Sensation & Perception Quiz

Primary Depth Cues

Primary Depth Cues

Cues that we need to interact with our environment. Cannot be mimicked by drawings.

-Oculomotor Cues (Accommodation Convergence)
-Binocular Disparity (includes stereopsis)

Occulomotor cues

Convergence and Accommodation

Accommodation

A primary depth cue. The curvature of our eye changes based on distance. The visual system keeps track of how we need to strain and tells us that the more we need to strain, the closer an object is. Accommodation tells us useful information about depth at close distances.

Convergence

Our eyes turn into each other when we focus on an object. The angle in which our eyes are turned in changes based on distance. This angle is called the vergence angle. If we are looking at two objects in front of us, the object closer to us has a greater vergence angle.

Vergence Angle

Relates to Convergence. When eyes are turned in to focus on an object, it is the orientation of the eye when compared to looking straight ahead. Where the visual angles meet when eyes are turned in. Vergence angle increases when looking at closer objects.

Vergence angle near vs far

The vergence angle is greater when looking at closer objects because we have to turn our eyes in more.

Binocular disparity

Each eye sees the world from a slightly different perspective. Based on distance, objects project images to different points on the retina of each eye. There is more disparity in closer objects. Fun fact, your dominant eye’s view is closer to how you see with both eyes!

Remember that binocular disparity includes Stereopsis, horoptor (and corresponding and non-corresponding points).

Stereopsis

Relates to binocular disparity. The visual system creates information about depth (depth perception) based on binocular disparity.

Horoptor

An imaginary surface that passes through your fixation point

Angle of Disparity

Difference between the projection of a point in one eye compared to the other. If an object is closer to you than the horoptor, it has a crossed disparity. If an object is farther than the horpotor, it is uncrossed.

Corresponding points

Within the horoptor. Casts points on the same area of the retina in each eye.

Non-corresponding points

Everything outside the horopter. Casts points that hit different areas of the retina of each eye.

Binocular disparity tells us about

Direction: Disparity tells you whether points in the visual field are closer or further than the fixation point (or horoptor).

Magnitude: Disparity tells you how much further or closer an object is to the fixation point (or horoptor).

Pictorial/monocular/secondary depth cues

Perceptual qualities that we can replicate in a 3D rendering

-Size and Distance
-Familiar Size
-Relative Size
-Linear Perspective
-Height in the Plane
-Horizon and eye height (Includes Horizon Ratio)
-Texture Gradient
-Occlusion
-Atmospheric Perspective
-Shading

Anaglyphic images

Images that are created to be viewed with 3D glasses. Creates stereopsis (mimics and creates depth from binocular disparity by creating non-corresponding points) There are two slightly different offset views/images in a picture. Glasses filter one image to each eye using color filters.

stereogram

The old version of a view master. Creates 3D images. There are two slightly displaced images. One image is cast to one eye while the other is cast to the other eye. Need a stereoscope to see images.

-Random dot stereogram: images of random dots that when using a stereoscope, one can see a 3D image.

Random dot stereogram

images of a pattern of random dots that when using a stereoscope, one can see a 3D image.

Auto Stereogram

Images with random patterns that when you isolate each eye, you can see a 3D image.

Size and Distance

Pictorial depth cue. Perception of size and depth are related. We know that the size of an object on the retina changes with distance. The visual angle increases when looking at close-up objects. We use size and distance to perceive objects as having a constant size.

Linear Perspective

Pictorial depth cue. When looking into the distance, we see our environment as two lines that slowly converge.Hint: think the shining.

Height In The Plane

Pictorial depth cue. Objects that have a base (bottom) that is higher in the environment are perceived to be farther away. We receive information about an object’s relative height from this cue. Hint: think buffalo on the plain

Horizon and eyeheight

We use our eyeheight as a metric to judge the size of an object by comparing our eyeheight to where the object intersects with the horizon. The horizon will always intersect with eyesight.

Horizon Ratio

Relates to horizon and eyesight. To obtain information about depth/size, we take the amount of the object that is above the horizon/the amount of the object below the horizon.

-The ratio is the same for objects of the same size. Hence, if we are looking at telephone poles, the ratio should be the same for all of them-so we perceive differences in distance. When we cannot see the horizon, linear perspective creates one for us.

Relative size

If two objects are similar in size, we perceive the one that casts a smaller retinal image to be farther away.

Closer objects and visual field

Closer objects take up more of our visual field and create larger images on the retina

Further objects and the retina

Further objects take up less of our visual field and create smaller images on the retina

Atmospheric Perspective

Pictorial depth cue. The atmosphere interacts more with objects that are further away, making them appear more hazy while the atmosphere interacts less with objects that are closer, making them appear sharper.

Texture Gradient

When texture is consistent, patterns compress and become less visible with distance.

Occlusion

Pictorial depth cue. When objects are occluded, we infer that they are closer.

Shading

Gives us information about where objects lie in relation to each other. Tells us depth and distance. Uses many other pictorial depth cues such as occlusion and height in the plain.

Familiar Size

Our prior knowledge about the size of objects influences how we judge distance.

For example, if you see a car in the distance, you can estimate its distance based on your memory of how big cars typically are.

Size Constancy

The perception of an object’s size stays constant when the perceived distance and retinal image size stay change. Usually if an object moves further away, the perceived distance changes along with retinal size. However, this does not change how we perceive the size of the object

Perceived size = retinal size X perceived distance S= R X D

Ponzo Illusion/Railroad

Utilizes the linear perspective. Lines come together, the illustion is that a vertical line at the top of the image is larger than one that is drawn at the bottom of the image.

-Linear perspective can help explain this illusion. We see the train track as getting further with distance and see converging lines. Because the tracks appear to converge, the line at the top appears to be larger because it’s in the distance

-Texture Gradient: As distance increases, the texture becomes less visible. So when the line is positioned at the top, it appears longer due to seeing less texture.

-Size Constancy: P=retinal size X Perceived size. Retinal size does not change, but perceived distance does, making the perceived size appear different.

Monster illusion

A form of the Ponzo illusion with two monsters. The one in the back appears bigger.

-Linear perspective helps us to know that the monster in the back is further away since the lines are getting closer. Tells us the monster in the back should be smaller if it was the same size as the one in the front.

-Height in the plain: We know that the monster in the back is higher up in the plain.

-Texture gradient: Dot on the walls becomes more spread out in back of image, telling us info about distance.

-Size Constancy P= Retinal size X perceived size. Retinal size does not change, but perceived distance does, making perceived size different.

Perceptual Development

Babies do not have convergence or accommodation at birth. They can minimally use these primary depth cues at one month, but don’t get good at it until 3 months

Perceptual Development Stereopsis

Cannot use at birth to use as a depth cue, but is pretty good at it at 4-5 months

How to measure infant development of depth cues

Random-dot stereograms: Babies under three months cannot see, but those over three can use binocular disparity to see the image

Visual Cliff: A fake cliff created from plexiglass. Babies who have locomotive abilities avoid the cliff-around 6 months. Depth cues develop with motor
-With babies who are not moving (3 months) they were moved accross cliff and their heart rate lowered, shows unhabituation/interest because they don’t understand due to depth cues.

Babies and Perception of Objects

Babies group objects by similarity and common fate. Is measured by the length of gaze of an object. Longer looking at equals interest.

Babies and grouping by lightness similarity

3 months babies look at one image and then the other that differs by orientation/light bands. Babies experience dishabituation at three months due to lightness similartiy

Babies and common fate

4 months. There is a box and some rods. If two separate rods move together behind a box, babies see a single rod and experience dishabituation to seeing two rods. If the rods do not move, babies see two different rods and experience dishabituation to seeing one rod.

Babies and Separate Adjacent Objects

10 month olds The duck on the truck. If the truck is stationary, baby perceives the duck and truck to be one objects If duck is moving, baby then sees them as one two objects

Why do we have motion?

Motion is important for survival and establishes a figure from the ground. It is also practical: moving vehicles and motion in general draws our attention and helps us identify ambiguous stimuli

Conditions for seeing motion

Autokinetic effect

motion aftereffects

apparent motion

induced motion

Autokinetic Effect

When in a dark room and looking at a single dot, the light will appear to move at is is influenced by us moving our eyes and body.

Motion Aftereffects

After staring at an object that moves in one direction, a stationary object will move in the opposite direction.

-Waterfall motion. Moving dots then motion upward. When we see image of waterfall, our motion neurons fatigue and aftereffect occurs in other direction so that neurons can try to recover

-Budda effect/demo

Apparent Motion

In a sequence of static images, we see motion.

-Flip books, patterned static images where eye movements create motion

Induced Motion

When the background moves in one direction we misattribute the motion of an object to be moving in another direction

-Car seems to move forward when we are stationary

-The backdrop moves to make it look like the object moves

-At the train station, it might seem like one train is moving in one
direction when it is moving in the opposite direction.

-A cloud moves but it looks like the moon moves in the opposite direction

Real Motion

Real motion that occurs from objects. Not perception based.

Ames Room

People see this trapezoid room as a rectangle when looking through a peephole. One corner is ½ closer than the other, so one person is a bit closer to the observer. Problem with size constancy. The retinal size of one person decreased, but not the perceived distance. Both look just as close, but one looks bigger because retinal size changed.

Moon Illusion

when it comes to the sky, we assume that the the height above us taps out at some point. The heavens are flattened. We assume the distance to our horizon is further, even when the visual angle is the same for an object above-elliptical perspective. Because we assume the distance is greater, but the image does not change its size on our retina, we see the moon bigger on the horizon.

Emmert’s Law

Light on our retina has a fixed retinal image. But the afterimage changes with distance. When we look at something close, the after image is smaller and when we look at something far, the image is larger. We are using assumptions about distance.

Motion with 3D objects

Motion gives us structure about 3D images

Surface Segregation: Motion tells us what parts belong to one object and what parts belong to another

Dynamic occlusion: When an object blocks another object, we still know which objects are closer and that the object in the back still exists even when moving. We get depth information by “moving occlusion”

Motion Grouping: Allows us to group by what is moving and what is stationary

Kinetic Depth Effect

Structure from a moving 2D rendition helps us to fully perceive depth from a 3D object.

Wire projection: A wire projection rotates a 2D cube and participant sees the shadow. The movement helps the participant obtain detail about the sides. But when stationary, all you see is a square

Point light Walker: Motion cameras on a person. When 3D we can see it is biological movement and that it is 3D. Movement created by perceptual movement. Without movement, we just see a stick figure.

Implied Motion

Representational Momentum: in still pictures we perceive that motion will continue in the same direction in next frame

Motion system in the brain

Exists in the dorsal stream. The middle temporal (MT) is the primary motion area.

MT Neurons

As dot coherence increases (movement more in one direction) firing rate increases and monkeys can judge direction of movement more accurately

Mirror Neurons

Dorsal stream in parietal area. Responds when a monkey performs and action or watches someone else perform an action (but has to be an action they have experience with)

Audiovisual Mirror Neurons: Mirror neurons also fire when monkey hears an action that they have performed earlier. ex: opening peanut shell

Usefulness of mirror neurons

Observational learning-we imitate and understand others. Premotor activity in the brain when observing an activity.

Ballerina: experts in a certain type of dance had mirror neurons that responded higher to watching videos in their genre. Shows us that we have greater premotor activation to activities we are familiar with and do.

Inferential/Constructivist Perspective

Our Brain is a computer. Every time we move, our brain must make a whole set of new computations for each object.

Ecological Perspective

Gibson. Visual system does not need to make a whole set of new computations every time we move or every time light changes. We use an optic array to orient ourselves. It does not matter what the structure is, but how it changes as we move.

Optic Array

The structure of light-is consistent accross a given environment and when we move. We study the strucutre of light rather than each object’s retinal image. We don’t need to recalculate, but look for slight differences in strucutre. For instance, let’s say I am surrounded by trees. I use this structure of light coming through trees to orient myself.

Optic Flow

A pattern is created when movement causes elements of the optic array to flow past us. When we move towards objects, the optic array expands (objects get bigger). When we move backward optic array contracts (objects get smaller)

Invariant Information

Some information is constant even when we move (like ballerinas and spotting) We use this as a gauge rather than relying on retinal image.

Horizon Ratio: Horizon stays the same proportion. We scale objects to our body.

Swinging Room

13-16 months Tells us that optic flow is important for balance. When the room swings towards us (optic array expands) we move back to correct. When the room swings away from us (optic array contracts) we move forward to correct.