Sensation and perception: motion & depth Flashcards
1
Q
what is contrast sensitivity in the periphery?
A
- In another study, we presented participants with a sine wave vertical bar at the edge of their visual field and measure how well they could detect it (To et al. 2011).
- We used a two-alternative forced choice method and presented the bar either in the first interval or the second. Participants had to tell us in which interval the bar appeared.
- Participants had to tell us in which interval the bar appeared.
- If they got it wrong, we increased the contrast. If they cannot see it at 100% contrast, we recorded it as not visible No sensitivity.
- If they got it right, we decreased the contrast. We continued to decrease the contrast until participants were 50% correct (chance level). This 50% correct contrast was used as their threshold. We calculated sensitivity as 1/threshold.
2
Q
is there an increased sensitivity in the periphery when an object moves?
A
- You are less sensitive to STATIONARY stimuli compared to MOVING stimuli in the periphery
- In the far periphery, you can only see moving objects.
3
Q
how does the visual system detect motion?
A
- Many researchers have proposed models to explain this…
- The first one is the Barlow-Levick model (1965). This is an example of a detector for rightward movement.
- When something appears in front of a RECEPTOR. It trigger a * signal that travels down the pathway.
- The DELAY holds onto the signal for a bit before allowing it to continue its way.
- MOVING RIGHT only activates when it receives two * at the same time.
- The rightward motion detector is just one part of a larger system. Imagine similar detectors for leftward, upward, and downward movements, each responding to its own specific direction.
- Together, all these detectors cover all possible directions.
4
Q
why is that useful?
A
- Autonomous vehicles rely on sophisticated motion detection algorithms to navigate complex, dynamic environments where detecting the direction of moving objects is crucial for tracking and responding to them.
5
Q
how do robots used automation?
A
- Robots can use direction-sensitive motion detectors to navigate dynamic environments more effectively by understanding the direction of approaching objects or people and knowing how to avoid them.
6
Q
how do we track moving objects?
A
- When we follow objects with our eyes, this is called smooth pursuit.
- This is a voluntary eye movement.
- Masson and Stone (2002) showed that when we see an object, our eyes will initially track its features (colour, shape etc). Later after these features have been associated with the object, our eyes will track the whole object’s trajectory,
7
Q
what are saccades and fixations?
A
- When we read, look at a scene, or search for an object, we continually make eye movements called saccades, which bring our central vision to new regions of interest.
- In between saccades, our eyes remain still during fixations that last between 200ms and 300ms.
8
Q
what are saccades?
A
- Rothman et al (2006) showed that in a block stacking task, participants would saccade to the next block before the reaching movement was complete.
- This demonstrates that some of our saccades are planned and can support our movements/actions.
9
Q
what happens to vision during saccades?
A
- During a saccade, your eyes can reach a peak speed of 500deg/sec, and the movement would blur the retinal image.
- But we do not perceive this blur when we move our eyes. One of the main mechanisms that prevent this is called Saccadic Suppression.
10
Q
what happens in between fixations?
A
- Latour (1962) presented participants with two light sources (A and B) and asked them to fixate (look) at the one that is lit.
- The lights would light up alternatively so the participants would move their eyes back and forth.
11
Q
what do you see between fixations?
A
- Latour measured how sensitive participants were to a flash C (placed midway between A and B).
- Flash C could appear before, during or after the participants moved their gaze between A and B.
12
Q
what is saccadic suppression?
A
- Reduced visual sensitivity from 50ms BEFORE saccade
- Lowest sensitivity immediately before a saccade
- When you make a saccade, there is a period when your visual sensitivity is decreased and your vision is suppressed.
- This is called saccadic suppression and it prevents you from seeing blur when you move your eyes
13
Q
what is corollary discharge theory?
A
- When your eyes move, the signal sent (to the motor cortex) to move your eyes is accompanied by another signal that tells the brain (motion processing centre) that any movement across your retina is caused by your eye movements, not by the objects.
- When objects moves, their projection on your retina moves. This signals to your brain that an object is moving.
- When your eyes move, the projection of objects also moves on your retina, but you do not think objects are moving.
14
Q
what are the types of optic flow?
A
- When you move, the objects in your environment create something called optic flow.
- These flow lines help you determine the direction and speed of your movements.
- Optic flow is very powerful and can make you feel like you are moving even when you are not.
15
Q
what is vection?
A
- when you are on a stationary train and the adjacent train starts moving, you may feel that you are moving yourself. This is called vection.
- Feeling as though you are moving backwards when stand at the end of the water and the waves recede at a beach.
16
Q
how does vection work across the visual field
A
- Brandt et al. (1973) presented participants with lines that simulated optic flow.
- They showed that vection is strongest when the optic flow patterns are presented across the visual field and in the periphery.
17
Q
what are motion cues?
A
- Motion smear/blur (Burr, 1980) and motion lines oriented in the direction of the representation motion (Kawabe & Miura, 2006) can provide a compelling sense of dynamic movement in a still image.
18
Q
what is implied motion?
A
- We can extract the dynamic state of an object in an ephemeral posture in still images.
- Implied motion is so powerful that it activates areas of the brain that process actual motion (Kourtzi & Kanwisher, 2000)
19
Q
what is motion parallex and depth perception?
A
- When you move, objects at different distances shift their positions across your retina at different rates:
- Close objectsmove quickly across your field of vision.
- Distant objectsmove slowly or may seem almost stationary.
20
Q
what is motion parallex in 2D scenes?
A
- Animators make foreground elements move faster than background elements.
- They also need to account for changes in the speed, position and perspective of the player.
21
Q
what are depth cues?
A
- We normally rely on several cues for depth perception, but motion parallax becomes useful when objects are very far away.
- When you are closer to objects, we can use monocular and binocular spatial cues. Whilst monocular cues are effective with just one eye, binocular cues rely on both eyes.
22
Q
what is atmospheric perspective?
A
- Objects that are far away look more blurry and can have a blueish haze.
- Example: The mountains in the background are less detailed and have a blue tint.
23
Q
what is occlusion?
A
- Close objects hide objects that are further away.
- Example: Cliff is hiding some of the lake because cliff is closer.
24
Q
what is relative size?
A
- Objects are larger when they are closer.
- Example: Trees that are closer are larger than those further away
25
Q
what is elevation?
A
- Objects are lower in the image are generally closer compared to objects that are higher up.
- Example: Character is closer than the mountains
26
Q
what is texture gradient?
A
- The texture of nearby objects is more detailed than the texture of distant objects.
- Example: The water ripples are more visible closer compared to far away.
27
Q
what is shading?
A
- We can infer shape and location of objects by knowing how they are lit and the location of the light source.
- Example: From its shadow, you can tell that the little bush is round and not flat.
28
Q
how do we get shapes from shading?
A
- Light normally comes from above (e.g. sun), so we use this knowledge to interpret the shape of shaded objects.
29
Q
what is linear perspective?
A
- Parallel lines converge in the distance.
- Example: the sides of the road are parallel, but they approach each other in the distance.
30
Q
what are the binocular cues (vergence)?
A
- When you move your eyes towards a close object, your eyes converge.
- When you move your eyes towards an object that is further away, your eyes diverge.
- This vergence can offer some information on the distance of the fixated object.
31
Q
what are the binocular cues (disparity)?
A
- We can also infer the position of objects from differences between the retinal images of each eye. This is called binocular disparity.
- The 3D percept we get from this disparity is called stereopsis (solid vision).
32
Q
what are the two types of disaprity?
A
- When objects are located closer than your fixation, there is crossed disparity.
- When objects are located further than your fixation, there is uncrossed disparity.
33
Q
what is HOROPTER?
A
- The HOROPTER is an imaginary plane where all objects have zero disparity.
34
Q
what are depth perception applications
A
- Tesla’s Autopilot and Waymo’s self-driving vehicles combine depth information from multiple sensors to make real-time driving decisions.
- Other cars also have automated functions that can help you park!
