Vf: Sensory Module B Quiz #4

*One class of visual illusions

*Simple line figures that produce large errors in perception

*Extent (length)

*Alignment

*Size

*Angle

*Shape

Are named after the person(s) who discovered them

Geometrical illusions

Ponzo illusion

Muller-Lyer Illusion

Extent

Extent

Morinaga Illusion

Alignment

Poggendorff illusion

Alignment

Ebbinghaus Illusion

Size

Jastrow Illusion

Size

Zollner Illusion

Angle

Tend to not see the lines are paralllel but they are

Twisted Cord Illusion

Angle

Horizontal-vertical & bisection illusion

Shape

Vertical line looks longer but they are actually the same length

Shepard's Illusion

Shape

Looks like different shapes but actually the same shape

Yes, there are many studies published

Many variations

Still disagreement

Different illusions may have different explanations

These say illusions are caused by eye movement errors

Another version relates illusiosn to errors in "efferent readiness"

Illusions occur when presentation is too short for eye movements (if show in a flash, no eye movements occur but illusion still occurs)

But there is evidence that perception is not based on efferent readiness

*Illusions arise from low-level sensory processes such as lateral inhibition

Such processes may contribute to SOME illusions, but this theory can be rejected for some illusions, such as the Poggendorff

(Variants of the illusion provide evidence that it is not caused by acute angle expansion)

Processing of 3D scenes

We have 2 ways (or modes) of attending to pictures of 3D scenes

We can attend to information about the depicted 3D scene or we can attend to informatin about the flat surface of the picture

Yes

Some 3D scene context influences our perception of the picture surface this prodcues illusion

Some (but not all) of them

*There is some evidence they can't explain

*Several other theories have some support

*Many illusions may have multiple causes

The amount of light incident on a surface

(Albedo)

The percentage of incident light reflected by a surface

The amount of visible light that comes to the eye from a surface

Illuminance

Reflectance

The perceived luminance of a surface

The perceived reflectance of a surface (its constant property)

The tendency for perceived reflectance to remain constant in spite of variations in surface luminance

Weak effect

Has little if any effect

Its relation to the surrounding areas

Ratios of luminance remain constant as illuminance changes --> this provides a basis for lightness constancy

Ratio of light of object relative to ratio of frame work it is in.

Ratio of both is the same

Low level: "sensory", early stages of visual processing, usually farily local

IE lateral inhibition - occurs with center/surround receptive fields

Reduces effects of absolute level of lumiannce

Enhances local relations such as ratios and so contributes to lightness constancy

The same mechanisms that produce constancy can also produce illusions

IE background context is varied

Simultaneous contrast is the result

Lateral inhibition contributes to this illusion

No

*Perceptual

*Are at a higher level of processing, are more sophisticated

*Tend to operate over a wider spatial range but are still somewhat local (not the global scene)

Lightness constancy

Relative luminance

No

*The visual system uses several rules to determine anchoring

*One rule is that the highest lumiancne area is seen as white

*The lightness of other areas is determined by their ratio in relation to the anchor

*This can lead to illusions with a limited lumiannce range IE gray to black

Yes (IE small areas of high lumiancne may be seen as different luminate rather than white)

Local frameworks

Perceived lightness

*Koffka's rings

*White's illusion

Koffka's rings

The one on the left is mostly surrounded by white but looks lighter. Opposite for one on right.

Opposite effect of what expect with simultaneous contrast. Think of it as grouping with the ones on the side

Yes

Both squares are also affected by the more global white surround

Various factors

Increases

(IE more "articulation" of the local surrounds increases their relative weights and so that increases the perceived lightness difference between the squares)

Can look at this white area --> "global surround" is white. Also have an effect will roduce more equal gray because has weighing of local and global surround. compelx "weighing" going on

Gives local surround a greater weight so color differenciation of the two grays are more obvious

One thing in front of another if careful would say partial oclcusion (back not completely hidden)

Where contour meet = contour junction (The action is there)

Perception attempts to find the simplest or most probably organization

"B"

Only part of the circle is visible, but you perceive it as a partially occluded whole circle, this is called "perceptual completion"

Tend to see that occlusion because of T junction

Relatable

Unrelatable

Belongs to the surface that is seen in front (occluding)

Behind the occluding surface

Curved illusory contours

Illusory curvature from straight inducers

Getting the curved subjective controus because of the angles between the straight inducers

*Modal completion

*Amodal completion

2

Visual system tends to split into two objects (rings) when split up do both forms of perceptual completion

The part seen as in front is modal

Part seen as back is amodal

Self splitting objects

Ambiguous

Can either see horizontal or vertical in front. One seeing in front has (modal contour) one seen in back have amodal

Can get that to switch because the depth order is ambiguous

The relations between regions

No

(Only when bring several regions together can create transparency)

Two necessary conditions for perceived transparency with regions of different luminance

1) contrast is reduced by transparent surface

2) Contrast polarity is maintained by the transparent surface

Transparent Illusory Contours

(Looks like the vertical edges are continuous)

Neon Color Spreading

Yes

(This is another kind of optic array ambiguity that is resolved by environmental constraints)

Process that decides what kind of edge produced each contour

Four kinds of edges. This scene contains four different kinds of luminance edges:

*Orientation edges (O) due to abrupt changes in surface orientation

*Depth edges (D) due to gaps between surfaces at different distances

*Reflectance edges (R) due to different surface pigments or materials

*illumination edges (I) due to shadows

"+" = convex edge

"-" = concave edge

" -> " = occluding edge (object on right; left is empty space)

Same

No

That don't obey the rules

They don't normally occur in the real environment

Consistent labeling

Fairly local

So more global contour relations are less saliant

Different

Different

Reflectance (IE different areas are painted differently)

or

Differences in "shading" (IE different areas are illuminated differently)

perceived direction of illumination

(Top has more light and bottom is seen as in shade. If 2 patches send out same amount of light, the top is receiving more more light so has lower reflectance (because get more light and send out less light) additionally the lower one is reflecting a higher percent)

mid-level grouping

The weak illusion in this variant is hard to explain by perceived direction of illumination or by grouping. This componenet may be due to low-level lateral inhibition

When an object casts a shadow on another surface this produces a luminance change. A shadow border usually has a "penumbra" which helps distinguish shadow borders from reflectance borders. This can be used to create powerful lightness illusions