Motion Perception

smooth pursuit eye movements

tracking a moving object using visual feedback

saccades

fast planned jumps (3persec)

brain suppresses visual info receives in between jumps which cause blurs

fixational eye movements

constant drift in eye position followed by micro saccades

Fraser-Wilcox Illusion/ Rotating Snakes

static images that appear to spin or pulse

microsaccades are not filtered out as the image is repetitive with asymmetric patterns of high and low contrast

=timing of visual signals is skewed so motion detectors are tricked into seeing its eye movements as motion on the image

Sherrington’s theory (eye muscle signal)

the brain compares retinal motion to the actual muscle movement signal

=even though the image on your retina moves, can perceive the world as not moving= no blur

Helmholtz’s theory (efference copy)

brain compares retinal motion to an efference copy- copy of command sent to muscle, not physical movement

poked eyeball situation

poking your eye makes the world appear to move

Sherrington predicts no motion as your eye muscles didnt move it

Helmholtz predicts motion as the retina moved but there was no efference copy to cancel it out so brain assumes world moved

paralysis situation

if eye muscles are paralysed and you try to move them the world appears to jump

sherrington predicts no motion as no muscle movement

helmholtz predicts motion as the brain sent command to eyes which generates efference copy but retina didnt move to cancel it out

Sutherland’s ratio model

claimed motion is based on pairs of opposing neurons

-fails to explain bidirectional adaptation (adapting to two diagonal motions at once)

Mather’s Distribution Shift Model

motion is based on population coding

adapting to a direction shifts the entire populations centre of mass

predicts that adapting to two diagonal directions result in a single unidirectional motion aftereffect opposite to the average of the adapted directions

how we detect motion (Reichardt Detector)

neural circuit:

bug flies over one receptive field

signal is held up by a delay (means that signal is direction and speed selective)

bug keeps flying and crosses second receptive field

triggers second signal

both signals travel to a comparator

comparator only fires if both signals arrive at the exact same time

aperture problem

neurons in V1/ primary visual cortex have very small receptive fields so when they look at a moving edge they can only perceive the motion as perpendicular to that edge

solving the aperture problem

MT/V5 integrates multiple local signals to work out the true global direction of the object

random dot kinematograms

test for global motion

V5 can detect the global motion of dots if 5-10% of them are moving together

dorsal stream development

develops late, making global perception vulnerable to conditions like dyslexia, autism and amblyopia

akinetopsia

damage to V5= motion blindness

patient LM had preserved spatial vision but saw static strobe like frames= required 90% coherence to detect motion