Adaptation II Flashcards
How does contrast adaptation work in V1?
When a preferred stimulus is presented repeatedly, neural responsivity is reduced (Movshon & Lennie 1979).
Following adaptation, some V1 neurones show an overall reduction in firing rate, akin to a form of neural fatigue.
Some show a lateral shift in their contrast response function, shifting response range towards adapted contrast.
Define selectivity in contrast adaptation in V1
Any responsivity changes depend on the neuronal response to an adapting stimulus, so contrast adaptation shows similar properties to the individual neuron’s response profile (e.g. orientation and spatial frequency tuning)
How do visual stimuli alter contrast sensitivity?
Repeated exposure to visual stimuli selectively alters contrast sensitivity, i.e. ability to detect patterns similar to adapting stimulus is impaired and there is no change in ability to detect different patterns.
What are the different aspects of contrast sensitivity?
Location specificity: adapting at one location doesn’t change sensitivity to stimuli presented at distant location.
Orientation tuning: elevation in contrast threshold falls off as difference between adapted and tested orientations is increased.
Spatial frequency tuning: contrast threshold elevation falls off as difference between adapted and tested SF is increased.
Partial interocular transfer: adapting to stimuli presented to one eye will affect contrast sensitivity measured in the other to a lesser extent.
Visual aftereffects - perceptual consequences
Contrast adaptation can also change the appearance of subsequently viewed suprathreshold stimuli.
Tilt-aftereffect: adaptation to particular orientation produces a repulsive bias in perception of nearby orientations.
Spatial frequency after-effect: adaptation to a spatial frequency produces repulsive biases in perception of those nearby.
Can contrast adaptation improve visual performance?
This adaptation shifts steepest part of contrast response function towards best contrast level. It allows those to better discriminate differences in contrast, and selective improvement here has been found in psychophysical studies.
Can CA improve efficiency at a neural level?
Adaptation equalises reponse level across neuron population, reducing metabolic cost of continued responses to regular features in environment.
How can contrast adaptation be used to learn more about visual processing?
Psychophysicists can use adaptation to determine whether complex visual aftereffects can be induced through adapting cortical areas beyond V1.
Adaptation to subjective contours
Von der Heydt, Peterhans & Gaumgartner (1984) - V1 neurones respond to real contours only, and 40% of V2 neurones respond to subjective contours (lines/edges perceived when no colour difference).
These produce tilt aftereffects similar to luminance contours.
Adaptation to shape (Suzuki 2001) and global form (Clifford and Weston 2005)
Found some evidence of distinct tuning properties with adaptation to shape (convexity and aspect ratio) induced tilt-aftereffects. Some tolerance to size/position change, broad orientation tuning.
Adaptation to global form likely involves mix of adaptation at multiple levels of form processing. Weak position invariance observed.
What are the principles of face adaptation ?
Adaptation to a distorted face makes original face look distorted in opposite direction. Face morphing approaches used to look at aftereffects e.g. Webster and McLean (1999)
What are some aftereffects seen in face adaptation? What are their properties?
Studies show that face distortion effects reflect mixture of low level (orientation/SF) and high level (face) adaptation processes (size tolerance, orientation tolerance and object specificity).
Define multidimensional face space
visual brain represents faces in different ways to basic image properties (e.g. orientation/SF).
norm-based representation: each face coded by deviations from prototype or average face on multiple dimensions. changes from given face to norm face on identity axis, with distinctiveness increasing the further away from the centre. Anti faces can be produced.
What are identity aftereffects in face adaptation?
Leopold et al., (2001) found adapting to anti-face makes average face look like original face. Recognition of original face improved but recognition of other identities was impaired.
Consistent with notion that adaptation shifts the norm in face space towards the adapted face.
Can normalisation account for properties of face adaptation that conflict with tilt/spatial frequency aftereffects?
Adaptation changes appearance of adapted face, making it look less distinctive.
Adaptation to neutral face doesnt alter appearance of distorted face. (Webster and McLeod 2011)
