Brain & Behaviour Essay
Vision (1) � Visual cortical areas
3 May, 2000
Prof Rolls
The early empiricists na�ly assumed perception to be a fairly simple task of extracting the components of an image and forming them into objects we recognise. Such a view was banished by the rigours of attempting to model such a process computationally � the problem is very much harder. Even from the point that the photoreceptors in the retina are first stimulated, the visual system is extracting and processing the information � a passive image, no matter how rich and detailed, is of little use to us. In order to transform these �tiny, distorted, upside-down images� (Gregory 1966) into the three-dimensional mental constructs we see, we have to construct this visual representation from �unconscious inferences� (Rock, 1984) and ambiguous data.
Zeki first showed that the visual system seems to operate using perhaps three parallel pathways, which can be roughly characterised as being for analysing:
at least 3 parallel pathways in the visual system (Semir Zeki)
visual system does not simply record images passively like a camera. the visual system transforms transient lights stimuli on the retina into mental constructs of a stable three-dimensional world
how is this processing accomplished?
simple idea: visual perception is achieved by a single hierarchical system of cells processing ifnormation from the retina to the striate and extristriate cortex with receptive field properties that range from simple to complex and super-complex
there is a transformation of receptive field properties along a serial pathway, but how far does this pathway reach?
is there a group of cells that receives input from the complex cells and makes us aware of the toal image?
is there a special supercomplex cell group for each familiar object on top of the hierarchical processing
there may be further elaboration of recetive field properties along a serial pathway as a result of higher-order cells in the occipital, inferotemporal and posterior parietal areas abstracting the computational results of the striate cortex
but in addition to serial procesing, cells in different areas along separate pathways of the visual cortex may respond to different perceptual attributes of objects � motion, fofrm or colour
Gestalt = an integrated perceptual structure
or unity conceived as functionally more than the sum of its parts.
Gestalt
psychologists � brain creates 3D experiences from 2D images by organising
sensations into stable patterns (perceptual
constancies)
rational principles of shape, colour, distance and movement of objects in the visual field
visual pathways
V4 + MT
incl primary visual cortex
Rolls & Treves, chapter 8 � feedforward connectivity
1. self-organisation using Hebbian learning rule, and feed-forward connectivity
formation of concentric receptive fields peripherally, and simple cells in V1
2. centre-surround (lateral inhibition) origanisation �/span> contrast enhancement, removal of the mean, redundancy reduction
3. simple cells �/span> re-representation of the visual scene, makes explicit the oriented edges, used by feature combination neurons later
4. convergence in feedforward connectivity �/span> larger receptive fields
provides translation invariance, global motion, colour constancy etc.
5. modularity
Perception of Motion, Depth + Form pg 441
fast, transient and non-linear responses
more contrast gain
few colour opponent cells
sensitive to motion (with low velocity)
at equiluminance (when M cells may not respond), motion may disappear
(but motion detection may be based on either the motion orhogonal to the contour, or on the motion of the terminators)
slower, sustained and linear responses (e.g. centre-surround additivity, resulting in� no response with uniform illumination)
all P cells are colour opponent
cytochrome oxidase
colour-opponent cells, not very orientation sensitive
blobs receive inputs from P, M and K (koniocellular) streams
to V2 thin stripes to V4
V2 thin stripes are colour selective
V4
surface colour independent of the spectral composition of the illuminant
V2 interstripes: oriented, non-colour-selective cells, often endstopped
position-sensitive
not position sensitive
can be simple or complex cells
respond best to short lines
spatial relationship of the retinal photoreceptors is preserved in the striate cortex (retinotopic map)
at least 20 complete or partial representations of the retina in the extrastriate cortex
retina: 2 types of gangial cells M (Pa) and P (Pb)
not concerned with colour
it just adds the 3 types of cones together
� the magnocellular layers of the LGN
neurons respond rapidly but transiently
relatively unsensitive to colour
bad for contours/borders of colour contrast (equiluminance)
specialised for detecting object motion and 3D object organisation
limited depth perception
poor for analysing stationary objects
where not what
lesions: selective deficit in motion perception and in eye movements directed towards targets
distinguishes between the 3 types of cones
� parvocellular layers of the LGN
sensitive to orientation of edges
most information about shape is from borders, so this system is important for perception of shape
slowly-adapting, capable of high resolution (important for seeing stationary objects in detail)
what not where
lesions in the infero-temporal lobe �/span> deficits re recognition of objects (incl faces)
important for depth perception
link between the P/M layers in the LGN and the different retinotopic maps in the cortex
stained V1 for the mitochondrial enzyme cytochrome oxidase: precise repeating pattern of 0.2mm diameter peg-like regions � blobs
very prominent in the superficial layers 2+3
intervening lighter inter-blob regions
3 pathways from the layers in the LGN � striate cortex
in V2, instead of blobs: thick/thin stripes, pale inter-stripes
specialised for colour
arises from the parvo-cellular sub division of the LGN
-. blobs of layers 2+3 in V1 � V2 thin stripes � V4 � infero-temporal lobe
detection of form (+ colour)
P � LGN � 4Cb � inter-blobs of V1 layers 2+3 � pale stripes of V2 � V4 � infero-temporal cortex
motion and spatial relationships
M � LGN � 4Ca of V1, then layers 4B + 6 � V2 thick stripes � V3 � MT (V5) � MST + others in the parietal cortex
the 3 pathways inter-connect at various levels
psychological evidence for separate pathways carrying different visual information
V1 = striate cortex
V4 � colour, orientation of edges
V5 (MT) � primarily visual movement � depth + motion
7a � integrating somatic and visual sensations
parietal � visuospatial
perception of motion disappears at equiluminance
motion processed separately from colour, presumably by the magnocellular system, independent of the parvocellular system
perspective, relative size of objects, depth, figure-ground relations, visual illusions � also disappear at equiluminance, slso seem to be mediated by the magnocellular system
I will briefly talk about the neural system for motion, as
an example of why
The visual system needs to be able to deal with and
distinguish motion of the visual field (i.e. motion of objects in the
environment) and motion due to the movement of the head and eyes.
Motion in the visual field is
analysed by a special neural system. In fact only the evolved primates can
respond to objects that do not move. Frogs, for example, cannot even see
objects unless they are moving. They do not �see� in the sense that we do �
when their tongues flick out to catch flies, it is more like a reflex reaction
to small, dark, fast-moving objects. Motion in the visual field is detected by
comparing the position of images at different times � the visual system should
be able to compare the previous localtion of an object with its current
location by extracting the necessary information from the retina. This is
complicated by the fact that information about the direction of motion from a
small receptive field can be ambiguous. For example, the aperture problem (Movshon,
1990) demonstrates that if a grating of diagonal lines is moved either
downwards, sideways or perpendicular to the gratings, then it will always
appear to move in the same right-downwards direction � in order to be sure,
information from two separate local areas needs to be taken. The difficulties
involved increase with more complex objects and surfaces moving in three
dimensions. Problems like the aperture problem highlight the need for a more
complex solution, prompting researchers like Marr and Movshon to propose that
information about motion in the visual field is extracted in two stages. The
first stage is concerned with one-dimensional moving objects and measuring the
motion of the components of complex objects. The second stage involves
higher-order neurons combining and integrating the components of motion
analysed by several of the initial stage neurons.
Motion of the head and eyes
involves co-ordinating the vestibulo-ocular reflex, the continual
micro-saccades that we don�t even notice, and larger voluntary movements of the
head so that we don�t misinterpret these as huge movements in the visual world
around us.
the motion pathway originatesin the M-type retinal ganglion cells � they have no specific sensitivity to motion themselves, but they respond best to targets whose contrast varies with time
it may be that the segregation is not as clearly delineated as this
1. striate + extrastriate cortex
2. what is non-linear?
linearity
= e.g. centre-surround additivity, resulting in� no response with uniform illumination)
3. C = (P-T)/(P+T)
contrast, peak, trough
4. colour-opponent
5. equiluminance
equiluminant stimuli vary only in colour, but not in luminance, so if an image was converted to black and white, 2 equiluminant colours would be an indistinguishable grey
6. koniocellular (K) streams
7. �convergence in feed-forward connecitivity enables larger receptive fields to form�
8. translation invariance
9. global motion
10. �modularity (including topographic maps) built by short range excitatory connections and longer range lateral inhibitory connections�
11. centre-surround organisation �/span> removal of the mean
12. cytochrome oxidase