just in teractive : Perception Notes: Motion, Ramachandran and Geometry Wars

� Rough Design Sketches for Posture-Improving Chair | Main | Explaining Brain Training �

6 January, 2009

  Perception Notes: Motion, Ramachandran and Geometry Wars

In his book Art and Visual Perception (Amazon), Rudolf Arnheim applies Gestalt Psychology to, well, art and visual perception. He examines the underlying ways that we see, the way we pattern information and respond to forms and visual cues.

You can get a direct impression of the kinds of perception experiments Arnheim was talking about from the Gestalt Psychology page on Wikipedia; there are, as of this writing, a number of quick visual experiments that play with perception.

Arnheim talks about the artist's eye: seeing the curve of a snake and the curve of a coastline: making metaphor and seeing beyond immediate context. This kind of talk echoed a recent lecture I heard on synesthesia, but Vilayanur S. Ramachandran MD, PhD. People seem to like to say that artists make connections between disparate objects.

Also, Arnheim alluded to our ability to perceive objects in motion, distinct from our ability to perceive color and shape. Separate visual functions.

This too echoed a lecture I listened to from Vilayanur S. Ramachandran MD, PhD. Ramachandran is a neuroscientist and a popularizer of brain research. He's written a number of books, and in 2003, the BBC invited him to participate in the Reith lectures. The results, audio and text transcripts, are posted online, and they make fascinating listening.

I pulled out one bit, concerning disparate visual processing - fascinating stuff. Here's a transcript of what I played for the class:

We primates are highly visual creatures and it turns out we have not just one visual area, the visual cortex, but thirty areas in the back of our brains which enable us to see, perceive the world. It's not clear why we need so many, why do you need thirty areas, why not just one area? But perhaps each of these areas is specialised for a different aspect of vision. For example, one area called V4 seems to be concerned mainly with processing colour information, seeing colours, whereas another area in the parietal lobe called MT or the middle temporal area is concerned mainly with seeing motion. How do we know this? Well the most striking evidence comes from patients with tiny lesions that damage just V4, the colour area, or just MT, the motion area.

So for example, when V4 is damaged on both sides of the brain, you end up with a syndrome called cortical colour blindness or achromatopsia, and patients with cortical achromatopsia see the world in shades of grey, like a black and white movie, but they have no problem reading a newspaper or recognising people's faces or seeing direction of movement. Conversely if MT, the middle temporal area is damaged, the patient becomes motion blind. She can still read books and see colours but can't tell you which direction something is moving or how fast.

For example there was a woman in Zurich who had this problem, she was terrified to cross the street because unlike of us here, she saw the cars on the street not as moving but as a series of static images as though lit by a strobe light in a discotheque. She couldn't tell how fast a car was approaching even though she could read its number plate or tell you what colour it was. Even pouring wine into a glass was an ordeal; you and I gauge the rate at which the wine level is rising and slow down appropriately but she can't do this - so the wine always overflows. All of these abilities that seem so simple and effortless to all of us normal people -- it's only when something goes wrong we realize how extraordinarily subtle the mechanisms of vision really are and how complex a process it really is.

Now even though the anatomy of these thirty "visual" areas, the "seeing areas" in the brain looks bewildering at first, there is an overall pattern which I will now describe. The message from the eyeball on the retina goes though the optic nerve and goes to two major visual centers in the brain. One of these I'll call it the old system, the old visual centre, it's the evolutionary ancient centre, the old pathway that's in the brain stem and it's called the superior colliculus. The second pathway goes to the cortex, the visual cortex in the back of the brain and it's called the new pathway. The new pathway in the cortex is doing most of what we usually think of as vision, like recognizing objects, consciously. The old pathway, on the other hand, is involved in locating objects in the visual field, so that you can orient to it, swivel your eyeballs towards it, rotate your head towards it. Thereby directing your high acuity central foveal region of the retina towards the object so then you can deploy the new visual pathway and then proceed to identify what the object is and then generate the appropriate behaviour for that object.

Let me now tell you now about an extraordinary neurological syndrome called Blindsight discovered by Larry Weiscrantz and Alan Cowey at Oxford. It's been known for more than a century that if the visual cortex which is part of the new visual pathway, if that's damaged you become blind. For example if the right visual cortex is damaged you're completely blind on the left side if you look straight everything to the left side of your nose, you're completely blind to.

When examining a patient named GY who had this type of visual deficit, one half of the visual field completely missing, where he was blind, Weizcrantz noticed something really strange. He showed the patient a little spot of light in the Blind region. Weiscrantz asked him "what do you see"? The patient said "nothing" and that's what you would expect given that he was blind but now he told the patient "I know you can't see it but please reach out and touch it" The patient said well that's very strange - he must have thought this is a very eccentric request. I mean, point to this thing which he can't see.

So the patient said, you know I can't, I can't see it how can I point to it? Weiscrantz said well just try anyway, take a guess. The patient then reaches out to touch the object and imagine the researcher's surprise when the patient reaches out and points to it accurately, points to the dot that he cannot consciously perceive. After hundreds of trials it became obvious that he could point accurately on 99% of trials even though he claimed on each trial that he was just guessing. He said he didn't know if he was getting it right or not. From his point of view it might as well have been an experiment on ESP. The staggering implication of this is that the patient was accurately able to point to an object that he denied being able to see. How is this possible? How do you explain his ability to infer the location of an invisible object and point to it accurately?

The answer is obvious. As I said GY has damage to his visual cortex - the new pathway - which is why he is blind. But remember he still has the other old pathway, the other pathway going through his brain stem and superior colliculus as a back-up. So even though the message from the eyes and optic nerves doesn't reach the visual cortex, given that the visual cortex is damaged, they take the parallel route to the superior colliculus which allows him to locate the object in space and the message then gets relayed to higher brain centres in the parietal lobes that guide the hand movement accurately to point to the invisible object! It's as if even though GY the person, the human being is oblivious to what's going on, there's another unconscious zombie trapped in him who can guide the hand movement with uncanny accuracy.

Aaron spoke before me and wondered out loud about the impact of moving onscreen overstimulating on our visual processing units. I wondered how games work with both this new visual pathway, and the old as well -

I had just gotten an Xbox 360 the weekend before, and Aaron and I had played a bunch of Geometry Wars. It's a 2D game with tons of onscreen action - exploding colors, millions of blinking pixels. All this visual stimulation, and you're supposed to keep track of your moving ship in the middle of it, and the moving ships coming towards you.

Using one brain pathway to confuse the other? Or stimulating one and freeing the other? It's not clear to me how Geometry Wars exploits human perceptual patterns, but it's definitely giving them a good workout.

Posted by justin at January 23, 2006 10:53 PM

Trackback Pings

TrackBack URL for this entry:
http://interactive.usc.edu/mt/mt-tb.cgi/1763

Comments

Post a comment