What are the two sides of the argument relating to what occurs in our 'blind spot' due to the lack of receptors in the optic disk region? As we do not actually perceive a 'black hole'.
- Vision 'fills-in' visual info in that region from info from surrounding regions (Ramachandran and Gregory)
- Visual system does not 'fill-in' - it simply doesn't represent those regions of visual space (hence we don't see anything there) (Dennett)
- Blind spot is peripheral region of retina with poor acuity so it's difficult to resolve this debate.
What physiological study may lend weight to the filling in hypothesis?
- Fiorani et al (1991)
- Single-cell recording
- V1 neurons found corresponding to blindpot region which fire in response to orientation of line overlapping the blindspot.
What is the study by Ramachandran and Gregory that seem to support the filling-in hypothesis?
- Breakdown of border coding in the 'twinkle' aftereffect
- Fixation on particular point in an image where a grey square is shown on a background of dynamic noise in periphery of image
- after a while, square replaced by 'twinkle' noise and even after the display is switched off, persisting patch of dynamic noise
Give neural evidence for filling-in.
- De Weerd et al (1995)
- Response of texture selective neurons in macaque V2 and V3
- Texture field containing hole is viewed in peripheral vision (artificial scotoma), hole appears to fill in with surrounding texture.
- V2 and V3 neurons with receptor fields covering hole increase in response (to a level comparable to that elicited by same texture without hole)
- Time course parallels the time course of perceived filling-in of hole by human observers
Give other examples of how the visual system fills in.
- Amodal completion: completion of an occulded border behind an occulder (Michotte 1964)
- Illusury/subjective contours/edges:(modal completion) Shape is perceived to be occluding other shapes even when the shape itself is not drawn.
- eg. white square overlapping in front of six grey circles, but certain parts of that white square are camouflaged by white background
[Onto the Visual Cortex]
Describe the seminal study which was the first ever recording of V1 responses systematically.
- Hubel & Wiesel (eg. 1962)
- Electrode inserted perpendicular to cortical surface in V1 of cat
- Cells in this colum tended to have similar receptive field locations, to prefer same orientation, to receive most of their input from same eye.
- However, when electrode placed parallel to cortical surface, all these would vary systematically, depending on which direction electrode moved.
- Orientation columns
- Hypercolumns: small slabs of cortical tissue 1x2mm which may be the basic unit of cortical processing
Tell me more about hypercolumns and what Hubel & Wiesel found.
- All cells in this colum have aprox same receptive field locations
- neighbouring hypercolumns coding neighbouring regions of visual space
- Hypercolumns varied along 2 dimensions parallel to cortical surface (in 1d - eye dominacne; in another d - orientation)
- Neighbouring groups of cells along this second dimension coded orientations about 15deg apart and this varied systematically - forming a 'pinwheel'.
- Cells showed differential preference to spatial frequencies within hypercolumn
So, just to recap, when we talk about hypercolumns, mention
- receptive field location
- eye dominance
- spatial frequency
- binocular disparity (range of cell within Hcolumn code different disparities)
Who discorevered 'blobs' within hypercolumns and what are they?
- Wong-Riley (1979)
- Cytochrome oxidase 'blobs' after staining
- Blob cells: (colour) high colour info, low selectivity for orientation or spatial frequencies
- Interblob cells: (form) low colour info, high spatial frequencies, and edges
Although not in lecture, briefly identify what the M pathway and P pathways are.
- Magnocellular: high contrast sensitivity, fast colour-blind, low resolution
- Parvocellular: colour-selective, high resolution, slow, low contrast sensitivity
- (CHECK if contrast sensitivity and resolution are same or not)
Look at this diagram, might be useful.
What are further subdivisions of cortical cells? (3)
- Simple cells: position, size orientation and size, RF mapped by light spots of light
- Complex cells: direction of motion, orientation not position, no response to light spots
- Hypercomplex cells: endstopped - respond to corners, angles or line-ends/bars moving in preferred direction
Hubel & Wiesel's discovery that some neurons in cortex respond only to oriented lines and others respond best to corners was important discovery. Why?
- Confirmed the idea that neurons respond to specific patterns of light and not to others (like with centre-surround receptive fields)
- Makes sense because purpose of visual system is to enable us to perceive objects in environment - many objects can be at least crudely represented by simple shapes and lines of various orientations.
What marks the border between V1 and V2 and the beginning of the extrastriate visual areas?
Line of Gennari
What is the main function of V2?
- Unlike V1 (which only codes features in the stiumulus), it goes 'beyond' retinal input
- Completion of amodally-completed shapes. Filling-in of camouflaged or occulded regions of space
- Integration across multiple hypercomplex/end-stopped cells? (Which codes for corners and lines etc).
- eg. Kaniza illusion
Main function of V5? Evidence?
- aka middle temporal area (MT)
- High-level motion processing
- 1. Patient L.M with akinestopsia (motion blindnes) after damage to V5 (Zihl, Crammon & Mai)
- V. interesting - how can you see the world in still images? Shows that motion is a construct of our visual system, not implicitly there in the stimulus.
- 2. Beckers & Zeki (1995) Large TMS pulse to MT (V5) causes motion perception to drop to chance
Main funciton of V4/V8? Controversies?
Damage around anterior V4/V8 causes acquired cerebral achromatopsia (cortical colour blindness)
- Specialised area for colour vision (more complex colour processing than V1, and colour constancy)
- Colour activates V4 more than achromatic greys (McKeefry & Zeki)
- Hadjikhani et al (1998)
- Claim it is a separate area - V8
Give 2 examples of how we are not conscious of early visual mechanisms,
- 1. Not aware of retinal size that object forms, just the brain's conscious interpretation
- 2. Amount of light - we only know interpretation of illumination vs lightness
Talk to me a bit about conscious vision's limitations with awareness of multiple objects at any one time (simultaneous vision).
- Phenomena such as change blindness show that we are aware of one or at most a few objects at a time.
- Change in stimulus --> perceptual 'transient' - heightens our awareness in location of change.
- But when there are too many transients in one scene, we fail to notice change
- Simons and Levin
However, there are times when we seem to be able to process multiple visual stimuli at the same time. When is this? What is the research paradigm called?
- Good way to measure is the visual search paradigm
- When an item differs from all items in terms of single unique feature that is processed at early stages of vision (eg colour, edge orientation), it perceptually 'pops out' from the others and is not affected by number of items - efficient, parallel search (simultaneous processing of multiple items)
Studies have show this parallel search happens not just for simple features, but for relatively high-level features. Give me two examples of these.
- Depth from shading - Ramachandran (1988)
- Completion process: (Rensink & Enns) Visual search for easily-detected fragments becomes difficult when the completed shape is similar to others in the display; conversely, search for fragments that are difficult to detect becomes easy when the completed shape is distinctive
However, what are 2 studies which show high-level features cannot be processed in this efficient parallel search way?
- Face configuration: Kuehn & Jolicoeur (1994)
- Letter identity: (look at Treisman & Gelade)
- Steep search slopes (big increase in RT with number of display items)
- Treisman: crucial feature NOT processed for all objects at once
- We have to pay particular ATTENTION to that item in order to see its high-level features
Sum up conscious vision
- Can process whole scenes at once in terms of simple features
- Conscious perception doesn't directly reflect early visual processing
- Our conscious perception reflects a limited amount of info selected by visual attention