Week 10 Motion Processing

Question 1

How does the Elaborated Reichardt Detector explain how motion is extracted?

Answer 1

1. We have two spatially opponent offset V1 simple cells that are connected through neural circuitry combined with multiplicative manner
2. One of the offset V1 cells goes through a time/temporal delay called Delta-T - cell will only activate when two signals arrive at the same time
   and left to right
3. The summation is the multiplicative output of the 2 V1 simple cells

Question 2

Why don’t we have additive summation cells?

Answer 2

If the cells had additive summation, all you would need is one signal in one cell but not the second motion cell, it would not be selective to motion processing.

Question 3

How does the Elaborated Reichardt Detector work with a moving stimulus?

Answer 3

1. A bar of light activates a simple cell in one location and as it moves across the receptive field, the time delay of the simple cell means activation in the second cell occurs at the same time
2. The time delay corresponds to how long it takes the stimulus to move from one receptive field to the second receptive field

Question 4

How do we change the speed tuning of the cells to be tuned to faster stimuli?

Answer 4

1. We can change the physical distance between the receptive fields to be further away go a shorter distance/faster speed in the same time (faster speed), closer together for a lower speed
2. Vary the temporal delay - Decrease the time delay on the first V1 cell to occur in a faster speed

Question 5

Where does motion processing first occur?

Answer 5

Motion processing first occurs in V1

Question 6

Consider a complex cell tuned to vertical orientation and left-to-right motion. When will it not fire?

Answer 6

1. If there’s no change in light falling on the cell, the cell will not fire
2. If the stimulus moves in a way where light on that specific receptive field does not change, the cell will not be sensitive to it.

Question 7

What is the aperture problem and how to solve it?

Answer 7

V1 motion cells have small receptive fields, they pull the output of V1 cells to do higher level form processing

Question 8

What are two behavioural stimuli designs to test the aperture problem?

Answer 8

1. The global motion stimulus
2. The global gabor stimulus

Question 9

What is the global motion stimulus and how does it test motion processing?

Answer 9

Q. What’s the global/overall motion of moving dots?

Global motion = motion produced by signal dots over noise dots

Signal dots = moving in the same direction
Noise dots = moving in random directions

Question 10

What is the DV and findings of the global motion stimulus?

Answer 10

DV: threshold measure: the number of signal dots needed to determine what is the global motion

Findings - Average threshold is 6% (only need 6/100 dots to determine global motion)

Question 11

What is the Global Gabor Stimulus?

Answer 11

Gabor = A sine wave variation of luminance in a Gaussian envelope
Sinewave variation of luminance = little and dark bars
Global motion direction moves (in lecture simulation = down)

Question 12

How to link global motion processing to V5?

1. ANIMALS
2. HUMANS
3. NEUROIMAGING

Answer 12

1. In primates/macques = single cell recording, shows columns of cells in V5 tuned to the same direction of motion, (similar to V1 cells being tuned to same orientation)
2. Microstimulation = training animal on motion task ie indicate direction of motion, stimulating a particular column with specific motion orientation and measuring whether the animal becomes biassed to a specific orientation of motion after the microstimulation
3. Lesion studies + cognitive task - localised to V5 cortical area and compare ability to extract motion with a control task such as colour discrimination

HUMANS: TMS transcranial magnetic stimulation to create a temporary lesion in a specific region of V5 cortex to measure its role in motion orientation

NEUROIMAGING - fMRI, PET and MEG

Question 13

Where does local motion occur and where does global motion occur?

Answer 13

Local motion extraction = V1
Pooling of global motion = V5

Question 14

Why do we have a local motion pooling stage? Why do we have motion transparency in non locally balanced dots?

Answer 14

We have a local-motion pooling exchange to get

1. finer resolution of the motion signal
2. local averaging occurs before extracting global motion in V5.

Question 15

What is optic flow?

Answer 15

The pattern of motion produced on our eyes/retina that occurs due to our motion through the world

Produced because of Retinal motion: Every time we move our head or eyes or the world moves in relation to us - leads to motion across the entire visual field + complex patterns.

Question 16

What are the 3 types of retinal flow motions?

Answer 16

1. Translational motion - moving parallel to something (frontal parallel motion)
2. Radial motion - expansion or contraction: moving towards/away from something (most common)
3. Rotational motion - rotating/spinning moving forward/ around an axis

Question 17

Where do you get the illusion of induced self motion?

Answer 17

We experience the illusion of induced self-motion when we are looking at a moving stimuli that is dominant in our visual field, ie. a moving boat or train nearby, that we confuse with our own motion, we think you’re moving in the opposite direction/backwards.

This primarily occurs with big stimuli dominating your receptive field, because the visual system assumes that the big stimulus is part of the background and you must be moving in relation to it.

Question 18

Why is induced self-motion most effective with slower bigger things in motion?

Answer 18

This illusion best works with surrounding stimuli that move slowly around us, because when we are moving quickly we have other systems to rely on confirming our motion, including

1. The inertial impact of fast motion
2. The vestibular system where fluid fills around our ears

Question 19

Optic flow is important for maintaining balance and posture: what are some ways to show this from experiments?

Answer 19

1. Remove optic flow: close their eyes, stand on 1 leg, when we only rely on the vestibular system its harder to maintain balance
2. Provide wrong visual cues, ie. put person in a swinging room, ie. the vestibular system cannot be completely relied upon for posture because after spinning it keeps rotating, causing dizziness and the perception that you are still rotating

Question 20

Where is optic flow processed and what part of the dorsal stream does it receive input from?

Answer 20

1. Optic flow is processed in MSTd (medial superior temporal dorsal) cortical areas and tuned towards specific types of optic flow
2. Receives input from V5 which controls motion processing
   Has bigger receptive fields for processing large RFs of motion

Question 21

Are we more sensitive to expanding or contracting patterns of optic flow?

Answer 21

Expanding hypothesis =
Oblique effect, way more sensitive to cardinal directions than oblique. We are evolutionarily adapted to take in frontal vision to assess threats in the environment

We are more sensitive to contracting hypothesis = If outcome is more sensitive, a contracting pattern would be more helpful for having greater control to stop falling forward compared to backwards / maintaining balance

Question 22

How could we test the expanding/ contracting hypothesis of optic flow with the Global Motion Stimulus?

Answer 22

1. Modify global motion stimulus so signal dots are arranged in an expanding or contracting pattern around a bunch of noise dots
2. See which modified design produces the lowest threshold rate - this would indicate which arrangement we are more sensitive to!

Question 23

How to test whether there is a connection between optic flow and vestibular signals in global motion processing?

Answer 23

1. Expose participants to an optic flow stimulus while activating their vestibular system
2. Use an earthquake machine to replicate the effect of moving forward or backwards and measure their ability to detect an optic flow signal when the vestibular system was consistent or inconsistent with the optic flow stimulus.

Question 24

What do experiments on optic flow and vestibular signals indicate?

Answer 24

Our ability to detect the optic flow stimulus was enhanced when the vestibular system was consistent, indicating that optic flow processing and vestibular signals interact with each other

Question 25

What causes motion sickness and how to avoid it in two ways?

Answer 25

A cue conflict: the optic flow signal is inconsistent with the vestibular signals produced by the motion.

Solution!
1. We need to look forward to making the optic flow and vestibular system consistent

2. Hold up the phone/stimulus so you can see the outside world in your peripheral vision

Question 26

How do motion sickness glasses work?

Answer 26

The fluid of motion sickness glasses replicates the vestibular signals making it consistent with the optic flow of driving in a car

Question 27

What is the third cortical pathway hypothesis and where is it thought to be located?

Answer 27

1. Suggested to focus on processing motion for social relevant information like moving faces, eye gazes discrimination, biological motion, audio-visual integration of speech

Follow the ventral pathway in the ventral region of the brain as it is necessary for form processing