Test 3 Study Guide

COLOR SECTION

Be comfortable with the following terms: luminance, reflectance, lightness, power spectrum, spectral reflectance, color. Which ones are physical attributes that can be directly measured, and which are perceptual and not directly measurable?

Luminance

is the intensity of light at the eye, regardless of wavelength. It is a physical quantity and can be directly measured with a photometer.

Reflectance

is the proportion of incident light that a surface reflects. It is a physical property of surfaces and can be measured.

Lightness

is the perceived reflectance of a surface. It is psychological and cannot be directly measured.

Power Spectrum

is the intensity of light as a function of wavelength. It is physical and can be measured with a spectrophotometer.

Spectral Reflectance

is the proportion of light a surface reflects as a function of wavelength. It is physical.

Color

is the perceived spectral reflectance of a surface. It is psychological and does not exist in the external world.

What is lightness? Describe the inverse optics problem for lightness.

Lightness is the perceived reflectance of a surface. It is the visual system’s conclusion about how much light a surface reflects.

The inverse optics problem for lightness refers to the fact that the same luminance at the eye can result from many different combinations of surface reflectance and illumination intensity. Because the visual system only has access to luminance at the eye, it must infer surface reflectance by estimating how much light is shining on the surface and then discounting that illumination.

Describe the inference-like processes of lightness perception using displays like the checkerboard in shadow and others like it. (HW 6 is another example)

The inference-like process of lightness perception works as follows:

The visual system registers the luminance at the eye from different regions of the image.
The visual system uses scene cues to infer the illumination conditions (for example, whether a region is in the shadow).
If two regions produce the same luminance at the eye but one appears to be under weaker illumination, the visual system infers that the surface in shadow must be reflecting a higher proportion of the incident light.
As a result, the surface in shadow is perceived as lighter even though the retinal input is the same.

This shows that lightness is not determined by luminance, but instead is a perceptual conclusion based on an inference about illumination and reflectance.

What is the difference between illumination edges and reflectance edges? What role do they play in the perception of lightness?

Illumination edges

are boundaries where the change in luminance is interpreted as being caused by a change in lighting (such as shadow).

Reflectance edges

are boundaries where the change in luminance is interpreted as being caused by a change in surface reflectance (such as different colored paper).

Illumination edges signal changes in lighting and lead the visual system to discount luminance differences.

Reflectance edges signal changes in surface properties and lead luminance differences to be interpreted as differences in lightness.

Compare and contrast color to lightness. What is similar about them and what is different about them?

Lightness and color are similar in that both are psychological perceptions rather than physical properties, and both are inferred by the visual system through an inverse optics problem.

They differ in what they represent:

Lightness → is perceived reflectance that is constant across wavelength
Color → is perceived spectral reflectance, which varies across wavelength

Lightness depends on luminance and perceived illumination intensity, while color depends on the power spectrum of light at the eye and the perceived power of the illuminant.

Explain the difference between additive and subtractive color mixing.

Additive color mixing

occurs when lights are combined. Mixing lights adds wavelengths together, increasing the total energy reaching the eye (for example, red light plus green light produces yellow).

Subtractive color mixing

occurs when pigments are combined. Mixing pigments removes wavelengths because pigments absorb light, subtracting energy from the signal that reaches the eye.

What is the spectral content of white light?

White light contains all wavelengths of visible light in approximately equal proportions, producing a flat power spectrum.

What is the Law of Three Primaries (psychophysics) and how does it relate to Trichromacy (physiology)?

The Law of Three Primaries states that any visible spectral color can be matched by adjusting the intensities of three primary light sources. This psychophysical finding led to the hypothesis that the visual system has three classes of photoreceptors.

This hypothesis was later confirmed physiologically as trichromacy, meaning humans have three cone types with different spectral sensitivities.

Explain trichromacy. How are different wavelengths of light encoded neurally in the first stage of visual processing (i.e., the photoreceptors)?

Trichromacy

refers to the presence of three types of cones photoreceptors, each with a different peak sensitivity to wavelength (short, medium, and long).

Different wavelengths are encoded by the pattern of activity across the three cone types, not by a single cone alone. This is a form of population coding.

Explain what a metamer is and why it is that a trichromatic system (i.e., three cone types) has fewer metamers than a dichromatic system (i.e., two cone types).

A metamer

is a pair of stimuli that are physically different (they have different power spectra) but are perceptually identical.

A trichromatic system has fewer metamers than a dichromatic system because it has three cone classes with different spectral sensitivities. With more cone types, there are more constraints on how different power spectra can produce the same pattern of cone activation.

A dichromatic system, with only two cone classes, has less information about spectral differences and therefore more physically different stimuli collapse into the same percept.

What determines whether someone is Red/Green color blind versus Blue/Yellow color blind? Which type is less common and why?

The type of color blindness depends on which cone classes are missing or altered.

Red/Green color blindness

is typically caused by the M and L cones having very similar or overlapping spectral sensitivities, making it difficult to discriminate between red and green.

Blue/Yellow color blindness

results from problems with the S cone.

Blue/Yellow color blindness is less common because it involves the S cone, which is genetically and physiologically more distinct from the M and L cones.

Trichromats cannot know how a dichromat experiences color, why not? (Actually, as we’ve said, trichromats can’t even know how fellow trichromats experience color).

Color is a psychological experience, not a physical property of the world. Because perceptual experience is private, there is no direct way to access how another individual experiences color. Even if two people have the same cone classes, there is no way to know whether their subjective experiences are identical.

What can we know objectively about ours and other people’s color vision?

We can objectively measure which wavelengths people can and cannot discriminate using behavioral and physiological tests. We can also measure cone sensitivities and patterns of neural activity. However, we cannot measure or compare subjective color experience itself.

Describe color space. How is it different from wavelength which is the physical stimulus that gives rise to color perception?

Color space

is a perceptual (psychological) space that represents how colors are experienced. It is often described as circular and multidimensional (such as the color spindle).

Wavelength

is a physical dimension of the electromagnetic spectrum and is linear. Equal physical differences in wavelength do not correspond to equal perceptual differences in color.

What is the Law of Complementarity (psychophysics) and how does it relate to Opponency (physiology)?

The Law of Complementarity

states that for any spectral color, there exists a complementary color such that when the two are combined, the result is white or gray.

The psychophysical law led to the hypothesis of opponent processing, which was later confirmed physiologically.

Opponency

refers to the neurons that respond to differences between pairs of colors, such as red-green, blue-yellow, and white-black.

The discovery of the Law of Three Primaries and the Law of Complementary led to a raging debate as to what was the correct theory of color vision, trichromacy or opponency. We now know that both are correct. Explain how they are both correct.

Both theories are correct because they describe different stages of the visual system.

Trichromacy

occurs at the level of the photoreceptors, where wavelength information is encoded by three cone types.

Opponency

occurs at later stages, starting with retinal ganglion cells and beyond, where signals are recombined into opponent channels.

Together, these mechanisms establish internal representations of the proximal stimulus.

Based on what you’ve learned in this section of the course, defend the statement that ‘Color does not exist in the external world’.

Color does not exist in the external world because it is a physiological percept, not a physical property. The external world contains only electromagnetic radiation of different wavelengths. Color arises from the interaction between the spectral properties of light and the specific properties of the human visual system.

Explain under what circumstances we could perceive a piece of paper that reflects ‘blue’ light (i.e., light with more energy in the short wavelength range than in other parts of the spectrum) as gray.

A surface that reflects more short-wavelength light could be perceived as gray if the illumination itself is perceived as blue. If the power spectrum of the light at the eye matches the perceived power spectrum of the illuminant, the visual system infers a flat spectral reflectance and therefore perceives the surface as gray.

Refer to the cubes that are in the apparent yellow and blue light (see slides from class) and write out the inference like process that leads to Tile B appearing yellow.

The light reaching the eye from Tile B has a particular power spectrum.
The illumination in the scene is perceived as blue.
Given that illumination, the visual system infers that Tile B must be reflecting relatively more medium-to-long wavelength.
This inferred spectral reflectance is perceived as yellow.

What is going on with ‘the dress’ example? What are people who perceive it as white and gold doing differently than those who are perceiving it as blue and black? Be specific in your explanation.

The difference arises from different assumptions about the illumination.

People who perceive the dress as white and gold interpret the illumination as blue and discount short-wavelength.
People who perceive the dress as blue and black interpret the illumination as broad-spectrum or yellowish and discount long-wavelength light.

Different inferred illumination leads to different perceived surface reflectance.

MOTION SECTION

Explain how motion can be used to identify where things are relative to each other, where they are headed, and where we are headed.

Motion provides information about relative position and heading through changes in retinal location over time. Patterns of motion across the visual field, such as optic flow, indicate where objects are relative to the observer and where both objects and the observer are moving. This allows the visual system to infer direction of movement, speed, and future position.

Explain how motion supports object identification and categorization: shape, type/category, action, causation, intention/animacy.

Motion supports object identification by revealing structure from motion, allowing the visual system to recover shape. Different motion patterns provide cues about what an object is doing, which helps identify actions, causal relationships, and whether motion appears intentional or animate. Motion also aids perceptual organization by separating objects from backgrounds.

What is ‘motion’ as a (proximal) stimulus? In other words, what is the information that the visual system starts with for ‘establishing an internal representation’ of motion?

Motion, as a proximal stimulus, is systematic change in retinal location over time. The visual system starts with changes in where stimulation falls on the retina across successive moments in time.

How does the visual system solve the problem of distinguishing between change in retinal position due to object motion versus change in retinal position due to eye movements?

The visual system uses extra-retinal information, specifically corollary discharge, to account for eye movements. This information allows the system to discount retinal changes caused by eye or head motion and attribute remaining changes to object motion. Neurons in areas such as V3 respond differently to retinal motion caused by object movement versus eye movement.

Describe how a Reichardt Mechanism works. How does it code for direction and speed of motion?

A Reichardt Mechanism consists of two lower-order units with spatially separated receptive fields and different temporal delays feeding into a higher-order unit. The mechanism fires when signals from both lower-order units arrive simultaneously at the higher-order unit.

What is apparent motion?

Apparent motion

occurs when two stimuli are flashed sequentially at different locations, producing the perception of motion even though no object physically moves through space.

Explain why Reichardt mechanisms cannot distinguish between ‘real’ motion and apparent motion.

Reichardt mechanisms respond to patterns of spatiotemporal stimulation, not to whether motion is physically continuous in the world. Because apparent motions produces the same pattern of delayed activation across receptive fields as real motion, Reichardt mechanisms respond identically to both.

Compare a Reichardt Mechanisms to a binocular neuron (neuron that codes for retinal disparity).

Both Reichardt mechanisms and binocular neurons respond when signals from two spatially distinct inputs arrive simultaneously at a higher-order neuron. Reichardt mechanisms use temporal delays to code motion across time, whereas binocular neurons use spatial offsets between the two eyes to code depth through retinal disparity.

What is the motion after effect and how can it be explained in terms of motion opponency embodied in Reichardt Mechanisms?

The motion aftereffect

occurs prolonged exposure to motion in one direction, causing a stationary stimulus to appear to move in the opposite direction. This results from fatigue of motion-selective opponent cells. When one direction-selective Reichardt mechanism is adapted, its opponent dominates, producing the illusion of motion in the opposite direction. This effect is associated with neurons in area MT.

What is the correspondence problem? Explain how Quartet Motion is ambiguous with regard to correspondence.

The Correspondence Problem

is the problem of determining which elements in one moment correspond to which elements at the next moment in time.

In Quartet Motion

multiple correspondence assignments are possible, leading to ambiguous motion percepts (for example, motion perceived as vertical or horizontal).

What is Ternus Motion? Explain the basic display and what element and group motion are. Explain the correspondence relationships that are implied by group and element motion, respectively.

Ternus Motion

is a display in which three elements are shown, followed by a shifted version after a brief interval

Element Motion

occurs when correspondence is assigned to individual elements, making one element appear to jump.

Group Motion

occurs when correspondence is assigned to the entire configuration, making the group appear to move together.

Each percept reflects a different solution to the correspondence problem.

What evidence from Ternus motion experiments shows that spatiotemporal cues are used when resolving correspondence?

The dependence of perceived motion on the interstimulus interval (ISI) shows that spatiotemporal proximity influences correspondence. Short ISIs favor element motion, while longer ISIs favor group motion, demonstrating that timing and spatial continuity are used to resolve correspondence.

What evidence from Ternus motion experiments shows that feature cues are used when resolving correspondence?

When feature information biases correspondence toward either group or element motion, feature cues can override spatiotemporal cues. This shows that correspondence is influenced by similarity of features, not just space and time.

Explain the evidence that global-shape cues also play a role in solving correspondence.

Global shape cues influence correspondence when motion is perceived in ways that preserve the overall configuration of a stimulus. Demos show that correspondence can be resolved based on the global structure of objects, even when local spatiotemporal cues are ambiguous.

Explain the role of correspondence in the wagon wheel illusion.

The wagon wheel illusion occurs when correspondence is resolved incorrectly due to coarse temporal sampling. When sampling is too slow, the visual system assigns correspondence in a way that makes rotation appear slower or reversed, producing the illusion of backward motion.