Chapter 8: Motion Perception
• Our visual system distinguishes the bug by a number of features, such as its
shape, its location in space, and its color. Previous chapters have established these
features as fundamental perceptual dimensions: characteristics of visual stimuli
that are directly encoded by neurons fairly early in the visual system.
• Motion aftereffect (MAE): The illusion of motion of stationary object that occurs
after prolonged exposure to a moving object.
• Just as colour aftereffects are caused by opponent processes for colour vision,
MAEs are caused by opponent processes for motion detection.
Computation of Visual Motion
• Because motion involves a change in position over time, a logical place to start is
with two adjacent receptors (call them neuronsAand B) separated by fixed
distance.Abug (or a spot of light) moving from left to right would first pass
through neuronA’s receptive field, and then a short time later it would enter
neuron B’s receptive field. In theory, a third cell that “listens” to neuronsAand B
• should be able to detect this movement.
However, out motion detection cell (call it M) cannot simply add up excitatory
inputs fromAand B. Given such a neutral circuit, M would fire in response to the
moving bug, but it would also respond to a large, stationary bug that covered both
receptive fields. To solve this problem, we need two additional components in our
neural circuit, as shown in Figure 8.3c. The first new cell, labeled “D” in the
figure, receives input from neuronAand delays transmission of this input for a
short period of time. Cell D also has a fast adaptation rate. That is, it fires when
cellAinitially detects light, but quickly stops firing if the light remains shining on
A’s receptive field. Cell B and D are then connected to neurons X, a multiplication
cell. This multiplication cell will fire only when both cells B and D are active. By
delaying receptorA’s response (D) and then multiplying it by receptor B’s
response (X), we can crate a mechanism that is sensitive to motion.
• This mechanism would be directed-selective: it would respond well to motion
from left to right, but not from right to left. The mechanism would also be tuned
to velocity because when the bug is moving at just the right speed, the delayed
response from receptorAand the direct response from receptor B occur at the
same time and therefore reinforce each other.
▯ • Now that we’ve constructed our neutral circuit, consider what happens when our
bug moves into cellA’s receptive field and then into cell B’s receptive field. Cell
Awill activate cell D, which after a short delay, will communicate its response to
cell X. If the bug is flying at just the right speed, it will move into cell B’s
receptive field just as cell D begins to fire. The signals from B and D will be
multiplied by X, and X will drive cell M, the motion detector, to fire. Note the M
will not respond to our large sedentary bug, because of the fast adaptation of the
D cell. Furthermore, M will respond to motion in only one particular direction
(left to right).Abug moving from cell B’s receptive field into cellA’s receptive
field would cause B and D to fire in the wrong order, so X would not receive its
two inputs simultaneously, and M would not fire. Out motion detector is also
velocity-sensitive: if the bug moves too fast or too slow, the outputs from B and D
will again be out of sync.
Apparent motion: The illusory impression of smooth motion resulting from the
rapid alternation of objects that appear in different locations in rapid succession.
• Apparent motion was first demonstrated by Sigmund Exner in 1875. Exner set up
a contraption that would generate electrical sparks separated from each other by a
very short distance in space and a very short period of time. Even though there
were two separate sparks – that is, two different perceptual objects – observers
swore that they saw a single spark moving from one position to another.
The Correspondence Problem
• Each movie has two frames that alternate back and forth. The only difference
between the two frames is that the only object in the movie, the red squares
studded with small circles, has been shifted diagonally by a short distance.
• The square moves down and to the right, and then back up and to the left, and
detectors sensitive to these directions pick up and signal this movement. Now
consider movie 2 (8.5b), where we cover most of the squares with a black “mask”
leaving three of the circles viewable through a small window. Beneath the mask,
the square moves exactly as before, but it you view the movie you will quite
clearly perceive up-and-down, not diagonal, motion.
• Corresponding problem: In motion detection, the problem faced by the motion
detection system of knowing which feature in frame 2 corresponds to a particular
feature in frame 1.
The difficulty for our motion
Detection system is this: how
Does it know which circle in frame
2 corresponds to which circles in frame 1? Because we have motion detectors for all directions, one detector will
sense the diagonal motion implied by matching the circle labeled “A” in Figure
8.5c with the circle labeled “C”. But another detector will sense the vertical
motion implied by matching circleAwith circle B. These detectors compete to
determine our overall perception.
• Aperture problem: The fact that when a moving object is viewed through an
aperture (or a receptive field), the direction of motion of a local feature or part of
the object may be ambiguous.
• Aperture:An opening that allows only a partial view of an object.
• Avariety of contours of different orientations moving at different speeds can
cause identical responses in a motion-sensitive neuron in the visual system.
Without the aperture, there’s no ambiguity and no problem. But when we view the
grating through the aperture, the system appears to impose some kind of shortest-
distance constraint, and thus the vertical-motion detector wins.
• Every V1 cell sees the world through a small aperture.
• The solution to this problem is to have another set of neurons listen to the VI
neurons and integrate the potentially conflicting signals.
• Only one direction – down and to the right - is consistent with what all four V1
cells are seeing here, and this is the direction we perceive when we see the object
as a whole.
Detection of Global Motion inArea MT
• Lesions to the magnocellular layer of the lateral geniculate nucleus (LGN) impair
the perception of large, rapidly moving objects. Information from magnocellular
neurons feeds into V1 and is then passed on to (among other places) the middle
temporal area of the cortex, an area commonly referred to as MT in nonhuman
primates. The human equivalent of MT has been localized using functional MRI
(fMRI), and variously labeled as MT+ or V5. Recent work suggests that this
motion-sensitive area may have at least two separate maps located on the lateral
surface at the temporal occipital (TO) boundary.
• The vast majority of neurons in the MT are selective for motion in one particular
direction, but they show little selectivity for form or colour.
• You might wonder why any motion detectors would fire in response to a
stationary rock, but as we will see in a bit, our eyes are constantly drifting
around, so there is always a small amount of retinal motion to stimulate motion
detectors at least slightly. • Interocular transfer: The transfer of an effect (such as adaptation) from one eye to
• The fact that a strong MAE is obtained when one eye is adapted and the other
tested means that the effect must be reflecting the activities of neurons in a part of
the visual system where information collected from the two eye is combined.
• The emerging evidence suggests that the MAE in humans is caused by the same
brain region shown to be responsible for global-motion detection in monkeys:
cortical area MT.
• Up to this point our discussion has focused on first-order motion – the change in
position of luminance-defined objects over time. In the nest section we describe
another interesting motion phenomenon: second-order motion, in which texture-
defined objects (also called contrast-defined objects) change position over time.