Lecture 5 Flashcards
object & face recognition
2 visual systems
1 identifies and 1 interacts
visual apperceptive agnosia
damage to the grouping stage of object recognition
visual associative agnosia
inability to recognise an object by sight - damage to the memory stage of object recognition
3 stages of object recognition
feature detection
grouping
recognition
stage 1 - feature detection
recognition of objects
detecting simple features, such as colour, motion, lines and orientation
visual apppercetive agnosia
damage to the LOC, inability to perceive objects through vision
patients with damage to or around area LOC can only detect simple features e.g. lines, the presence of closure but they cant perceive even the simplest forms
stage 3 - recognition
matching the perception to the representation of the object we hold in long-term memory
Area IT
neurones have 10 times larger receptive fields
cells are activated by combinations of complex forms combined with colours and textures
visual associative agnosia
damage to area IT in the anterior temporal lobe following stroke, cardiac arrest of head injury can lead to visual associative object agnosia, the inability to recognise a visual form as something
how do we recognise objects, regardless of changes in viewpoint?
two theories:
view - independent theories e.g. recognition by components
view-dependent theories e.g. view based
recognition by components
every object can be described in terms of simpler forms, called geons and their configuration
objects are perceived and stored in memory as collections of distinct geons in specific spatial configurations
therefore, recognition will be accurate regardless of viewpoint
occlusion
something infront, we can still see what the object is
parahipocampul place area
helps us get around the world, views/scenery
geons
simple shapes hat are put together to build objects, its not just which geons are included but also the configuration
non-accidental properties
decomposing objects into similar forms so we are able to recognise them, even from different viewpoints as long as NAPS are present
evaluation of RBC
- recognition of similar objects is hard to explain
- some objects are hard to define in terms of geons
- object recognition may take different amounts of time depending on viewpoint
- other features in addition to geons help us indetify objects
view-based theory
- recognise objects based on a few learned viewpoints
- recognition is the result of a match between the current view and the views of the same object already stored in memory
- recognition speed and accuracy will duped on the deviation of the percieved view from memory
which theory is correct
depends on the task and stimuli
categorical discriminations - view independent
difficult discriminations - viewpoint dependent
face recognition
recognised the face as belonging to a particular individual
face perception
realise that a visual stimulus is a face
are faces special?
we can identify different people, despite similarities
we can identify the dame person, despite drastic changes
face pareidolia
we tend to see faces in chance arrangements of objects and parts
inversion paradigm
help us to study how faces are processed
using upright and inverted faces and asking people to judge whether there is something odd or not - Thatcher illusion
inversion effect with other objects - yin 1969
ptps saw 2 pictures simultaneously, one that had been studied and one that was new
they had to select the one they had studied
inversion disrupted faces nut not object recognition
why are we invacuate at perceiving inverted faces?
objects are perceived and recognised on the basis of the shape of their component parts and their spatial configuration
but faces are perceived on the basis of the overall configuration of features, but the features themselves are not accurately perceived and represented
when faces are inverted, the configuration of features breaks down so we have to rely on the individual features
evidence from fMRI
area of the brain for facial processing known as the fusiform face area. maximally activated when seeing faces
Gauthier et al (1999)
ptps trained to name each greeble, a face like stimuli that each has the same part but they are arranged in slightly different configuration
Sanocki (1993)
objects in the world can appear in a infinite amount of orintations, sizes, shapes, colours and so on
looking at the global context and then the local context
prosopagnosia
inability to recognise faces
always brain damage in the right hemisphere and often damage to the homologous region of the left hemisphere as well
cells in the visual cortex
sensitive to various kinds of stimulus features sick as orientation and position within the visual field
