Visual Perception Flashcards
What happens after the retina - Parvocellular (P) Pathway
- Sensitive to colour and fine detail
- Most input comes from cones
What happens after the retina - Magnocellular (M) Pathway
- Most sensitive to motion
- Most input comes from rods
Pathway from eye to the brain
Retina –> Optic nerve –> Optic chiasm –>Lateral Geniculate Nucleus (LGN) –>Cortical area
- Left visual cortex comes from the
left sides of the two retinas, and signals reaching the right
visual cortex comes from the right sides of the two retinas
Property of Visual Neurons - Receptive fields
The region of the sensory space (i.e. retina) within which light will cause the neuron to fire.
Property of Visual Neurons - Retinotopy
Things that are near to each
other in space are processed
in cells that are physically
near to one another
Property of Visual Neurons - Lateral inhibition
A reduction of activity
(inhibition) in one
neuron that is caused by
a neighbouring neuron
* Useful for enhancing
contrast at edges of
objects
1st Stop - Lateral Geniculate Nucleus
Part of the thalamus – a subcortical relay for most of the brain’s sensory input and motor output
- Cells have a centre-surround receptive field
- Responds to differences in light across their receptive field (e.g. light in centre, dark in surround)
- Maintains a retinotopic map
- Correlates signals from the retina in space and time
- Provides the early 3D representation of space for action
Primary Visual Cortex (V1)
Extracts basic information from the visual scene (e.g. edges, orientations, wavelength of light)
- Sends this information for later stages of processing of shape, colour, movement, etc.
- Maintains retinotopy
Single-cell recordings by Hubel and Wiesel (1979)
- Indicate that some cells respond to simple features (e.g. points of light)
- Others combine those features into more complex ones (e.g. adjacent points of light may
combine into a line)
What is the result of damage to V1?
- Leads to cortical blindness - patient cannot consciously report
objects presented in region of
space - Patient can still make visual discriminations in ‘blind’ area - Blindsight (Weiskrantz, 1974)
- This is due to other routes between eye and brain
- Geniculostraite route may be specialised for conscious vision + others act unconsciously
Blindsight
Filling-in of ‘blind’ regions similar to filling-in of normal blind spot
Functional Specialisation Theory (Zeki, 1992-93)
Different parts of the visual cortex are specialised for different visual
functions
- V1 + V2: Early stage of visual perception like shapes
- V3 + V3a: Responsive to form (especially of moving objects)
- V4: Responsive to colour
- V5/MT: Responsive to visual motion
Central Assumption of FST
Colour, form, and motion are
processed in anatomically
separate parts of the visual
cortex
Brain Imaging (PET) of Human V4 and V5 - Zeki et al, 1991
V4 more active for coloured than
greyscale images –>specialised for
colour
* V5 more active for moving dots compared with static dots –>
specialised for motion
V4: The colour centre of the brain + Cortical achromatopsia
Patients with cortical
achromatopsia can’t see
colours because of damage to V4,
but often also due to damage to
V2 and V3 (despite a fully functioning retina)
* Case studies indicate intact implicit colour processing in patients with achromatopsia.
* Conclusion: V4 is involved in colour processing but the link between colour processing and V4 is not perfect
V5/MT: The motion centre of the brain + case study of LM
- Damage to V5/MT leads to akinetopsia
Patient LM:
- Bilateral damage to V5/MT
- Was good at locating stationary
objects
- Had good colour vision
- Motion perception was grossly
deficient
A Challenge for Functional Specialization– The Binding Problem
- Sighted people don’t perceive colour of things separately so how are different features bound to create coherent object processing?
Possible solution: Coherent perception depends on synchronised neural activity between brain regions, which my depend on attention
Beyond visual cortex: Parietal processing pathway
Also known as dorsal
Concerned with movement processing - vision for action
Beyond Visual Cortex - Temporal processing pathway
Also known as ventral
- Concerned with colour + form processing
Case study of DF - Vision for perception vs action
-Had a lesion on lateral occipital cortex - trouble locating and identifying objects
- But, conscious perception was different from information available to her motor system - had where but not what
Model of Object Recognition - Step by Step
- Early visual processing (colour,
motion, edges etc.) - Perceptual segregation: grouping of visual elements (Gestalt principles, figure–ground segmentation)
- Matching grouped visual description onto a representation of the object stored in the brain (called structural descriptions)
- Attaching meaning to object based on prior semantic knowledge
Perceptual Segregation
- Separating visual input into individual objects
- Thought to occur before object
recognition
Gestalt Psychology
Fundamental principle: the “Law of
Prägnanz”:
- “Of several geometrically possible
organisations, that one will actually
occur which possesses the best,
simplest, and most stable shape”
(Koffka, 1935, p. 138)
– Assumes a set of rules that operate early in visual processing
Gestalt Psychology: Problems
- Segmentation processes aren’t always bottom-up and following the laws of perceptual organisation
- Most evidence is only descriptive not explanatory
- Relies on introspection + evidence from 2D drawings
- Some segmentation clearly occurs via top-down prior knowledge
Object recognition deficits - Agnosia
- Impairment in object recognition
(without primary visual deficits) - Different kinds of impairments should arise depending on the
stage at which object
recognition is damaged
Apperceptive agnosia
– Impairment in constructing a perceptual representation from vision (e.g. grouping)
– Seeing the parts but not the whole
– Associated with lateral occipital lobe damage
Associative agnosia
- Impairment in the process which maps a perceptual representation onto knowledge of the object’s functions and associations
– Seeing the whole but not its meaning
– Associated with occipitotemporal damage
Case study of Apperceptive agnosia - HJA (Riddoch & Humphreys,1987)
Bilateral ventral-medial occipital damage
- HJA could recognise objects from touch but had a marked impairment in visual object recognition, particularly for line drawings over silhouettes.
- HJA had problems grouping or
organising information (e.g., recognising any objects presented together with other objects)
Case Study of Associative Agnosia - LH
- Had visual object agnosia + damage to occipitotemporal regions
- Preserved ability to copy drawings of objects, but unable to
name them or show what they are for- i.e. no access of semantics
Object perception: Caveats
- Most research comes from white male researchers + pps
- Westerners prioritise processing/categorising objects while East Asians prioritise relationships between objects + context
- Less activation of object perception regions in EAs during scene viewing
- Generalisable?
Different categories of objects
- Some evidence that
patients with damage to the
ventral visual stream (“what
stream”) have impairments in
only one particular category of
objects - Lesions to different areas of the
what stream can be associated
with agnosia for different types
of objects, like naming faces,
animals and tools
The Problem with Faces
Face recognition is a within-category discrimination – all faces look very similar
- Other object recognition is between categories – e.g. distinguishing a pen from a cup
- Faces require different types of
processing to other objects?
- Faces are Important from a
social/evolutionary perspective that they may have its own mechanism
Are Faces Special? - Neuropsychological Evidence
- Prosopagnosia - impairment of face processing that doesn’t come from damage to early visual processing
- De Renzi (1986): Patient failed to recognize his own family but
could do so by voice or clothes - Could match different views of faces and name other objects
- Impairment at the stage of matching to stored information
Are Faces Special - Neuroscience Evidence
Fusiform Face Area
- Part of the ventral (what) stream
- Responds to faces more than
other types of objects in
functional imaging experiments
Holistic Processing
- Features (parts) of faces are processed (and subsequently
remembered) less than for other types of objects, like houses
Holistic Processing in Face Recognition - Inverted Faces
- Sighted ppl identify inverted faces slower + less accurately
- Interpreted as evidence of processing
- Qualitative differences in processing of upright + inverted faces
- Spatial-relational information is
disproportionately affected by
inversion, and therefore face
recognition suffers
Faces are special due to Visual Expertise - Gauthier et al, 1999
- Suggest that faces are special as we are now experts at within-category discriminations
- Claims it involves fusiform face area
Criticism that we are Visual Expertise - Case Studies
Not all prosopagnosic patients are impaired on within-category discrimination
- WJ - had sheep and could distinguish between them
- RM - could distinguish between 5000 mini cars but couldn’t identify famous/his own/his wife’s face
