ventral temporal cortex
–Often damaged in prosopagnosics (but face processing involves a brain
network so damage not limited to FFA)
–Generally more responsive to faces>objects HOWEVER FFA less responsive to faces than other objects in 20% of individuals (Kanwisher et al., 1997)
–Developmental prosopagnosics have less selectivity for faces in 12 areas (including FFA) (Jiahui et al., 2018)
What brain areas (networks) are activated during face processing? (Duchaine & Yovel, 2015)
1.OFA=occipital face area
2.pSTS-FA and aSTS-FA=posterior and anterior temporal superior sulcus face areas
3.temporal FFA=fusiform face area
4.IFG-FA=inferior frontal gyrus face area
5.ATL-FA=anterior temporal lobe face area
Define the controversial Expertise hypothesis: face perception
brain areas and processes allegedly specific to faces are also involved in processing ANY objects for which we possess expertise
What findings support the Expertise hypothesis?
Holistic processing found for objects of expertise (e.g., interpreting medical images)
Fusiform face area is sometimes more activated for objects where
individuals have expertise
(McGugin et al., 2014): fusiform expertise area?
Bruce & Young’s (1986) model on face perception: Define Structural encoding/Expression analysis/ Facial speech analysis/Directed visual processing
SE-Various facial representations and descriptions
EA-Inferring an emotional state from facial features
FSA-Speech perception can be aided by observing lip movements
DVP-Specific facial info may be selectively processed
Bruce & Young’s (1986) model on face perception: Define Face recognition units/Person identity nodes/Name generation/Cognitive system
FRU-Structural info about known faces
PIN-Info about individuals (e.g., their occupation, interests)
NG-A person’s name
CS-Contains additional info (e.g., that actors and actresses tend
to have attractive faces) influencing which other components receive attention
What were Bruce & Young’s (1986) findings in their model + face perception?
*Familiar faces should be easier to recognise>unfamiliar ones as more info available (e.g., face recognition units/person identity nodes) greater than predicted effects
*Separate processing routes for facial identity and facial expression (BUT expressive facial info can interfere with identity perception (Redfern & Benton, 2017))
*Personal info (e.g., occupation) should be retrieved before an
individual’s name (not always the case)
*Several forms of impaired face recognition based on brain areas damaged (Davies-Thompson et al. (2014) linked 3 forms of impaired face recognition to specific areas of damage as predicted)
Give 3 strengths of Bruce & Young’s (1986) model
■Extremely influential model
identifying several key structures and processes
■Developed a plausible serial
stage approach to understanding face processing and recognition
■Provided a detailed account of
differences in the processing of
familiar and unfamiliar faces
Give 5 Bruce & Young’s (1986) model
■Facial identity and facial expression not entirely independent
■Oversimplified; there may be multiple systems for facial expressions (e.g., emotional system may play a far more
integral role)
■Gaze perception should be included
■Assumption that name processing is last occasionally wrong
■Model omits initial face detection stage: is it a face?
How could recognition be viewed as an evolutionary tool?
-Cognitive abilities are limited
-Recognition saves time
-Recognition saves energy
-Recognition is to detect relevant info
What’s the importance of face recognition?
-Recognise the emotional state of the person
-Communication
-Theory of mind
Why are we interested in patterns?
Patterns reflect processes=Processes shape reality=Understanding processes makes us more adaptive= Mastering the processes is
technology
Explain Complex adaptive systems
-Simple unit
-Coordination between units creates behaviour (swarm intelligence)
-System might evolve under (artificial) evolutionary pressure
-The behaviour of the system emerges from the interaction between the numerous elements
-It’s a self-organising system (RECOGNISE THEN REACT TO PATTERN)
