Infering real world properties from retinal images, visual system examines statistical regularties in the world and uses them to disambiguate images to achieve form constancy
1 of 24
Similarity between colour and form perception
Visual system discards irrelevant information to get at real object qualities e.g. luminance and orientation
2 of 24
Example of form constancy
foreshortening; objects tend to produce shorter near vertical than horizontal retinal images because of foreshortening
3 of 24
Howe & Purves (2002)
3D laser scanner - vertical lines on the retina tend to correspond with longer objects in the world than horizontal lines - regularity
4 of 24
Form constancy
object doesnt appear to change shape as they change orientation, distance etc - gets rid of irrelvenat orientation information e.g. position, size
5 of 24
Why are Non human primates: Macacques used in research?
Similar visual system to humans however much smaller
V1-TE increase in complexity of processing (connections are bi directional) Increase in latency from V1 to TE (40ms - 100ms)
7 of 24
V1 - TE receptive field size
Receptive field size increases from V1 - TE (0.5 - 70)70 = Most of visual field
8 of 24
V1 (Motter 2007)
Centre is expanded, retinotopic space - adjacent points on the retina are represented by adjacent points on the visual cortex(no form constancy)
9 of 24
Visual cortex magnification (V1)
VC magnifys what is in the centre of vision - takes up more cortical teritory than stuff which in the periphery
10 of 24
V4
less active in simple patterns e.g. lines/bars/parallel more active in complex bars
11 of 24
TE (Tanaka et al 1992)
Simplification procedure - complex shapes - very active, less complex - active, simplified - nothing(starting complex then getting simplier and monitoring activation) VE - non retinotopic
12 of 24
Distibured representation of form in feature space
Different columns of the surface are activated depending on object (position and size invariant)
13 of 24
Human retinotopy (Eccentricty & Polar Angle)
Eccentricty - distance from a point to the centre of vision; Polar angel angle from horizontal
14 of 24
Eccenticity
Centre of visual field - mapped more posteriorally (V1) Periphery - more anteriorally (TE)
15 of 24
fMRI adaptation
Neurons get fatigued and respond less when stimulated over and over
16 of 24
No object representations
V1 & V2
17 of 24
Object representations
V4 & LOC
18 of 24
Human LOC = macauque TE
In humans everything is shifted posteriorally becuase the ITL has more complex cogntiion
19 of 24
How can we see if an area has form contancy?
If a visual area has form constancy then repeating the same objects in different sizes and viewpoints over and over should produce strong adaptation
20 of 24
Wilkinson et al (2002) Human V4
Not just for colour perception - paitent with damage to V4 unable to distinguish between differing radial patterns
21 of 24
No form constancy
v1 and v2
22 of 24
Little form constancy
V4
23 of 24
Full Form constancy
LOC
24 of 24
Other cards in this set
Card 2
Front
Similarity between colour and form perception
Back
Visual system discards irrelevant information to get at real object qualities e.g. luminance and orientation
Comments
No comments have yet been made