Read Mind Hacks™: Tips & Tools for Using Your Brain Online
Authors: Tom Stafford,Matt Webb
Tags: #COMPUTERS / Social Aspects / Human-Computer Interaction
Mind Hacks™: Tips & Tools for Using Your Brain (16 page)
Motion detection uses contrast information first, not color.
The moral of this story is that if you want people to see moving objects, make them
brighter or darker than the background, not just a different color.
Motion is important stuff for the brain. Information about movement gets routed from the
eye to the visual cortex — the final destination for all visual information — along its own
pathway (you can take a tour round the visual system
[
Understand Visual Processing
]
), the
magnocellular
pathway
. (Like a lot of things in neuroscience, this sounds more technical than
it is;
magnocellular
means “with large cells.”)
Color and form information travels along the
parvocellular pathway
(yup, “small cells”) to the visual cortex, which means any motion has to be processed
without access to that information. This functional division makes sense for a brain that
wants to know immediately if there’s a movement, and only secondly what exactly that moving
something looks like. Problems arise only when movement processing is trying to figure out
what sort of motion is occurring but the clues it needs are encoded in color and so not
available.
Stuart Anstis has constructed just such a problematic situation, and it leads
to the nifty stepping feet illusion
1
(
http://psy.ucsd.edu/~sanstis/Foot.html
; Shockwave). Blue and yellow blocks move smoothly in tandem from side to side.
Click the Background button to bring up the striped background, and look again. It should
look like
Figure 2-24
.
Even though they’re still moving in the same direction, the blocks now appear to be
alternately jerking forward, like little stepping feet. Like a lot of illusions, the
effect is stronger in your peripheral vision; fix the center of your gaze at the cross off
to the side and the stepping feet will be even clearer.
The easiest way to see why the stepping feet occur is to look at the same pattern, but
without any color — the yellow becomes white and the blue becomes black. Michael Bach’s
animation of stepping feet (
http://www.michaelbach.de/ot/mot_feet_lin
; Flash) allows you to remove the color with a click of the Color Off
button.
With no color, there’s no illusion: the moving blocks appear like stepping
feet even when you look straight at them. When the black (previously blue) block overlaps
a black stripe, you can’t see its leading edge so it isn’t apparent that it’s moving.
Given no cues, your motion processing centers assume no movement. Then as the black block
begins to move over a white stripe, you can suddenly see the leading edge again, and it’s
moved from where your brain had thought it was. That’s when you see the block apparently
jump forward and then move normally — at least until it overlaps the black stripe again. The
same is true for the white (previously yellow) block over white stripes, only it moves
when the black block looks still and vice versa.
So that’s what the blocks look like in black and white. Losing the movement
information of the leading edge over one stripe in two makes the blocks look like stepping
feet. And that’s what the motion-sensitive and color-insensitive magnocellular pathway
sees. The color information is added back in only later, reattached in the visual cortex
after the motion has been computed. In the end, you’re able to simultaneously see the
stepping feet motion via one pathway and the colors via the other.
Low-contrast patterns in general produce a less vigorous response from the
motion-sensitive parts of the brain,
2
which may explain why objects seen in fog appear to drift serenely, even
though they may actually be moving quite fast.
- Anstis, S. M. (2003). Moving objects appear to slow down at low
contrasts.
Neural Networks, 16
, 933–938. - Thiele A., Dobkins, K. R., & Albright, T. D. (2000). Neural
correlates of contrast detection at threshold.
Neuron, 26
,
715–724.
- Stuart Anstis’s publications online (
http://psy.ucsd.edu/~sanstis/SAPub.html
). - Anstis discusses the effect of contrast on motion perception (
http://psy.ucsd.edu~sanstis/PDFs/YorkChapter.pdf
).
Shading in pictures combined with the continuous random jiggling our eyes make can
generate compelling movement illusions.
We’ve all seen optical illusions in which parts of a completely static picture appear to
drift and swirl. One of the most famous examples is Professor
Akiyoshi Kitaoka’s rotating snake illusion (
Figure 2-25
), commonly passed around via email,
but, sadly, rarely with explanation.
http://www.ritsumei.ac.jp/~akitaoka/index-e.html
This is really a story about why you don’t see everything moving all the time rather
than about why you see movement sometimes when it isn’t there. Your eyes constantly move in
your head
[
To See, Act
]
, your head
moves on your body, and your body moves about space. Your brain has to work hard to
disentangle those movements in incoming visual information that are due to your movement and
those due to real movement in the world.
Another source of confusion for our visual system is a constant random drift in the
exact focus of our eyes.
1
This happens between saccades (see
Figure 2-5
, for example, in
To See, Act
). Our muscles are constantly sending little corrective signals
to keep our eyes in the same place. These signals never keep the eyes exactly still,
producing so-called
fixational movements
. This is a good thing. If
visual input is completely constant (i.e., if your eyes become paralyzed), the neurons in
the eye stop responding to the constant input (because that is what they do
[
Get Adjusted
]
) and everything fades
out.
One way your brain cuts down on confusion is shutting down visual input during
rapid eye movements
[
Glimpse the Gaps in Your Vision
]
.
Another mechanism is used to cancel out visual blur that results from head movements.
Signals from how your head is moving are fed to the eyes to produce opposite eye movements
that keep the visual image still.
Try this experiment. Hold the book in one hand and shake your head from side to side.
You can still read the book. Now shake the book from side to side at the same speed at
which you shook your head. You can’t read a word, even though the words are moving past
your head in the same way, and at the same speed, as when you were shaking your head. The
vestibular-ocular reflex
feeds a signal from your inner ear
[
Keep Your Balance
]
to your eyes in
such a way that they move in the opposite direction and at the correct rate to correct the
visual displacement produced by the movement of your head.
You can readily demonstrate that this is a reflex hardwired to your inner ear, rather
than a clever compensatory mechanism that depends on the motor signals you are sending to
shake your head. If you get a friend to move your head from side to side while you relax
completely (be sure your friend is careful and gentle!), you’ll see that you can still
read. This compensation doesn’t depend on your knowing to where your head is going to
move.
Normally your brain uses the structure of the current scene combined with the assumption
that small random movements are due to eye movement so as not to get distracted by these
slight constant drifts. To actually see these fixational movements, you have to look at
something without any structure and without any surrounding frame of reference.
We need to get a handle on various principles of vision and motion computation before
we can understand the rotating snakes illusion. Fortunately, each step comes with a
practical demonstration of the principle.
You will need a small point of light. A lit cigarette in an ashtray is
ideal — slow-burning, small, and dim enough not to illuminate anything else near it. Place
it at the other end of a completely darkened room so that all you can see is the light,
not the table it is sitting on or the wall it is in front of. Stand back at the other
end of the room from the light and watch it. You’ll see it start to move of its own
accord. This movement is due to the random drift of your eyes, which can’t be
compensated for by your brain because it has no frame of reference.
You can get the same effect by looking at a single star through a tube.
Without the other stars visible as a reference, it can look as if the single star is
dancing slightly in the night sky.
This
autokinetic
effect is famous for being influenced by
suggestion. If you’re introducing this effect to someone else, see if you can make him
see the kind of motion you want by saying something like, “Look, it’s going round in
circles” or “Hey, it’s swinging back and forth.”
So while we normally have these jiggly eye movements going on all the time, we use
the structure of what we’re seeing to discount them. But certain visual structures can
co-opt these small random movements to create illusions of movement in static pictures.
The rotating snakes illusion is one, but to understand the principle, it’s easier to
start with an older visual illusion called the Ouchi illusion, shown in
Figure 2-26
.
2
the design
Here the central disk of vertical bars appears to move separately from the
rest of the pattern, floating above the background of horizontal bars. You can increase
the effect by jiggling the book.
Your fixational eye movements affect the two parts of the pattern in different ways.
The dominant direction of the bars, either horizontal or vertical, means that only one
component of the random movements stands out. For the “background” of horizontal bars,
this means that the horizontal component of the movements is eliminated, while for the
“foreground” disk the vertical component of the movements is eliminated. Because the
fixational movements are random, the horizontal and vertical movements are independent.
This means that the two parts of the pattern appear to move independently, and your
visual system interprets this as meaning that there are two different objects, one in
front of the other.
The rotating snakes illusion (
Figure 2-25
) uses a different kind of
structure to co-opt these small random eye movements, one that relies on differential
brightness in parts of the pattern (color isn’t essential to the effect
3
). To understand how changes in the brightness of the pattern create an
illusion of motion in the periphery, see
Figure 2-27
.
In this simple pattern, the difference in the shading of the figure creates the
impression of illusory movement. It makes use of the same principles as the rotating
snakes, but it’s easier to work out what’s happening. Brighter things are processed
faster in the visual system (due to the stronger response they provoke in neurons
[
Why People Don’t Work Like Elevator Buttons
]
), so where the spokes meet, as one fades out into white and meets the
black edge of another, the white side of the edge is processed faster that the black
edge. The difference in arrival times is interpreted as a movement but only in the
peripheral vision where your resolution is low enough to be fooled. The illusion of
motion occurs only when the information first hits the eye, so you need to “reset” by
blinking or quickly shifting your eyes. It works really well with two patterns next to
each other, because your eye flicks between the two as the illusory motion in the
periphery grabs your attention. Try viewing two copies of this illusion at the same
time; open
http://viperlib.york.ac.uk/Pimages/Lightness-Brightness/Shading/8cycles.DtoL.CW.jpg
in two browser windows on opposite sides of your desktop.
You are now equipped to understand why Professor Kitaoka’s rotating snakes illusion
(
Figure 2-25
) works. Because the shape
has lots of repeating parts, it is hard for your visual system to lock on to any part of
the pattern to get a frame of reference. The shading of the different parts of the squares
creates illusory motion that combines with motion from small eye movements
that are happening constantly. The effect is greatest in your peripheral
vision, where your visual resolution is most susceptible to the illusionary motion cue in
the shading of the patterns. Your eyes are attracted by the illusory motion, so they flit
around the picture and the movement appears everywhere apart from where you are directly
looking. The constant moving of your eyes results in a kind of reset, which triggers a new
interpretation of the pattern and new illusory motions and prevents you from using
consistency of position across time to figure out that the motion is illusory.
of your eye
4
Professor Kitaoka’s web page (
http://www.ritsumei.ac.jp/~akitaoka/index-e.html
) contains many more examples of this kind of anomalous motion and his
scientific papers in which he explores the mechanisms behind them.
We are constantly using the complex structure of the world to work out what is really
moving and to discount movements of our eyes, heads, and bodies. These effects show just
how artificial patterns have to be to fool our visual system. Patterns like this are
extremely unlikely without human intervention.
Professor Kitaoka has spotted one example of anomalous motion similar to his
rotating snakes illusion that may not have been intentional. The logo of the Society For
Neuroscience, used online (
http://web.sfn.org
), appears to drift left and right in the corner of their web site! Now you
know what to look for, maybe you will see others yourself.
- Martinez-Conde, S., Macknik, S. L., & Hubel, D. H. (2004).
The role of fixational eye movements in visual perception.
Nature Reviews
Neuroscience, 5
, 229–240. - Figure reprinted from: Ouchi, H. (1977).
Japanese Optical
and Geometrical Art: 746 Copyright-Free Designs.
New York: Dover. See
also
http://mathworld.wolfram.com/OuchiIllusion.html
. - Olveczky, B., Baccus, S., & Meister, M. (2003). Segregation
of object and background motion in the retina.
Nature, 423
,
401–408. - Faubert, J., & Herbert, A. (1999). The peripheral drift
illusion: A motion illusion in the visual periphery.
Perception,
28
, 617–622. Figure reprinted with permission from Pion Limited,
London.